Chiral Recognition in Cinchona Alkaloid Protonated Dimers: Mass

Jan 11, 2010 - ... Structures, EC project 15637) founded through the NEST (New and Emerging Science and Technology) program is gratefully acknowledged...
0 downloads 0 Views 866KB Size
Download PDF
3306

J. Phys. Chem. A 2010, 114, 3306–3312

Chiral Recognition in Cinchona Alkaloid Protonated Dimers: Mass Spectrometry and UV Photodissociation Studies† D. Scuderi,*,‡ P. Maitre,‡ F. Rondino,| K. Le Barbu-Debus,§ V. Lepe`re,§ and A. Zehnacker-Rentien*,§ Laboratoire de Chimie Physique, UMR8000, Baˆt. 350 and Laboratoire de Photophysique Mole´culaire, CNRS UPR3361, Baˆt. 210 UniV Paris-Sud, Orsay, F-91405, and Dipartimento di Chimica e Tecnologia del Farmaco, UniVersita` di Roma “La Sapienza”, P.le Aldo Moro 5, Rome I-00185 ReceiVed: October 1, 2009; ReVised Manuscript ReceiVed: December 17, 2009

Chiral recognition in protonated cinchona alkaloid dimers has been studied in mass spectrometry experiments. The experimental setups involved a modified 7T FT-ICR (Fourier transform-ion cyclotron resonance) mass spectrometer (MS) and a modified Paul ion trap both equipped with an electrospray ionization source (ESI). The Paul ion trap has been coupled to a frequency-doubled dye laser. The fragmentation of protonated dimers made from cinchonidine (Cd) and the two pseudoenantiomers of quinine, namely, quinine (Qn) and quinidine (Qd), has been assessed by means of collision-induced dissociation (CID) as well as UV photodissociation (UVPD). Whereas CID fragmentation of the dimers only leads to the evaporation of the monomers, UVPD results in the additional loss of a neutral radical fragment corresponding to the quinuclidinyl radical. The effect of the excitation wavelength and of complexation with H2SO4 has been studied to cast light on the reaction mechanism. Complexation with H2SO4 modifies the photoreactivity of the dimers; only evaporation of the monomeric fragments, quinine, and cinchonidine is observed. Comparison between the mass spectra of the cinchona alkaloid (CdQnH+) or (CdQdH+) dimers resulting from the UVPD of (CdQnH2SO4H+) and that of bare (CdQnH+) helps propose a fragmentation mechanism, which is thought to involve fast proton transfer from the quinuclidine part of a molecular subunit to the quinoline ring. CID and UV fragmentation experiments show that the homochiral dimer is more strongly bound than the heterochiral adduct. Introduction Besides the well-known importance of quinine as an antimalarial drug,1 cinchona alkaloids are endowed with numerous interesting applications in various fields of chemistry. As chiral molecules, they play an important role in asymmetric synthesis chemistry. Cinchona alkaloids are employed as modifiers of Pd catalyzers for the enantioselective hydrogenation of ketones or carboxylic acid’s substituted C-C double bond, with very large enantiomeric excess.2 Optical methods resting on vibrational spectroscopy have been used to characterize the conformation of the cinchona alkaloid and its modification upon molecular interaction.3-5 Enantiomeric separation has become increasingly important, especially in pharmaceutical industries. In this context, chiral stationary phases showing very efficient separation properties have been built on cinchona alkaloids.6,7 Cinchona carbamatetype stationary phases have been developed for separating aromatic acids or N-blocked amino acids. Effort has been put into studying the enantioselective recognition mechanism. The interaction between the cinchona alkaloid and the molecules to separate has been studied in detail by NMR, X-ray crystallography, as well as molecular dynamics.8,9 Recently, the role of solvation in the enantioselective recognition mechanism has been addressed by comparing the solution-phase enantiomeric preference of quinine carbamate toward amino acids with that †

Part of the “Benoît Soep Festschrift”. * Corresponding authors. E-mail: [email protected] (D.S.); [email protected] (A.Z.-R.). ‡ Laboratoire de Chimie Physique, Univ Paris-Sud. § Laboratoire de Photophysique Mole´culaire, Univ Paris-Sud. | Universita` di Roma “La Sapienza”.

observed under solvent-free conditions in an ESI-MS experiment, as measured by collision-induced dissociation (CID).10 Evaluation of the performances of chiral selectors like cinchona derivatives is one of the applications of mass spectrometry for studying chiral recognition.11 Indeed, the past decade has shown a flurry of studies aimed at studying the enantioselectivity of molecular interaction by measuring the relative abundance of diastereoisomeric ions,12 their collisional dissociation,13 or ion molecule reactions.14 Studies combining optical spectroscopy and ion traps have not been reported so far for diastereoisomeric complexes, in contrast with other biologically relevant systems, mainly amino acids and peptides, either in the IR15,16 or in the UV.17-20 Indeed, the only studies of diastereoisomeric complexes by UV spectroscopy deal with neutral species.21-26 We have therefore undertaken the study of protonated diastereoisomeric cinchona alkaloid dimers, produced by electrospray ionization (ESI) in a 7T FT-ICR mass spectrometer (FT-ICR MS) or a Paul ion-trap. Besides the measurement of the relative stability of the diasterosimeric complex by CID, the electronic and vibrational spectra have been obtained by UV photodissociation (UVPD) spectroscopy. The molecules under study, namely, cinchonidine (Cd), quinine (Qn), and quinidine (Qd), and for the sake of completeness, cinchonine (Cn), are shown in Figure 1. Cinchona derivatives show two basic nitrogen atoms, on the quinuclidine (pKA ) 9.7) and quinoline (pKA ) 5.7) rings.27 The monomer is therefore protonated in the quinuclidine ring. The gas-phase proton affinities of 983.3 and 953.2 kJ/mol parallel the pKA values.28 Cinchonine and cinchonidine, or quinine and quinidine, are pseudoenantiomers. This means that they differ only by the

10.1021/jp9094497  2010 American Chemical Society Published on Web 01/11/2010

Chiral Recognition in Cinchona Alkaloid Dimers

Figure 1. Scheme of the molecules under study and numbering of the atoms.

chirality of the two carbon atoms that link the quinuclidine and the quinoline rings. (See Figure 1.) The chirality of the three asymmetric atoms located on the quinuclidine ring is the same in both pseudoenantiomers. However, previous studies have shown that the two pseudoenantiomers behave like true enantiomers.10 Studying the dimers requires that one of the enantiomers be mass tagged, so that the mass of the molecule can be correlated with its absolute configuration29 and the corresponding mixture of the diastereoisomeric adducts can be mass resolved. Mass tagging has been achieved by substituting cinchonidine by a methoxy substituent on the quinoline ring, which results in quinine. The substituent is remote enough from the chiral center so that the enantiospecific interaction is expected to be the same for the two molecules. We will therefore consider the (CdQnH+) and (CdQdH+) dimers like homochiral and heterochiral dimers, respectively. In this article, we shall focus on the comparison of the fragmentation pattern obtained by CID and UVPD in the cinchona alkaloids dimers as well as on chirality effects. The investigated dimers are the homochiral (CdQnH+) or the heterochiral (CdQdH+) complex. Experimental Section Cinchonidine, quinine, and quinidine were obtained by SigmaAldrich and used without purification. We prepared 5 × 10-4 M solutions of the cinchona alkaloids by diluting 1 mL of stock solutions (5 mM) in 10 mL of H2O/MeOH 50:50 solution and adding 20 µL of formic acid 98%. CID-MS2 experiments were performed in a modified 7T FT-ICR mass spectrometer (Bruker Apex Qe) and a modified Paul ion trap (Bruker, Esquire 3000+). UVPD spectra have been recorded in the Paul ion trap. Details of the performances of the modified 7T FT-ICR mass spectrometer30 and Paul ion trap31 can be found elsewhere, and only important parameters for the present experiment will be discussed below. We generated ions in both cases by electrospraying the solutions under identical ESI conditions for all studied species. The ESI conditions used in the 7T FT-ICR MS were as follows: flow rate of 150 µL/h, spray voltage of -4000 V, drying gas

J. Phys. Chem. A, Vol. 114, No. 9, 2010 3307 flow of 4 L/s, nebulizer pressure of 1.5 bar, and dry gas temperature of 150 °C; whereas the ESI conditions used in the Esquire 3000+ were as follows: flow rate of 150 µL/h, drying gas flow of 5 L/mn, nebulizer pressure of 3.5 bar, capillary voltage of -4500 V, and drying gas temperature of 100 °C. For recording the CID-MS2 mass spectra of the homochiral (CdQnH+) and the heterochiral (CdQdH+) complexes in the 7T FT-ICR MS, ions were first mass selected in a quadrupole mass filter and fragmented through multiple collisions with Ar in a pressurized hexapole by using a DC offset on the order of -2 V. Ions were detected in the ICR cell at a background pressure of 10-9 mbar. For recording CID-MS2 mass spectra of the homochiral (CdQnH+) and the heterochiral (CdQdH+) complexes in the Bruker Esquire 3000+, precursor ions were mass selected in a 5 Da window, and fragmentation was obtained by collision with He gas inside the Paul trap by applying a radiofrequency (RF) excitation of variable amplitude during 150 ms. In what follows, we define [Fi] as the abundance of the fragment ions Fi produced by UVPD and [P] as that of the parent ion P. The fragmentation yield QFi for a given fragment Fi and the total fragmentation yield Qtot are defined by

QFi ) [Fi]/(Σ [Fi] + [P]) Qtot ) Σ [Fi]/(Σ [Fi] + [P]) The UVPD spectra presented here correspond to the plot of the total fragmentation yield Qtot as a function of the excitation wavelength. Specific fragmentation has been studied by measuring the yield QFi of a selected fragment Fi as a function of the excitation wavelength. A 1.2 mm wide hole drilled in the ring electrode of the Paul ion trap allows for the optical access to the center of the trap. The trapped ions are excited by a frequency-doubled dye laser (Sirah, Spectra Physik, containing DCM), pumped by the second harmonic of a Nd/YAG (GCR 190, Spectra Physik). Different focusing conditions have been tested for optimizing the overlap between the ion cloud and the laser beam. It eventually turned out that the best spectra were obtained with a nonfocused UV beam. The beam was simply skimmed by an iris to avoid unwanted electron scattering due to interaction of the beam with metallic parts of the trap. Nonfocusing conditions also limit the spectral density and thus saturation effects. Indeed, the study of the fragmentation efficiency as a function of the laser beam intensity has shown that saturation was already effective at low power (less than 1 to 2 mW), and the power has been kept as low as possible to minimize it. We have recorded the UVPD spectra by measuring the total yield Qtot at excitation wavelengths separated by 2.5 nm, with an averaging time of 5 min. The laser power has been adjusted at each excitation wavelength so that it was kept at a constant value on the order of the milliwatt. Multistage mass spectroscopy was carried out using the standard Bruker Esquire Control software. UVPD spectra were obtained through MS2 experiments, where precursor ions were mass selected in a 5 Da window, the RF excitation amplitude was set to zero, and ions were irradiated for 1s. Results and Discussion Collision-Induced Dissociation of the Dimers. The CIDMS2 spectra of the (CdQnH+) and the (CdQdH+) diastereoisomeric proton bound dimers obtained in the 7T FT-ICR mass

3308

J. Phys. Chem. A, Vol. 114, No. 9, 2010

Scuderi et al.

Figure 2. CID-MS2 mass spectra of (a) (CdQnH+) and (b) (CdQdH+) recorded in the 7T FTICR mass spectrometer.

spectrometer are presented in Figure 2. Besides the strong peak at m/z 619 due to the proton-bound dimer, fragment ions appear at m/z 295 (CdH+) and 325 (QnH+ or QdH+), respectively. For both diastereoisomers, the main dissociation path under CID conditions is the breaking of the intermolecular bond leading to the two monomeric fragments. However, the ratio between the peaks at m/z 325 and 295 in (CdQnH+) and (CdQdH+) shows that the proton prefers quinine or quinidine to cinchonidine. Methoxy substitution of the quinoline ring slightly increases the basicity of the quinuclidine nitrogen, which seems at first look to be surprising because the two centers are remote from each other. Comparison of Figure 2a and 2b shows that fragmentation is by far more efficient in the case of (CdQdH+) than for (CdQnH+). The fragmentation yields may slightly vary when the experimental conditions are changed (collision energy, backing pressure, etc.), but it is systematically larger for (CdQdH+) than for (CdQnH+), which indicates a higher binding energy for the latter diastereoisomer. Similarly, the relative intensity of the m/z 325 (QnH+) or (QdH+) and 295 (CdH+) peaks is larger for (CdQdH+) than for (CdQnH+). Electronic Absorption and UV-Induced Photofragmentation. Comparison between CID and UV-Photodissociation. The chromophore of cinchona alkaloids is the quinoline ring, which is known to display close-lying weak nπ* and stronger ππ* transitions in the gas phase, in the range of 300-320 nm.32 However, the nπ* transition is so weak that only the ππ* transition is observed in the condensed phase as well as rare

Figure 3. UVPD-MS2 mass spectra of (a) (CdQnH+) and (b) (CdQdH+) excited at λ ) 310 nm in the Paul ion trap.

gas matrixes.33 Here as well, we expect to see mainly the ππ* transition. Substitution lowers the ππ* transition energy, as observed for jet-cooled hydroxyquinoline;34 it is therefore expected that quinidine or quinine display lower transition energy than cinchonidine in the gas phase. As previously mentioned, protonation takes place on the quinuclidine nitrogen atom and is not expected to modify strongly the absorption spectrum. With this in mind, we have first set the excitation wavelength of the laser near the maximum of the ππ* transition of quinoline observed in the gas phase (λ ) 310 nm) and recorded the mass spectrum resulting from the fragmentation of the cinchona alkaloid dimers.

Chiral Recognition in Cinchona Alkaloid Dimers Figure 3 shows the results obtained for the (CdQnH+) and the (CdQdH+) diastereoisomeric complexes in the Paul ion trap. Measurement of the total yield Qtot of UVPD parallels the CID results; the (CdQdH+) complex (Figure 3b) undergoes more extensive fragmentation than the (CdQnH+) complex (Figure 3a). However, the fragmentation pattern obtained here clearly differs from what is observed in CID experiments. Whereas only protonated monomers (m/z 295 and 325) are observed in the CID-MS2 mass spectra, an additional fragment ion is observed in UVPD-MS2 spectra. This fragment is a radical cation at m/z 483 resulting from the loss of a neutral radical fragment (-136), corresponding to the quinuclidinyl C9NH14 radical, from the proton bond dimer. It thus seems that under UV irradiation, there is a homolytic cleavage of the C8-C9 bond. Interestingly, similar homolytic bond cleavages have been observed upon UV photoactivation of tyrosyl-containing peptides.19,35,36 The main photofragment observed after UV activation of mono- and diprotonated peptides was a radical cation resulting from the homolytic cleavage of the CR-Cβ bond of tyrosyl residues. The loss of quinuclidinyl radical raises several questions. The first one is whether the quinuclidinyl radical leaving group arises from the Cd or the Qd moiety. Second, it is not clear whether the charge of the m/z 483 fragment is borne by the protonated quinoline part, which is left from the quinuclidinyl radical loss, or by the other moiety of the complex. Last, the m/z 295 and m/z 325 fragments could result either from spontaneous dissociation of the m/z 483 fragment or from the direct dissociation of the complex, as observed in CID. The possible fragmentation processes are sketched in Figure 4a,b. To get closer insight into the UVPD process, UVPD/CID MS3 experiments have been carried out in the Paul ion trap where the m/z 483 photofragment was mass-selected and subsequently fragmented through CID with radiofrequency amplitude of 0.25. The UVPD/CID MS3 mass spectrum displays two CID fragments at m/z 295 and 325. Similar observations were made by comparing CID-MS2 and UVPD/CID MS3 mass spectra of tyrosyl-containing peptides.19,35,36 The b/y fragments, which are generally observed under CID conditions, were also observed upon collision-induced fragmentation of the radical cation resulting from the loss of tyrosyl radical. The abundance ratio (1:1) of the two m/z 295 and 325 fragments (Figure S1 in Supporting Information) may be informative and suggests that the homolytic C8-C9 bond cleavage occurs with a similar probability on both Cd and Qd in the proton bound dimer. The location of the charge in the protonated photofragment is not clear in this stage. The loss of neutral quinuclidinyl radical means that the UVPD process corresponds to the rupture of the C8-C9 bond in the R-position of the quinuclidine nitrogen, accompanied by a substantial rearrangement involving proton transfer from the quinuclidine nitrogen to another part of the complex. Proton transfer accompanied by a C8-C9 breakage is reminiscent of what has been observed in the ππ* state of protonated tryptophane, in which a proton transfer from NH3+ to the indole ring has been observed, followed by the rupture of the CR-Cβ bond.20 In tryptophane, the proton transfer was supposed to occur on the C2 or C4 carbons of the indole ring. In the cinchona alkaloids presented here, there are more favorable acceptor sites. A possible candidate could be the nitrogen atom of the quinoline ring of the excited species. The ring nitrogen of quinoline derivatives shows a dramatic increase in its basicity upon electronic excitation. Indeed the increase is on the order of 6 pK units in aqueous solution. For example, the pKA of 6-methoxyquinoline increases from 5.2 to 11.8 upon

J. Phys. Chem. A, Vol. 114, No. 9, 2010 3309

Figure 4. Scheme of the fragmentation processes at play in the studied cinchona alkaloid dimers, on the example of the (CdQdH+) complex (a) CID-MS2 of the (CdQdH+) complex, (b) UVPD-MS2 of the (CdQdH+) complex, (c) CID-MS2 of the (CdQdH2SO4H+) complex, and (d) UVPD-MS2 of the (CdQdH2SO4H+) complex. “Nu” stands for the neutral quinuclidinyl radical.

electronic excitation37 or that of phenylene-1,4-ethylenebis(4quinoline) from 4.4 to 11.05.38 The quinoline nitrogen of cinchona derivatives is thus expected to be more basic than that of quinuclidine in the electronic excited state, which would be in favor of the proton transfer to the electronically excited subunit. The monomeric fragments observed in CID-MS2 experiments at mass m/z 295 (CdH+) and m/z 325 (QnH+ or QdH+) are still present in the UVPD-MS2 mass spectra. The protonated monomers may arise either from direct dissociation competing with the quinuclidinium radical loss, or from a sequential process, as indicated in Figure 4b. Direct dissociation would consist of fast internal conversion (IC) from the S1 state to the ground state, followed by statistical dissociation from the hot ground-state potential energy surface, in a way similar to CID. However, in contrast with CID-MS2 experiments, where the m/z 325 fragment clearly dominates over the m/z 295 fragment, UVPD-MS2 spectra show a ratio of about 2 between the two monomers for both the homochiral and heterochiral complexes. This could arise from the fact that in addition to dissociation in S0 after IC, in a way similar to CID, dissociation toward the monomers can also happen on the S1 energy surface after or before proton transfer to the quinuclidine ring. This can also arise from the fact that sequential fragmentation takes place in the m/z 483 protonated cation. The m/z 483 cation resulting from the loss of quinuclidine could further react by loosing a second neutral quinuclidine fragment and a protonated monomer at mass

3310

J. Phys. Chem. A, Vol. 114, No. 9, 2010

m/z 295 (CdH+) or m/z 325 (QnH+ or QdH+). Experiments aimed at detecting the different photofragments in coincidence would probably cast light on the mechanism.39 Despite the overall fragmentation efficiency being more efficient for (CdQdH+) than (CdQnH+), no difference in the branching ratio to the different fragmentation paths (loss of quinuclidine vs evaporation of the monomers) could be clearly observed between the two quasi enantiomers. Dependence of the Fragmentation Efficiency upon the Excitation Energy: Action Spectra. The dependence of the photofragmentation efficiency has been studied as a function of the excitation wavelength to get better insight into the mechanism of the reaction. Figure 5 shows the UVPD resulting from the UV-induced dissociation of the (CdQnH+) (Figure 5a) and (CdQdH+) (Figure 5b) diastereoisomeric complexes toward the (CdH+) and (QnH+) monomers as well as the ion at m/z 483 corresponding to the loss of quinuclidine. The (Cd)2H+ (Figure 5c), the (Qn)2H+ (Figure 5e), and (Qd)2H+ (Figure 5d) dimers have been shown for the sake of comparison. All spectra presented in what follows have been recorded in the 300-330 nm range. Total Fragmentation Yield. For all systems studied here, the fragmentation yield is high. When increasing the laser power (in saturating conditions), a fragmentation yield close to unity can be obtained. The action spectrum of (Cd)2H+ (Figure 5c) recorded by measuring the total fragmentation yield is typical of a quinolinelike absorption. The spectrum of (Qn)2H+ (Figure 5e) bears strong resemblance with the spectrum of protonated quinine derivatives recently studied in solution;40 it extends over a broader region than that of (Cd)2H+. The spectrum of the mixed (CdQnH+) and (CdQdH+) diastereomeric complexes (Figure 5a and b) is the superposition of that of the subunits. The resemblance between the shapes of the action spectrum observed here and that of the absorption spectrum of the neutral subunit confirms the hypothesis of protonation on the quinuclidine ring, with little influence on the electronic transition located on the quinoline ring. Last, we have to notice that the spectrum of (Qn)2H+ (Figure 5e) extends much more toward the red than that of (Cd)2H+(Figure 5c). This effect might be related to conjugative effects due to the methoxy substituent in 2-position of the quinoline ring. The presence of the substituent in the two-position has been shown to shift down the energy of the first electronic transition also in the case of naphthalene derivatives.41,42 Loss of Monomeric Fragment. The loss of monomers extends over the whole absorption range and does not vary strongly with the excitation wavelength. Moreover, the intensity ratio between the (CdH+) and (QnH+) or (QdH+) fragments (Figure 5a,b) is more or less constant over the energy range studied here. This shows that evaporation toward a given monomer does not depend whether it has been electronically excited. This observation tends to confirm the hypothesis of fast IC to the S0 ground state, resulting in a vibrationnally hot complex that dissociates toward the monomeric subunits, like in CID. Loss of Neutral Quinuclidinyl Radical Fragment. We shall now discuss the efficiency of loss of quinuclidine fragment as a function of the excitation energy. In (Cd)2H+, the process seems to follow the trends of the absorption spectrum, as measured by the total fragmentation yield Qtot. (See Figure 5c.) In (Qn)2H+or (Qd)2H+, the process is much less efficient and is hardly observed as a background in the action spectrum. (See Figure 5d,e). In the mixed (CdQnH+) and (CdQdH+) dimers, the process starts at ∼320 nm, which matches the onset of the

Scuderi et al.

Figure 5. UVPD-MS2 spectra of protonated cinchona alkaloid dimers (a) (CdQnH+), (b) (CdQdH+), (c) (Cd2H+), (d) (Qd2H+), and (e) (Qn2H+) showing the total yield of fragmentation (9), fragmentation to protonated CdH+ (b), fragmentation to protonated QnH+ or QdH+ (2), and loss of quinuclidinyl radical (Nu) accompanied by proton transfer (1). (See the text.) The solid line is a graphical extrapolation of the experimental points. The error is estimated to be (0.05.

Chiral Recognition in Cinchona Alkaloid Dimers Cd absorption. Moreover, the fragmentation yield Q483 follows the profile of the Cd absorption spectrum. This observation seems to indicate that electronic excitation located on the (CdH+) moiety is necessary for the process to take place. The S0-S1 transition of quinine lies at a much lower energy than that of cinchonidine. It is therefore possible that after excitation of quinine, energy relaxation by fast IVR happens within the quinine moiety; there is not enough energy left in the reaction coordinate for the reaction to take place. The initial proton transfer must occur from the protonated quinuclidine to the quinoline ring of cinchonidine. However, we have to keep in mind that CID of the m/z 483 photoproduct shows that loss of quinuclidine happens on both moieties. This means that rearrangement of the complex accompanied by a second proton transfer toward quinine may also happen. Effect of H2SO4. Quinine and quinidine, as already reported in the experimental section, were obtained from Sigma-Aldrich and used without purification. Because they were purchased in the sulfate forms, the protonated complex between the cinchona alkaloids dimers and H2SO4 was easily formed under the electrospraying conditions used for the experiments. The interactions in these complexes are expected to involve the sites that also play a role in the proposed proton transfer mechanism, namely, the nitrogen atoms of the cinchona alkaloids. One expects then a difference in the reactivity of the dimers when they are complexed with H2SO4. We have therefore studied these complexes with the same experimental techniques, CID-MS2 and UVPD-MS2. Figure 6a shows the CID-MS2 mass spectrum of the (CdQdH+H2SO4) complex ion recorded in the Paul ion trap. Fragmentation of the isolated complex at m/z 717 has been obtained using a RF excitation amplitude of 0.5. It mainly shows dissociation toward the (CdH+) and (QdH+) monomers (Figure 4c). It can be deduced from this fragmentation pattern that no loss of H2SO4 or neutral Cd or Qd is observed. The neutral leaving fragment is always CdH2SO4 or QdH2SO4. This is an indication that H2SO4 is strongly bound to Cd or Qd. Figure 6b shows the mass spectrum resulting from the photodissociation of the (CdQdH+H2SO4) complex ion, isolated and excited at λ ) 310 nm. In contrast with the bare dimer, the H2SO4 complex does not show any fragment at m/z 483 resulting from the loss of quinuclidine molecule (Figure 4d). The only significantly observed fragments are the protonated monomers and dimer (CdH+), (QdH+), and (CdQdH+). The major difference between CID and UVPD is that the (CdQdH+H2SO4) complex ion undergoes loss of H2SO4 when excited in the UV but not when dissociated by collisions. The loss of quinuclidine in the excited state of the dimer seems to be prevented by complexation with H2SO4. To better understand the process, the (CdQdH+) dimer produced by photofragmentation of the (CdQdH+H2SO4) complex has been isolated and studied by CID by applying a RF excitation amplitude of 0.5. The observed fragments, shown in Figure 6c, correspond to the loss of water molecule (fragment at m/z 602), the loss of monomers (fragments at m/z 295 and 325), and the loss of two water molecules from the m/z 483 fragment (fragment at m/z 447). This pattern is very different from the CID-MS2 pattern observed for the (CdQdH+) complex. This shows that the structure of the (CdQdH+) dimer in the (CdQdH+H2SO4) complex differs from that of the bare (CdQdH+) dimer. Complexation between CdQd and H2SO4 must involve strong molecular interactions at the basic sites of Cd or Qd, either through the formation of an ion pair or through strong hydrogen bonding. It is therefore not surprising that the proposed proton transfer process, which involves the abovementioned nitrogen atoms, is modified by the presence of H2SO4.

J. Phys. Chem. A, Vol. 114, No. 9, 2010 3311

Figure 6. (a) CID-MS2 mass spectra of (CdQdH+H2SO4) obtained by applying a radiofrequency excitation amplitude of 0.5. (b) UVPD-MS2 of (CdQdH+H2SO4) excited at λ ) 310 nm. (c) CID-MS3 mass spectrum of (CdQdH+) resulting from UVPD of (CdQdH+H2SO4) at λ ) 310 nm, obtained by applying a radiofrequency excitation amplitude of 0.5. The CID-MSn and UVPD-MS2 mass spectra have been recorded in the Paul ion trap.

The conformation of cinchona alkaloids plays an important role in its biological as well as chemical activity, especially the rotation around the C8-C9 bond linking the quinuclidine and quinoline parts. This angle has been shown to depend on the state charge of the cinchona alkaloids and the presence of an interacting molecule or that of a counterion.5,6,43-45 The protonated form only exists in an open structure, which has been shown to accommodate an anion like carboxylate or chlorate

3312

J. Phys. Chem. A, Vol. 114, No. 9, 2010

very easily.45,46 The complex is a bidentate structure involving Coulomb interaction between NH+ and the anion and hydrogenbonding from the Cd hydroxyl to the anion. The structure of the (CdQdH+H2SO4) complex probably involves interaction between H2SO4 and the protonated nitrogen of quinuclidine, which in turn modifies the excited state proton transfer invoked to explain the photoinduced dissociation process. Conclusions Most of the work devoted to UVPD of protonated ions has focused on amino acids or peptides, which have attracted considerable attention.47 We have presented here results concerning another class of important biologically relevant molecules, namely, cinchona alkaloids. We have shown that UV excitation leads to a fragmentation pattern that strongly differs from that obtained by CID. CID of the photoproducts, dependence of the photoinduced fragmentation upon the excitation wavelength, as well as complexation of the (CdQdH+) dimer with H2SO4 have been studied to help us understand the mechanism of the photoinduced reaction. It is proposed that the driving force for the reaction is the dramatic increase in the basicity of the quinoline nitrogen upon electronic excitation. The excited system evolves on the ππ* manifold and undergoes fast proton transfer, followed by C8-C9 breakage, which results in the loss of neutral quinuclidinyl radical. Complexation with H2SO4 modifies the photoinduced process because of competitive interaction at the reaction site. Quantum chemical calculations would be highly desirable to confirm this hypothesis. Both CID and UVPD experiments show that the total fragmentation yield of the dimers Qtot is larger for (CdQdH+) than for (CdQnH+). These results show that the homochiral dimer is more strongly bound than the heterochiral adduct. CID exclusively produces the protonated monomers at m/z 295 (CdH+) and m/z 325 (QnH+ or QnH+), with the m/z 325 fragment being systematically more abundant. Moreover, the Q325/Q295 ratio is larger for the heterochiral complex (CdQdH+), which results from subtle differences in interaction between the two diastereomers. A full explanation of the CID- or UV-induced fragmentation patterns observed in the cinchona alkaloid dimers demands that their structure be elucidated. The vibrational spectra, as obtained by IRMPD spectroscopy in the NH and OH frequency region, as well as the structures calculated by quantum chemistry methods will be reported in the near future. Acknowledgment. Financial support by the European Commission to the EPITOPES project (Electron Plus Infrared TO Probe and Elucidate Structures, EC project 15637) founded through the NEST (New and Emerging Science and Technology) program is gratefully acknowledged. We thank Edith Nicol for her help with the operation of the FT-ICR mass spectrometer. Supporting Information Available: UVPD/CID MS3 mass spectrum of the m/z 483 radical cation. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Monoterpenoid Indole Alkaloids: Supplement to Part 4; Saxton, J. E., Eds.; John Wiley & Sons: New York, 1994; Vol. 25. (2) Sonderegger, O. J.; Ho, G. M. W.; Burgi, T.; Baiker, A. J. Catal. 2005, 230, 499. (3) Bu¨rgi, T.; Vargas, A.; Baiker, A. J. Chem. Soc., Perkin Trans. 2 2002, 1596. (4) Wirz, R.; Burgi, T.; Lindner, W.; Baiker, A. Anal. Chem. 2004, 76, 5319.

Scuderi et al. (5) Meier, D. M.; Urakawa, A.; Turra, N.; Ruegger, H.; Baiker, A. J. Phys. Chem. A 2008, 112, 6150. (6) Maier, N. M.; Schefzick, S.; Lombardo, G. M.; Feliz, M.; Rissanen, K.; Lindner, W.; Lipkowitz, K. B. J. Am. Chem. Soc. 2002, 124, 8611. (7) Gyimesi-Forras, K.; Maier, N. M.; Kokosi, J.; Gergely, A.; Lindner, W. Chirality 2009, 21, 199. (8) Akasaka, K.; Gyimesi-Forras, K.; Lammerhofer, M.; Fujita, T.; Watanabe, M.; Harada, N.; Lindner, W. Chirality 2005, 17, 544. (9) Bicker, W.; Chiorescu, I.; Arion, V. B.; Lammerhofer, M.; Lindner, W. Tetrahedron: Asymmetry 2008, 19, 97. (10) Schug, K. A.; Lindner, W. J. Sep. Sci. 2005, 28, 1932. (11) Tao, W. A.; Cooks, R. G. Anal. Chem. 2003, 75, 25A. (12) Liang, Y. J.; Bradshaw, J. S.; Dearden, D. V. J. Phys. Chem. A 2002, 106, 9665. (13) Tao, W. A.; Gozzo, F. C.; Cooks, R. G. Anal. Chem. 2001, 73, 1692. (14) Speranza, M.; Gasparrini, F.; Botta, B.; Villani, C.; Subissati, D.; Fraschetti, C.; Subrizi, F. Chirality 2009, 21, 69. (15) Correia, C. F.; Balaj, P. O.; Scuderi, D.; Maitre, P.; Ohanessian, G. J. Am. Chem. Soc. 2008, 130, 3359. (16) Wu, R. H.; McMahon, T. B. ChemPhysChem 2008, 9, 2826. (17) Marian, C.; Nolting, D.; Weinkauf, R. Phys. Chem. Chem. Phys. 2005, 7, 3306. (18) Stearns, J. A.; Guidi, M.; Boyarkin, O. V.; Rizzo, T. R. J. Chem. Phys. 2007, 127, 154322. (19) Joly, L.; Antoine, R.; Broyer, M.; Dugourd, P.; Lemoine, J. J. Mass Spectrom. 2007, 42, 818. (20) Gregoire, G.; Lucas, B.; Barat, M.; Fayeton, J. A.; DedonderLardeux, C.; Jouvet, C. Eur. Phys. J. D 2009, 51, 109. (21) Seurre, N.; Sepiol, J.; Le Barbu-Debus, K.; Lahmani, F.; ZehnackerRentien, A. Phys. Chem. Chem. Phys. 2004, 6, 2867. (22) Seurre, N.; Le Barbu-Debus, K.; Lahmani, F.; Zehnacker-Rentien, A.; Sepiol, J. J. Mol. Struct. 2004, 692, 127. (23) Seurre, N.; Le Barbu-Debus, K.; Lahmani, F.; Zehnacker, A.; Borho, N.; Suhm, M. A. Phys. Chem. Chem. Phys. 2006, 8, 1007. (24) Zehnacker, A.; Suhm, M. A. Angew. Chem., Int. Ed. 2008, 47, 6970. (25) Speranza, M.; Satta, M.; Piccirillo, S.; Rondino, F.; Paladini, A.; Giardini, A.; Filippi, A.; Catone, D. Mass Spectrom. ReV. 2005, 24, 588. (26) Speranza, M.; Rondino, F.; Satta, M.; Paladini, A.; Giardini, A.; Catone, D.; Piccirillo, S. Chirality 2009, 21, 119. (27) Co, M. In The Merck Index: An Encyclopedia of Chemicals, Drugs, and Biologicals, 11th ed.; Budavari, S., O’Neil, M. J., Smith, A., Heckelman, P. E., Eds.; Merck: Rahway, NJ, 1989; pp 1283. (28) NIST Chemistry WebBook. http://webbook.nist.gov/chemistry/ (accessed March 6, 2009). (29) Sawada, M.; Takai, Y.; Yamada, H.; Nishida, J.; Kaneda, T.; Arakawa, R.; Okamoto, M.; Hirose, K.; Tanaka, T.; Naemura, K. J. Chem. Soc., Perkin Trans. 2 1998, 701. (30) Bakker, J. M.; Besson, T.; Lemaire, J.; Scuderi, D.; Maitre, P. J. Phys. Chem. A 2007, 111, 13415. (31) Mac Aleese, L.; Simon, A.; McMahon, T. B.; Ortega, J. M.; Scuderi, D.; Lemaire, J.; Maitre, P. Int. J. Mass Spectrom. 2006, 249, 14. (32) Hiraya, A.; Achiba, Y.; Kimura, K.; Lim, E. C. J. Chem. Phys. 1984, 81, 3345. (33) Anton, M. F.; Nicol, M. J. Lumin. 1979, 18/19, 131. (34) Lahmani, F.; Douhal, A.; Breheret, E.; Zehnacker-Rentien, A. Chem. Phys. Lett. 1994, 220, 235. (35) Joly, L.; Antoine, R.; Allouche, A. R.; Dugourd, P. J. Am. Chem. Soc. 2008, 130, 13832. (36) Joly, L.; Antoine, R.; Allouche, A. R.; Broyer, M.; Lemoine, J.; Dugourd, P. J. Am. Chem. Soc. 2007, 129, 8428. (37) Valeur, B. Molecular Fluorescence; Wiley-VCH: Weinheim, Germany, 2002. (38) Fayed, T. A.; Etaiw, S. E. H.; Landgraf, S.; Grampp, G. Photochem. Photobiol. Sci. 2003, 2, 376. (39) Lucas, B.; Barat, M.; Fayeton, J. A.; Perot, M.; Jouvet, C.; Gregoire, G.; Nielsen, S. B. J. Chem. Phys. 2008, 128, 164302. (40) Qin, W. W.; Vozza, A.; Brouwer, A. M. J. Phys. Chem. C 2009, 113, 11790. (41) Lahmani, F.; Breheret, E.; Zehnacker-Rentien, A.; Ebata, T. J. Chem. Soc., Faraday Trans. 1993, 89, 623. (42) Lahmani, F.; Le Barbu-Debus, K.; Seurre, N.; Zehnacker-Rentien, A. Chem. Phys. Lett. 2003, 375, 636. (43) Caner, H.; Biedermann, P. U.; Agranat, I. Chirality 2003, 15, 637. (44) Weselucha-Birczynska, A. J. Mol. Struct. 2007, 826, 96. (45) Olsen, R. A.; Borchardt, D.; Mink, L.; Agarwal, A.; Mueller, L. J.; Zaera, F. J. Am. Chem. Soc. 2006, 128, 15594. (46) Ferri, D.; Burgi, T.; Baiker, A. J. Chem. Soc., Perkin Trans. 2 1999, 1305. (47) Reilly, J. P. Mass Spectrom. ReV. 2009, 28, 425.

JP9094497