ARTICLE pubs.acs.org/JPCA
UV Photodissociation Dynamics of Deprotonated 20 -Deoxyadenosine 50-Monophosphate [50-dAMPH] S. Sunil Kumar, M. Perot-Taillandier,† B. Lucas, S. Soorkia, M. Barat, and J. A. Fayeton* Institut des Science Moleculaires d’Orsay, CNRS UMR 8214, Universite Paris Sud 11, F-91405 Orsay Cedex, France ABSTRACT: The UV photodissociation dynamics of deprotonated 20 -deoxyadenosine 50 -monophosphate ([50 -dAMPH]) has been studied using a unique technique based on the coincident detection of the ion and the neutral fragments. The observed fragment ions are m/z 79 (PO3), 97 (H2PO4), 134 ([AH]), 177 ([dAMPHAH2O]), and 195 ([dAMPHA]), where “A” refers to a neutral adenine molecule. The relative abundances are comparable to that found in previous studies on [50 -dAMPH] employing different excitation processes, i.e., collisions and UV photons. The fragmentation times of the major channels have been measured, and are all found to be on the microsecond time scale. The fragmentation mechanisms for all channels have been characterized using velocity correlation plots of the ion and neutral fragment(s). The findings show that none of the dissociation channels of [50 -dAMPH] is UV specific and all proceed via statistical fragmentation on the ground state after internal conversion, a result similar to fragmentations induced by collisions.
1. INTRODUCTION Stability of biological molecules under UV irradiation has been one of the most widely pursued research problems over decades. However, it is very surprising that there exist only a few UV photodissociation studies14 of nucleotides, which constitute the building block of the molecule of life, DNA. In a typical mass spectrometric study of biological molecules such as proteins and nucleic acids, one usually prepares a gas-phase ensemble of the protonated or deprotonated forms of the molecule of interest (oligopeptides or oligonucleotides) and excites it under some perturbation such as collision with atoms, low energy electrons, or photons, and detects the fragment ions as a function of their m/z ratio. Such mass spectrometric studies enable one to extract information that helps in protein and nucleic acid sequencing.57 Similar studies, together with quantum chemical calculations, can be used to propose6,8,9 fragmentation mechanisms and determine the time scales of fragmentation. However, such indirect methods do not always deliver unambiguous information about the fragmentation processes. Deprotonated mononucleotides have been studied under various excitation methods such as collision induced dissociation (CID),812 infrared multiphoton dissociation (IRMPD),13 and UV photodissociation (UVPD).2,3,4 Some quantum chemical calculations are also available.1416 The experimental studies have shown that the dissociation products of these molecules are very similar. Commonly observed dissociation channels are the loss of an anionic or neutral base, the loss of neutral base + water, and the formation of PO3 or H2PO4. However, the dissociation thresholds and the branching ratios for various channels differ from one nucleotide to the other and also depend on whether the phosphate group is located at the 30 or 50 position8,9,11,17 of the sugar moiety. Previous [50 -dAMP-H] CID r 2011 American Chemical Society
studies by Ho and Kebarle8 suggest that the fragmentation channels except the loss of neutral base + water proceed through binary fragmentation. Using a 266 nm laser, Nielsen and co-workers4 investigated the photofragmentation of adenosine 50 -monophosphate anion ([50 -AMPH]) with the heavy ion storage ring, ELISA. They measured 16 μs for the overall decay time for this molecular ion, 4 and their results indicate that [50 -AMPH] predominantly dissociates statistically. However, the authors do not exclude the possibility of nonstatistical channels. The photofragmentation mass spectrum of [50 -AMPH] obtained by Aravind and co-workers 2 is similar to that of its DNA counterpart, 20 -deoxyadenosine 50 -monophosphate ([50 -dAMPH], shown in Figure 1). These authors also measured the overall decay time for [50 -AMPH] under photodissociation at 266 nm and found that the decay yield shows two dissociation lifetimes, 54% with a 95 ns lifetime and 46% with a 2.4 μs lifetime. They suggested that the fast component (95 ns) could be attributed to the breakup of the glycosidic bond (CN) and the slower component could be attributed to the remaining channels. Marcum and co-workers3 studied UV photodissociation of 20 -deoxyribose 50 -monophosphate ([50 -dBMPH], where “B” stands for any of the four nucleobases) and measured the branching ratios of all the observed fragmentation channels. They noticed that there exists a huge difference between the relative abundances of the anionic base compared with the CID results by Ho and Kebarle.8 To conclude, they cautioned that the formation of anionic adenine could be a nonstatistical process. Received: May 21, 2011 Revised: July 13, 2011 Published: July 21, 2011 10383
dx.doi.org/10.1021/jp204738q | J. Phys. Chem. A 2011, 115, 10383–10390
The Journal of Physical Chemistry A In [50 -dAMPH], the UV photons excite the adenine base.18 Relaxation mechanisms of adenine followed by UV excitation have been extensively studied both theoretically and experimentally.1922 The UV excitation leaves adenine in the optically active ππ* excited state from where it undergoes an ultrafast relaxation through conical intersection to the ground state.2123 This relaxation pathway is characterized by a nuclear geometry in which the six-membered ring in adenine is strongly puckered.21 Other pathways involving the coupling of ππ* and nπ*/πσ* states have also been shown to contribute to fast relaxation of adenine from its excited state.22 Similar explanations hold for other nucleotides, where the UV photons excite the corresponding nucleobase.22 In this paper, we present an experimental investigation of the [50 -dAMPH] photofragmentation induced by a 263 nm laser. The coincidence detection of both the ion and the neutral fragments allows for identification of each fragmentation channel and provides information on the fragmentation mechanisms, whether binary or sequential, as demonstrated previously.24,25 The experimental setup also enables unambiguous determination of the fragmentation times associated with each channel, which allows a conclusion on the statistical character of the fragmentation. In this paper the term “statistical” will be used to describe a process in which the fragmentation occurs after intramolecular vibrational relaxation to the ground state.
ARTICLE
hexapole for about 1 ms, and then injected between two pulsed electrodes to form ion bunches of 200 ns duration. The ion bunch is accelerated to 2.5 keV and mass selected using an electric chopper. The ions are then postaccelerated to a kinetic energy of 5 keV for better detection efficiency. The pulsed ion beam intercepts the 263 nm laser pulse (100 μJ/pulse, 200 ns duration, and 1 kHz pulse repetition rate) in the interaction region located just after the postacceleration (Figure 2). Notice that, at the present wavelength, the available energy is not sufficient for photodetachment.26 The interaction box is made of a set of plates that can be polarized to measure the fragmentation time in a range from a few tens up to hundreds of nanoseconds for each fragmentation channel as detailed elsewhere.27 Neutral fragments go straight to a first position sensitive detector (PSD) located along the beam axis. Ion fragments are mass and energy analyzed in a 45° electrostatic analyzer and received on a second PSD. The neutral detector is a fast PSD in which the pulses produced by a multichannel plate (MCP) are received on a double delay-line configuration that enables the determination of the position and the arrival time of several neutral fragments originating from the same fragmentation
2. EXPERIMENT The experimental setup has been described in detail elsewhere,24 and only a brief description is given here. In earlier experiments, the setup was used in the positive ion mode, whereas in the present study it has been modified to work in the negative ion mode. The sample is prepared by dissolving 50 -dAMP in a mixture of acetonitrile and water (50:50) at a concentration of 150 μM. Deprotonated [50 -dAMPH] ions are produced in an electrospray source. These ions are stored in a
Figure 1. Schematic representation of deprotonated 20 -deoxyadenosine 50 -monophosphate [50 -dAMPH].
Figure 3. Fragmentation time spectrum of the m/z 195 ion fragments. The interaction box is polarized with a 13 V/mm electric field. The righthand-side peak corresponds to delayed fragmentation occurring after the beam has left the interaction box. The left-hand-side structure corresponds to fast fragmentation produced in the interaction box. For this process the time scale on the x-axis has been calculated using the positiontime relation given in ref 27 (this relation is valid only in the interaction box, as indicated by the double slash on the x-axis). Dotted blue line: Monte Carlo simulation providing a fragmentation time of τ = 0.8 μs. Solid red line: the exponential fit of the left-hand-side structure gives τ = 0.8 ( 0.2 μs, which is consistent with the simulation.
Figure 2. Scheme of the experimental setup. Shown are the postacceleration assemblies, the interaction box in which the UV laser interacts with the ion beam, the electrostatic analyzer, and the two position sensitive detectors (PSDs). The y- and z-axes are also indicated (see text). 10384
dx.doi.org/10.1021/jp204738q |J. Phys. Chem. A 2011, 115, 10383–10390
The Journal of Physical Chemistry A
Figure 4. Intensity profiles along the y direction of the of the m/z 134 ion fragments on the PSD. The right-hand-side profile corresponds to fragmentations occurring after the beam has left the interaction box. The left-hand-side profile corresponds to fast fragmentations occurring inside the interaction box. Experimental data: open dots, Monte Carlo simulation for three different fragmentation times τ; dashed line, τ = 1.0 μs; solid line (red), τ = 1.3 μs; dotted line, τ = 1.6 μs.
process. The ion detector consists of an “MCP + resistive anode” configuration that provides the position and the arrival time of the ion fragments. An event is defined as a coincidence between an ion fragment and one or two neutral fragments generated in the dissociation of a given ion during one laser pulse. For each event, arrival times and positions of neutral and ion fragments are recorded, i.e., tn, yn, and zn for the neutral fragment(s) and ti, yi, and zi for the ion fragment. If more than one neutral fragment is produced in a given fragmentation event, coincidence between one ion and the neutrals can be recorded as long as the difference in arrival times between the two neutrals is larger than 40 ns, the dead time of the PSD. The flux of incident ions and photons is kept low enough to record at most one event per ion pulse to avoid false coincidence events. 2.1. Determination of Fragmentation Times. Information on time scales of fast fragmentation processes can be obtained by having the interaction of the laser with the target ion beam produced inside a constant electric field as discussed elsewhere.27 The length of the present configuration of the interaction box allows unambiguous determination of the fragmentation time in a range from a few tens of nanoseconds up to 1 μs. For example, Figure 3 shows the fragmentation time spectrum for m/z 195 fragments, corresponding to the loss of neutral adenine base. This result was obtained with the interaction box polarized with an electric field of 13 V/mm. The fragmentation time is determined by fitting (solid curve) the distribution of events that happen inside the interaction box, using an exponential decay function convoluted with a Gaussian-type function.27 The measured decay time of 0.8 μs is found to be consistent with a Monte Carlo simulation (dotted curve). Longer fragmentation times are determined by polarizing the interaction box at a constant voltage and measuring, with the electrostatic analyzer, the ratio of fragmentation events occurring inside the box and downstream after leaving the box. For a given voltage applied to the interaction box, two peaks appear in the energy analysis of the ion fragments. A Monte Carlo simulation code has been written to simulate fragmentation processes and determine fragmentation times from the relative intensities of the two peaks.
ARTICLE
Figure 5. Relative intensities of the various fragmentation channels observed in the present experiment compared to those observed in the photodissociation experiments of Marcum and co-workers,3 (A) and in the ion trap collisional activation experiments of Habibi-Goudarzi and McLuckey, (B)11 and (C),28 respectively.
Simulations shown in Figure 4 illustrate how sensitive is the ratio of the number of events happening inside and outside the interaction box with different fragmentation times ranging from 1.0 to 1.6 μs. In fact, the present results are obtained using the latter method, which presents a better time sensitivity in the scale of fragmentation of the present molecule. However, the two methods give similar results within the 300 ns uncertainty of the experiment. 2.2. Determination of the Fragmentation Mechanisms: vv Correlations. Information on the fragmentation mechanisms, such as the number of fragmentation steps, can be obtained by analyzing the correlation of the arrival times and positions, (t, y, and z) of the neutral and ion fragments detected in coincidence. Consider a two-body breakup M0 f mi + mn, with M0 = 330 and mi and mn are the masses of the ion and the neutral fragment, respectively. The momentum conservation provides the following relationship between the velocities of the ion and the neutral fragments:25,27 vzi mn ¼ vzn mi where vzi and vzn are the components of the velocities along the z-axis normal to the electric field inside the electrostatic analyzer (Figure 2). The z components of the velocity are chosen because they are not affected by the electric field and hence provide a simple relationship between the velocity components. In principle, for a two-body breakup, such vv correlation (in a vzi versus vzn plot) should be a straight line with a slope equal to mn/mi. Experimentally, the correlation is observed as an elongated ellipse due to the finite size of the beam, the velocity, and the angular spreads of the fragments. However, mn/mi slope is conserved. A contour plot with a slope different from that of a binary breakup, or showing a random correlation, is an indication of a concerted or sequential breakup involving two or more neutral fragments.25
3. RESULTS 3.1. Intensities of the Various Fragmentation Channels. The fragment ions observed in our UV photodissociation 10385
dx.doi.org/10.1021/jp204738q |J. Phys. Chem. A 2011, 115, 10383–10390
The Journal of Physical Chemistry A
ARTICLE
Table 1. Measured [dAMPH] Fragmentation Times for the Various Channels within an Uncertainty of 0.3 μs m/z fragment ions
fragmentation time (μs)
Table 2. Fit of the vv Distribution Maps for Different Fragmentation Channels Studied in the Present Work m/z fragment ion
no. of steps
expected slope (binary breakup)
fitted slope
1
79
3.18
3.11 ( 0.07
97 134
1.4 (first step) 1.3
>1 1
97 134
2.40 1.46
uncorrelated 1.47 ( 0.05
177
1.0, 0.1 ( 0.1a
2
177
not a binary breakup
195
0.8
1
195
0.69
79
1.5
0.70 ( 0.03
a
This error value is given as an upper limit. Within experimental uncertainties, the fragmentation time cannot be determined precisely.
Scheme 1. Formation of the Fragment Ion PO3 As Proposed by Ho and Kebarle8 a
a
Figure 7. vv correlation diagrams for (a) m/z 195 and (b) m/z 134 fragments. The red dashed lines drawn over the distributions have slopes corresponding to the neutral/ion fragment mass ratios for a two-body breakup.
The present study confirms the proposed mechanism.
Scheme 2. Formation of the Fragment Ions [dAMPHA] and [AH] As Proposed by Ho and Kebarle8 a
Figure 6. vv correlation diagrams for (a) m/z 79 and (b) m/z 97 fragments. The red dashed lines drawn over the distributions have slopes corresponding to the neutral/ion fragment mass ratios for a two-body breakup. Notice that the spreading of the (b) distribution signs a nonbinary fragmentation (see text). 0
(PO3),
experiment on [5 -dAMPH] are m/z 79 97 (H2PO4), 134 ([AH]), 177 ([dAMPHAH2O]), and 195 ([dAMPHA]), where “A” refers to a neutral adenine molecule. The same fragment ions were observed in CID8,11 and previous UVPD3 experiments. Figure 5 compares the relative intensities of the various channels observed in our experiment with those found in the UVPD experiments by Marcum and co-workers,3 obtained at a similar wavelength, and with those obtained in the ion trap collision activated dissociation (CAD) experiments by Habibi-Gaoudarzi and McLuckey (see ref 11 and Figure 7 in ref 28). The experimental data of Habibi-Gaoudarzi and McLuckey shown in Figure 5 are those obtained at a resonance excitation amplitude of 100 mV.11 It can be noted that the relative intensities of the various fragmentation channels are similar in all three experiments despite the entirely different excitation processes in CAD and UVPD, suggesting similar
a
See text.
fragmentation mechanisms for both photodissociation and collision experiments, i.e., internal conversion. 3.2. Fragmentation Times. The direct dissociation of an electronic excited state is a fast nonstatistical process occurring on a fragmentation time scale below 100 ns. Our time- and position-resolved coincidence method gives a direct measurement of the fragmentation time for each channel separately and allows unambiguous determination of whether there exist such nonstatistical fragmentations. Table 1 summarizes the fragmentation times associated with the various channels observed in our experiment. It is noteworthy that all fragmentation times are on the order of 1 μs, including the formation of the anionic adenine. 3.3. Determination of the Fragmentation Mechanisms. The characterization of the fragmentation mechanisms is achieved by using the vv correlation technique as explained in section 2.2. In the remainder of this section, the analysis for each fragment ion is presented. 10386
dx.doi.org/10.1021/jp204738q |J. Phys. Chem. A 2011, 115, 10383–10390
The Journal of Physical Chemistry A
ARTICLE
Figure 8. (ac) [(ti tn), Yi] correlations (see text) for m/z 97 fragment. The experimental correlation (a) is well accounted for by the m/z 330 f m/z 97 + [135 + 98] fragmentation scheme as simulated in (c). In contrast, the two-step scheme m/z 330 f m/z 195 + 135 followed by m/z 195 f m/z 97 + 98, as simulated in (b), does not account for the experiment.
m/z 79. CID studies have identified the m/z 79 fragment as PO3. It has been suggested that the fragment ion is produced via a two-body breakup as illustrated in Scheme 1.8 The vv correlation between the velocities of the ion fragment (PO3) and the corresponding neutral (mass = 330 79) fragment for a two-body breakup is shown in Figure 6a. The shape of the distribution and its fitted slope (Table 2) confirm the two-body breakup. m/z 134 and 195. The m/z 134 and 195 fragments correspond to the adenine anion and the (sugar + phosphate) anion, respectively. The vv correlations (Figure 7 and Table 2) confirm the binary fragmentation for these two channels as previously proposed in CID studies.8 Several pathways for anionic adenine and (sugar + phosphate) anion production from [50 -dAMPH]have been proposed. A generally accepted mechanism proceeds via a complex process, which is illustrated in Scheme 2.8,9 In this mechanism, the O in the phosphate group of the excited [50 -dAMPH] molecular ion attacks the H atom at the 20 position of the ribose sugar. This is followed by the breakup of the glycosidic bond, i.e., the CN bond between the sugar and the base, and a rearrangement of the molecule to form a proton bound complex between the adenine anion and the (sugar + phosphate). The latter then dissociates into either m/z 134 or m/z 195 depending on to which anion the proton is abstracted. The fragmentation times measured for these two channels (Table 1), determined by the present experiment, are similar, which is consistent with the proposed mechanism. m/z 97. In previous works on [50 -dAMPH], it was proposed that the m/z 97 corresponding to H2PO4 is formed by a two-body fragmentation as illustrated in eq 7 of ref 8. The vv correlation for the formation of this fragment ion is shown in Figure 6b. This diffuse vv correlation can hardly account for a simple binary breakup m/z 330 f m/z 97 + 233. Two mechanisms can then be proposed: either the ion fragment formed in the first step loses an additional neutral fragment m=z 330 f m=z 195 þ 135;
m=z 195 f m=z 97 þ 98
ð1Þ or the neutral fragment formed in the first step is unstable and dissociates in a second step m=z 330 f m=z 97 þ 233;
233 f 135 þ 98
ð2Þ
These two mechanisms (Figure 8b,c) have been simulated with the following parameters. For mechanisms 1 and 2, a 100 meV kinetic energy release has been chosen for each step and a 1.4 μs has been chosen for the first time constant. The second step of mechanism 1 should be fast and on the order of 300 ns to be consistent with the experiment. For the second step of mechanism 2, the experimental procedure does not allow accurate determination of the fragmentation time. One can only say that it should be smaller than 20 μs, the flight time of the neutrals. This simulation clearly excludes the first hypothesis and supports mechanism 2. In fact, the first step of mechanism 1 occurring in the polarized box would produce an intermediate m/z 195 ion fragment corresponding to the central left-hand-side structure in the simulation pattern of Figure 8b that is not seen in the experimental pattern of Figure 8a. It can be argued that the neutral fragment formed in the first step is left with significant internal energy to dissociate in the second step.29 In addition, the vv correlation plot corresponding to Figure 8b, although diffuse, is clearly stretched along the line corresponding to a binary fragmentation (not shown). This is again in favor of a multistep fragmentation as depicted in mechanism 2. The first step of this mechanism corresponds to a binary fragmentation giving the m/z 97 fragment, with the width of the correlation pattern being due to the second step. However, notice that the masses of the two neutral fragments had to be guessed since they cannot be determined by the experiment. A reasonable guess would be the breakup of the glycosidic bond because it is known to be a very weak bond.8 This scenario is also consistent with the hypothesis proposed in the CID works of Ho and Kebarle.8 m/z 177. Figure 9a shows the vv contour plot corresponding to the correlation between the ion and one neutral fragment. This complex pattern demonstrates a nonbinary breakup with one light neutral fragment being more deviated. The procedure for generating a two-step correlation diagram consists in selecting only triple coincidence events. The [(ti tn), Yi] correlations (not shown here) indicate that the second step is fast (∼100 ns, see Table 1). Here, it is not possible to distinguish between the first and second neutrals in time and hence which fragment is emitted in the first step. However, we can still differentiate the two neutrals by their respective deviations. The lighter fragment will always be the most deviated one. This has been explained in detail in a previous publication.25 In this case, we need to do a hypothesis on the mechanism, which can be validated with the vv 10387
dx.doi.org/10.1021/jp204738q |J. Phys. Chem. A 2011, 115, 10383–10390
The Journal of Physical Chemistry A
ARTICLE
Figure 9. vv contour plots for m/z 177: (a) correlations between the ion and one neutral fragment, either n1 (less deviated) or n2 (more deviated). The slope (M0 mi)/mi of the red dashed line is given by the ratio between M0 mi, the mass of the neutral fragment, and mi, that of the ion fragment for an assumed binary fragmentation. The events for vzi < 0.5 mm/μs are due to an experimental artifact. (b,c) and (d,e) correspond to the analysis of triple coincidence assuming mechanisms 3 and 4, respectively, where only events with |vn2| > 0.3 mm/μs are selected. (b) Correlations between the velocities of the center of mass of the ion and the more deviated neutral, (i + n2), and the less deviated neutral (n1) for the first step assuming mechanism 3. The corresponding second step is shown in (c). (d) Correlations between the velocities of the center of mass of the ion and the less deviated neutral, (i + n1), and the more deviated neutral (n2). The corresponding second step is shown in (e). The expected slopes for the two mechanisms are also shown.
correlations. The two possible pathways considered are m=z 330 f m=z 195 þ 135;
m=z 195 f m=z 177 þ 18
ð3Þ m=z 330 f m=z 312 þ 18;
m=z 312 f m=z 177 þ 135
ð4Þ The contour plots corresponding to the correlation between the ion fragment and the more and less deviated neutral
fragments are shown in Figure 9b,c and Figure 9d,e assuming mechanisms 3 and 4, respectively. The expected slopes for the two pathways are also drawn. Comparison between the slopes and the experimental patterns favors mechanism 4, which indicates that the lighter neutral is emitted first. Mechanism 3 is clearly excluded in this experiment within the 5 μs observation time. This finding appears to be at variance with mechanism 3 proposed by Habibi-Goudarzi and McLuckey11 based on ion trap collision activated studies (CAD). If mechanism 3 were feasible, the second step would appear at later times. Notice that the 10388
dx.doi.org/10.1021/jp204738q |J. Phys. Chem. A 2011, 115, 10383–10390
The Journal of Physical Chemistry A intermediate m/z 312 ion fragment is observed neither in the present experiment nor in the previous UVPD and CID studies, suggesting that it is a very unstable fragment.
4. DISCUSSION The relative abundances of the various fragment ions measured in the present and previous3 UVPD experiments are similar (Figure 5) to those of ion trap CAD experiments of HabibiGoudarzi and McLuckey.11,28 However, the CID results of Ho and Kebarle8 made in single collision conditions are in marked difference. Under ion trap collisional excitation, the molecular ensemble is heated slowly so that the ion fragments with the lowest dissociation threshold will be dominant in the mass spectrum. On the other hand, under single collision conditions, the center of mass energy transferred to the molecule can lead to the formation of any of the ions having a dissociation threshold less than the transferred energy. Comparing the threshold curves obtained for [50 -dAMPH] by Ho and Kebarle (see Figure 9 of ref 8), it can be noted that the abundance of m/z 134 ([BH]) is larger than that of any other ion fragment. UVPD of [dBMPH] produces much less [BH] compared to [dBMPHB] ion fragments (Figure 5).3 Marcum and coworkers,3 on the basis of the difference between the CID results of Ho and Kebarle8 and the results of their UVPD experiments, proposed that the formation of anionic adenine could be a nonstatistical process. In the present study, the determination of the fragmentation time clearly demonstrates that all dissociation channels of [50 -dAMPH] are statistical. The fragmentation times measured in the present study for all the channels of [dAMPH] are on the order of microseconds. This result is at variance with that of Aravind and co-workers2 on the fragmentation of the AMP anion, its nucleic acid counterpart. From a Monte Carlo simulation of the neutral yield, they concluded that fragmentation times could be deduced with two exponential decays, i.e., a fast component (τ = 95 ns) and a slow component (τ = 2.4 μs), with comparable probabilities. They suggested that the fast process could be attributed to the glycosidic bond breakup. No such fast component was observed in the present experiment. While the fragmentation channels are similar, one might argue that the observed difference in the decay times might be attributed to the presence of the OH group at the 20 -position in AMP. However, if that is true, a reduction in the fragmentation time of the CN bond by a factor of 10 is quite surprising. The relative abundances of the various fragment ions in both collision11,28 (except for the results of Ho and Kebarle8) and photodissociation3 experiments on nucleotides are consistent with the conclusion based on slow fragmentation (approximately microseconds). The excited state lifetimes of nucleic bases, nucleosides, and nucleotides have been shown to be on the order of picoseconds.18,20,30,31 Thus it is evident that all fragmentation channels of [50 -dAMPH] including that of the anionic adenine result from an internal conversion of the molecule as characterized by the long fragmentation times. 5. CONCLUSION UV photodissociation dynamics of [50 -dAMP-H] has been studied using a coincidence detection of the ion and the neutral fragments. The relative intensities as well as the fragmentation times for each channel have been measured. The fragmentation mechanisms of all the channels have been identified. It was found
ARTICLE
that all of the [50 -dAMPH] fragmentation channels are slow, with the same order of magnitude, meaning that the system relaxes through internal conversion to the ground state before statistical fragmentation takes place, a result consistent with the high stability of nucleotides against UV photofragmentation. This result is at variance with that of Aravind and co-workers,2 who measured a fast component in the decay of photodissociated [AMPH], the nucleic acid counterpart, with similar ion fragments as that of [dAMPH] but with different relative abundances.9 On the basis of previous studies on [50 -dAMPH] and similar molecules, most of the results presented in this paper are likely to be applicable to a class of molecules [n0 -dBMPH] and [n0 -BMPH], where n can be 3 or 5 and “B” represents any one of the four bases in DNA/RNA.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected]. Present Addresses †
Institut Parisien de Chimie Moleculaire, CNRS UMR 7201, Universite Pierre et Marie Curie, Paris, F-75252 Paris Cedex 05, France.
’ ACKNOWLEDGMENT We gratefully acknowledge useful discussions with Christophe Jouvet during this work. The present work was carried out at the Centre de Cinetique Rapide ELYSE. S.K. is thankful for financial support provided by Universite Paris Sud 11 and CNRS for carrying out this work. ’ REFERENCES (1) Andersen, J. U.; Andersen, L. H.; Hvelplund, P.; Lapierre, A.; Møller, S. P.; Nielsen, S. B.; Pedersen, U. V.; Tomita, S. Hyperfine Interact. 2003, 146147, 283. (2) Aravind, G.; Antoine, R.; Klaerke, B.; Lemoine, J.; Racaud, A.; Rahbek, D. B.; Rajput, J.; Dugourd, P.; Andersen, L. H. Phys. Chem. Chem. Phys. 2010, 12, 3486. (3) Marcum, J. C.; Halevi, A.; Weber, J. M. Phys. Chem. Chem. Phys. 2009, 11, 1740. (4) Nielsen, S. B.; Andersen, J. U.; Forster, J. S.; Hvelplund, P.; Liu, B.; Pedersen, U. V.; Tomita, S. Phys. Rev. Lett. 2003, 91, 048302. (5) Biemann, K.; Martin, S. A. Mass Spectrom. Rev. 1987, 6, 1. (6) Wu, J.; McLuckey, S. A. Int. J. Mass Spectrom. 2004, 237, 197. (7) Limbach, P. A. Mass Spectrom. Rev. 1996, 15, 297. (8) Ho, Y.; Kebarle, P. Int. J. Mass Spectrom. Ion Processes 1997, 165166, 433. (9) Rodgers, M. T.; Campbell, S.; Marzluff, E. M.; Beauchamp, J. L. Int. J. Mass Spectrom. Ion Processes 1994, 137, 121. (10) Chipuk, J. E.; Brodbelt, J. S. J. Am. Soc. Mass Spectrom. 2007, 18, 724. (11) Habibi-Goudarzi, S.; McLuckey, S. A. J. Am. Soc. Mass Spectrom. 1995, 6, 102. (12) Liu, B.; Nielsen, S. B.; Hvelplund, P.; Zettergren, H.; Cederquist, H.; Manil, B.; Huber, B. A. Phys. Rev. Lett. 2006, 97, 133401. (13) Chiavarino, B.; Crestoni, M. E.; Fornarini, S.; Lanucara, F.; Lemaire, J.; Maitre, P.; Scuderi, D. Int. J. Mass Spectrom. 2008, 270, 111. (14) Close, D. M. J. Phys. Chem. A 2008, 112, 8411. (15) Hou, R.; Gu, J.; Xie, Y.; Yi, X.; Schaefer, H. F. J. Phys. Chem. B 2005, 109, 22053. (16) Alexandrova, A. N.; Tully, J. C.; Granucci, G. J. Phys. Chem. B 2010, 114, 12116. 10389
dx.doi.org/10.1021/jp204738q |J. Phys. Chem. A 2011, 115, 10383–10390
The Journal of Physical Chemistry A
ARTICLE
(17) Phillips, D. R.; McCloskey, J. A. Int. J. Mass Spectrom. Ion Processes 1993, 128, 61. (18) Crespo-Hernandez, C. E.; Cohen, B.; Hare, P. M.; Kohler, B. Chem. Rev. 2004, 104, 1977. (19) Cheong, N. R.; Nam, S. H.; Park, H. S.; Ryu, S.; Song, J. K.; Park, S. M.; Perot, M.; Lucas, B.; Barat, M.; Fayeton, J. A.; Jouvet, C. Phys. Chem. Chem. Phys. 2011, 13, 291. (20) Barbatti, M.; Aquino, A. J. A.; Szymczak, J. J.; Nachtigallova, D.; Hobza, P.; Lischka, H. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 21453. (21) Marian, C. M. J. Chem. Phys. 2005, 122, 104314. (22) Satzger, H.; Townsend, D.; Zgierski, M. Z.; Patchkovskii, S.; Ullrich, S.; Stolow, A. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 10196. (23) Brøndsted Nielsen, S.; Sølling, T. I. ChemPhysChem 2005, 6, 1276. (24) Lepere, V.; Lucas, B.; Barat, M.; Fayeton, J. A.; Picard, V. J.; Jouvet, C.; Carcabal, P.; Nielsen, I.; Dedonder-Lardeux, C.; Gregoire, G.; Fujii, A. J. Chem. Phys. 2007, 127, 134313. (25) Perot, M.; Lucas, B.; Barat, M.; Fayeton, J. A.; Jouvet, C. J. Phys. Chem. A 2010, 114, 3147. (26) Yang, X.; Wang, X. B.; Vorpagel, E. R.; Wang, L. S. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 17588. (27) Lucas, B.; Barat, M.; Fayeton, J. A.; Jouvet, C.; C-arc-abal, P.; Gregoire, G. Chem. Phys. 2008, 347, 324. (28) McLuckey, S. A.; Habibi-Goudarzi, S. J. Am. Chem. Soc. 1993, 115, 12085. (29) Kadhane, U.; Perot, M.; Lucas, B.; Barat, M.; Fayeton, J. A.; Jouvet, C.; Ehlerding, A.; Kirketerp, M. B. S.; Nielsen, S. B.; Wyer, J. A.; Zettergren, H. Chem. Phys. Lett. 2009, 480, 57. (30) Pecourt, J. M.; Peon, J.; Kohler, B. J. Am. Chem. Soc. 2001, 123, 10370. (31) Peon, J.; Zewail, A. H. Chem. Phys. Lett. 2001, 348, 255.
10390
dx.doi.org/10.1021/jp204738q |J. Phys. Chem. A 2011, 115, 10383–10390
