Rovibrationally selected ion-molecule collision study

Rovibrationally selected ion-molecule collision study using the molecular beam vacuum ultraviolet laser pulsed field ionization-photoion method: Charge transfer reaction of N2 +(X 2Σg +; v+ = 0–2; N + = 0–9) + Ar Yih Chung Chang, Yuntao Xu, Zhou Lu, Hong Xu, and C. Y. Ng Citation: The Journal of Chemical Physics 137, 104202 (2012); doi: 10.1063/1.4750248 View online: http://dx.doi.org/10.1063/1.4750248 View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/137/10?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Communication: Rovibrationally selected study of the N2 +(X; v+ = 1, N + = 0−8) + Ar charge transfer reaction using the vacuum ultraviolet laser pulsed field ionization-photoion method J. Chem. Phys. 134, 201105 (2011); 10.1063/1.3596748 The study of state-selected ion-molecule reactions using the vacuum ultraviolet pulsed field ionization-photoion technique J. Chem. Phys. 125, 132306 (2006); 10.1063/1.2207609 Rovibrational state-selected study of H 2 + (X,ν + =0–17, N + =1)+ Ar using the pulsed field ionizationphotoelectron-secondary ion coincidence scheme J. Chem. Phys. 118, 2455 (2003); 10.1063/1.1542884 A selected-ion-flow-drift-tube study of charge transfer processes between atomic, molecular, and dimer ion projectiles and polyatomic molecules ethane, propane, and n-butane J. Chem. Phys. 109, 4246 (1998); 10.1063/1.477073 A state-selected study of the ion–molecule reactions O + ( 4 S, 2 D, 2 P)+N 2 J. Chem. Phys. 106, 1373 (1997); 10.1063/1.474087

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.120.194.195 On: Tue, 10 Feb 2015 20:35:54

THE JOURNAL OF CHEMICAL PHYSICS 137, 104202 (2012)

Rovibrationally selected ion-molecule collision study using the molecular beam vacuum ultraviolet laser pulsed field ionization-photoion method: Charge transfer reaction of N2 + (X 2 g + ; v+ = 0–2; N+ = 0–9) + Ar Yih Chung Chang, Yuntao Xu, Zhou Lu, Hong Xu, and C. Y. Nga) Department of Chemistry, University of California, Davis, Davis, California 95616, USA

(Received 15 June 2012; accepted 20 August 2012; published online 12 September 2012) We have developed an ion-molecule reaction apparatus for state-selected absolute total cross section measurements by implementing a high-resolution molecular beam vacuum ultraviolet (VUV) laser pulsed field ionization-photoion (PFI-PI) ion source to a double-quadrupole double-octopole ionguide mass spectrometer. Using the total cross section measurement of the state-selected N2 + (v+ , N+ ) + Ar charge transfer (CT) reaction as an example, we describe in detail the design of the VUV laser PFI-PI ion source used, which has made possible the preparation of reactant N2 + (X 2 g + , v+ = 0–2, N+ = 0–9) PFI-PIs with high quantum state purity, high intensity, and high kinetic energy resolution. The PFI-PIs and prompt ions produced in the ion source are shown to have different kinetic energies, allowing the clean rejection of prompt ions from the PFI-PI beam by applying a retarding potential barrier upstream of the PFI-PI source. By optimizing the width and amplitude of the pulsed electric fields employed to the VUV-PFI-PI source, we show that the reactant N2 + PFI-PI beam can be formed with a laboratory kinetic energy resolution of Elab = ± 50 meV. As a result, the total cross section measurement can be conducted at center-of-mass kinetic energies (Ecm ’s) down to thermal energies. Absolute total rovibrationally selected cross sections σ (v+ = 0–2, N+ = 0–9) for the N2 + (X 2 g + ; v+ = 0–2, N+ = 0–9) + Ar CT reaction have been measured in the Ecm range of 0.04–10.0 eV, revealing strong vibrational enhancements and Ecm -dependencies of σ (v+ = 0–2, N+ = 0–9). The thermochemical threshold at Ecm = 0.179 eV for the formation of Ar+ from N2 + (X; v+ = 0, N+ ) + Ar was observed by the measured σ (v+ = 0), confirming the narrow Ecm spread achieved in the present study. The σ (v+ = 0–2; N+ ) values obtained here are compared with previous experimental and theoretical results. The theoretical predictions calculated based on the Landau-Zener-Stückelberg formulism are found to be in fair agreement with the present measured σ (v+ = 1 or 2; N+ ). Taking into account of the experimental uncertainties, the measured σ (v+ = 1 or 2, N+ ) for N+ = 0–9 at Ecm = 0.04–10.0 eV are found to be independent of N+ . © 2012 American Institute of Physics. [http://dx.doi.org/10.1063/1.4750248] I. INTRODUCTION

The main goal of state-selected studies of ion-molecule collisions is to gain fundamental insight into chemical reactivity of atomic and molecular ions as they are excited in different forms of energies, be it translational, rotational, vibrational, and electronic.1–9 The major experimental difficulty of these studies was concerned with the preparation of reactant ions in single internal quantum state with high purity and with sufficiently high intensities for single-collision measurements.3–9 We note that the majority of previous state-selected ion-molecule collision experiments was conducted at center-of-mass kinetic energies (Ecm ’s) higher than 1.0 eV. Considering that many state-selected ion-molecule reactions of relevance to planetary atmospheres occur at low temperatures, it is important to extend the Ecm for cross section measurements of state-selected ion-molecule processes down to thermal energies.10 The need for the determination of state-selected cross sections at thermal energies requires a) Author to whom correspondence should be addressed. Electronic mail:

[email protected]. 0021-9606/2012/137(10)/104202/13/$30.00

the preparation of state-selected reactant ions with high kinetic energy resolutions. To our knowledge, a general experimental scheme for the preparation of a state-selected reactant ion beam with the rotational selectivity and narrow kinetic energy spread close to thermal energies has not been achieved in previous state-selected ion-molecule reaction studies.2–9 This report provides an account on the successful development of such an experimental scheme for absolute total cross section measurements of rovibronically selected ion-molecule processes. Due to the well-known selection rules for photoionization of diatomic and simple polyatomic molecules and the fine control of photon energies,11–16 vacuum ultraviolet (VUV) photoionization has become a favorable method for the preparation of state-selected reactant ions.1, 4–9 The pulsed field ionization (PFI)-photoelectron (PFI-PE) and PFIphotoion (PFI-PI) techniques, being the state-of-the-art methods for high-resolution photoionization studies,15, 16 are capable of achieving the energy resolutions of 1–4 cm−1 , which are sufficient to resolve rotational photoionization transitions for many diatomic and simple polyatomic species.13–16 The PFI-PI and PFI-PE measurements, which are also referred in

137, 104202-1

© 2012 American Institute of Physics


104202-2

Chang et al.

the literature as mass-analyzed threshold ion (MATI) and PFIzero kinetic electron (ZEKE) measurements, respectively, give the same spectroscopic information except that the PFIPI measurement also carries the mass identity of the resulting ion formed in the photoionization process.16 Thus, the VUV-PFI-PI technique can be employed as a high-resolution method for the preparation of rovibronically selected reactant ions for ion-molecule collision studies.8, 9 Taking advantage of the dark-gap in the high-resolution VUV synchrotron source at the Advanced Light Source, Ng and co-workers have previously introduced the synchrotron based dark-gap VUV-PFI-PE scheme11, 12, 17–23 for ion spectroscopy measurements. Due to the pseudocontinuum nature of the synchrotron radiation, they have also introduced the VUV-PFI-PE-photoion coincidence (VUV-PFI-PEPICO) scheme for unimolecular dissociation studies,12–14, 24–28 and the VUV-PFI-PE-secondary ion coincidence (VUV-PFIPESICO) method9 for ion-molecule reaction studies. Employing this VUV-PFI-PESICO scheme, Ng and co-workers have obtained absolute total cross sections for the ion-molecule reactions H2 + (v+ ) [HD+ (v+ )] + Rg → RgH+ [RgD+ ] + H with vibrational states selected in the range of v+ = 0–17, where Rg = He, Ne, and Ar.9, 29–32 Although successful laser PFI-PI detection schemes have been reported in the literatures,33, 34 the general application of the laser PFI-PI schemes for the preparation of stateselected reactant ions for absolute cross section measurements of ion-neutral collisions has not been totally successful.8 Ng and co-workers have recently coupled a VUV laser PFI-PI ion source to the triple quadrupole-double octopole (TQDO) ion-guided mass spectrometer8 for absolute total cross section measurements of state-selected ion-molecule reactions. Using this apparatus, they have been able to prepare a beam of rovibrationally selected reactant NO+ (v+ = 0–2, N+ ) PFI-PIs with sufficiently high intensities for absolute total cross section measurements of the charge transfer (CT) reactions NO+ (v+ = 0–2, N+ ) + CH3 I (C6 H6 ) covering a Ecm range of 0.5–5.0 eV.8 However, the laboratory kinetic energy (Elab ) resolution (Elab ) for the reactant NO+ (v+ = 0–2, N+ ) PFI-PI beam achieved in the latter experiment was poor [Elab ≈ 1 eV (FWHM)], preventing ion-molecule collision experiments to be performed at low kinetic energies. In addition to reporting on the development of the molecular beam (MB) VUV laser-based PFI-PI double-quadrupole double-octopole (DQDO) apparatus for absolute total cross section measurements of rovibronically selected ion-molecule collisions, we also present absolute total cross sections of the N2 + (X2 g + ; v+ = 0–2; N+ = 0–9) + Ar → Ar+ (2 P1/2,3/2 ) + N2 reaction, σ (v+ = 0–2, N+ = 0–9), measured using this newly established apparatus. Preliminary results of the present experiment have already been communicated.35 The N2 + (X2 g + ; v+ ) + Ar CT reaction and its reverse are among the most studied ion-molecule processes.7, 36–40 At Ecm > 1 eV, the [Ar + N2]+ reaction system has been examined to the state-to-state detail.38–40 However, total cross section measurements for the Ecm range from thermal energies to 1 eV are lacking. The σ (v+ = 0–2, N+ ) values at Ecm = 0.04–10.0 eV obtained here are compared to previous experimental and theoretical results.36, 41–47

J. Chem. Phys. 137, 104202 (2012)

II. EXPERIMENTAL CONSIDERATION

The VUV laser PFI-PI DQDO apparatus used in this work consists of three major components: a tunable VUV laser system, a VUV laser PPI-PI source, and a DQDO ion-guide mass spectrometer. The DQDO ion-guide mass spectrometer was modified from the TQDO ion-guide mass spectrometer,7 which had been used extensively for absolute total cross section measurements of ion-molecule reactions involving state-selected ions prepared by the VUV photoionization method. The arrangement and procedures for generating tunable VUV coherent radiation by nonlinear fourwave mixing schemes have also been described in detail previously.12–14, 48 Thus, in this experimental session, only a brief description of the DODQ mass spectrometer and the VUV laser system is given below, with special attention given to the design and operation of the VUV laser PFI-PI source. A. VUV laser system

In this study, coherent VUV radiation was generated by means of resonance-enhanced four-wave sum-frequency mixing (2ω1 + ω2 ) scheme using Kr gas as the nonlinear medium.12–14, 48 Here, ω1 represents the fixed UV frequency at 47 046.43 cm−1 (212.556 nm) obtained by frequency doubling of the fundamental output of the first dye laser (Sirah, CSTR-D-30) with an optical bandwidth 0.05 cm−1 (FWHM). The energy of 2ω1 matches the 5p ← 4p resonance transition of Kr at 94 092.85 cm−1 . The ω2 frequency is tunable in the UV range of 317.4–276.9 nm, which is generated by frequency doubling of the fundamental output of the second dye laser (Sirah, CSTR-D-24) with optical bandwidth 0.08 cm−1 (FWHM). The first and second dye lasers were optically pumped by the respective 355 and 532 nm outputs of the same injection seeded Nd:YAG Laser (Spectra Physics, model 290) operated at a repetition rate of 15 Hz. An autotracker (Sirah, AUTO-FSC) equipped with the β-barium borate (BBO) crystal and beam path compensator was used for scanning the ω2 frequency and tracking the beam path. The ω1 and ω2 frequencies were merged by a dichroic mirror and aligned to intersect the Kr gas jet in the four-wave mixing chamber. The VUV laser radiation thus generated entered directly into the photoionization chamber without a window and is aligned to intersect perpendicularly the N2 supersonic molecular beam in the photoionization/photoexcitation (PI/PEX) center. The VUV sum frequencies (2ω1 + ω2 ) of interest thus generated covered the range of 125 500–130 200 cm−1 (15.56–16.14 eV) with an optical bandwidth of 0.36 cm−1 (FWHM). Since no grating was used to select the VUV frequencies of interest, the sum frequencies entered the PI/PEX region together with the fundamental frequencies (ω1 and ω2 ), the difference frequencies (2ω1 – ω2 ), and the tripled frequency (3ω1 ). B. The DQDO ion-guide mass spectrometer

Figure 1 depicts the schematic diagram of the DQDO ion-guide mass spectrometer along with VUV laser PFI-PI ion source generally designed for absolute total cross section measurements of state-selected ion-molecule reactions. This


104202-3

Chang et al.

J. Chem. Phys. 137, 104202 (2012)

FIG. 1. Schematic diagram for the VUV laser PFI-PI DQDO apparatus. (1) Beam source chamber [to 5000 l/s diffusion pump (DP)], (2) pulsed nozzle, (3) skimmer 1, (4) skimmer 2, (5) PE/PI center, (6) VUV laser beam path, (7) photoionization chamber [to 500 l/s turbomolecular pump (TMP)], (8) ion lenses (E1 + I1–I9) of the PFI-PI ion source, (9) reactant QMS, (10) dc quadrupole ion bender, (11) MCP ion detector, (12) reaction chamber (to 500 l/s TMP), (13) lower reaction gas cell, (14) short radio-frequency (rf)-octopole, (15) upper reaction gas cell, (16) long rf-octopole, (17) product QMS chamber (to 5000 l/s DP), (18) product QMS, (19) Daly-type ion detector, (20) detector chamber (to 500 l/s TMP).

apparatus consists of, in sequential order, a pulsed molecular beam source (1) for the generation of a pulsed supersonic N2 beam, a VUV laser PFI-PI reactant ion source for the preparation of state-selected reactant N2 + ions, a reactant quadrupole mass spectrometer (QMS) (9) for the selection of reactant ions, a dc quadrupole ion bender (10) for monitoring the ion transmission, a set of short (14) (length = 8.64 cm) and long (16) (length = 19.55 cm) rf octopole ion guides arranged in tandem for guiding the product ions formed in the reactant gas cell [lower (13) or upper gas cell (15)] to the ion detector, a product QMS (18) for the mass selection of product or reactant ions, and a Daly-type ion detector (19). The latter ion detector has a structure of the Daly ion detector except that the photomultiplier tube is replaced by a set of dual MCPs. Both the short and long octopoles are constructed of eight electro-polished stainless steel rods (diameter = 0.2 cm) and are symmetrically spaced on an inscribed circle with a diameter of 0.6 cm. The two octopoles are powered by a single rfpower supply, but can have different dc potentials. As pointed out above, there are two reaction gas cells. The lower gas cell (length = 9 cm) is located in the middle of the short rf octopole ion guide and the upper one (length = 5.25 cm) is centered at the junction of the short and long rf ion guides. The short reaction gas cell is designed to minimize secondary reactions between slow product ions (particularly those formed in CT reactions) and thermal neutral reactants in the gas cell. Since the dc potentials applied to the short and long octopoles can be different, slow thermal product ions formed in the upper reaction gas cell can be extracted to avoid trapping of the product ions in the gas cell. This arrangement also enables TOF analysis of axial and radial velocity distribution of product ions formed in the lower gas cell. Since this experiment was a study of CT reactions, the upper gas cell was used. The dc quadrupole ion bender (10) situated between the reactant QMS and the short rf octopole can be used to deflect the reactant ion beam to the MCP ion detector (11) situated on top of the ion bender. The comparison of the ion signals observed at the MCP detector of the ion bender and at the Daly-type ion detector (19) allows the assessment of ion transmission through the ion lenses, the ion guides, and the QMS units of the DQDO spectrometer. The DQDO ion-guide mass spectrometer is divided into five chambers, namely, the beam source chamber (1), the pho-

toionization chamber (7), the reaction chamber (12), the product QMS chamber (17), and the ion detector chamber (20). The beam source and the product QMS chambers are each evacuated by a water-cooled 10-in. diffusion pump (4000 l/s), and the other 3 chambers are evacuated by separated turbomolecular pumps (520 l/s). The pressure of the reactant Ar used in the upper gas cell was ≤ 2 × 10−4 Torr as monitored by an MKS Baratron. During the experiment, the pressure in the beam source chamber, the photoionization chamber, the reaction gas cell chamber, product QMS chamber, and the ion detector chamber were maintained at ≤1 × 10−5 , ≤1 × 10−6 , ≤2 × 10−6 , ≤1 × 10−6 , and ≤1×10−7 Torr, respectively.

C. VUV laser PFI-PI source

The VUV laser PFI-PI ion source consists of the supersonic molecular beam production system, and the set of 10 ion lenses [see item (8) in Fig. 1] (lenses E1–I9, spaced 1 cm apart) inside the photoionization chamber. All ion lenses are aperture lenses with a circular aperture (diameter = 0.9 cm); and the apertures of E1, I1, and I2 are covered by gold grids (transmission = 90%). These three ion lenses are the most essential elements of the PFI-PIs source. The rest of the ion lenses I3–I9 are used to focus and transport N2 + ions from the ion source into the reactant QMS. To facilitate the subsequent discussion, we show in Figs. 2(a) and 3(a) a magnified schematic view of ion lenses E1, I1, and I2, where the traveling directions of the VUV laser beam and the N2 molecular beam are shown by the red and blue arrows, respectively. The N2 MB was produced by a pulsed valve (repetition rate = 15 Hz) and traveled along the central axis (defined as the z-axis here) of the DQDO spectrometer. After passing through two conical skimmers and E1, the N2 MB intersected perpendicularly the VUV laser beam at the PI/PEX center, producing high-n Rydberg N2 * (n) species as well as prompt N2 + ions. The first and second skimmers have the respective diameters of 2 and 4 mm and are spaced by 5 cm. The use of the double skimmer arrangement limits the divergence of the N2 and N2 * (n) beam, which is essential for achieving a higher kinetic energy resolution for the reactant N2 + PFI-PI ion beam.


104202-4

Chang et al.

J. Chem. Phys. 137, 104202 (2012)

FIG. 2. (a) Schematic diagram of lenses E1, I1, and I2, where the travel directions of the VUV laser, N2 MB, N2 + prompt ion beam, and N2 + MATI (or PFI-PI) beam are shown by arrows. (b) The pulsing scheme applied to lenses E1, I1, and I2: Time zero signifies the firing of the VUV laser. The PFI/EX field pulse with the amplitude of 60 V/cm and width of 0.45 μs (turned on at t = 2.30–2.75 μs) was applied to E1. I1 and I2 were kept at the dc voltages of 0.0 and 5.8 V, respectively. (c) The TOF spectrum for N2 + observed using the ion MCP detector on top of the dc quadrupole ion bender shows a single ion peak, consisting of N2 + MATIs (or PFI-PIs) and prompt ions at 36.7 μs.

1. Preparation of a pure rovibrationally selected reactant N2 + (X 2 g+ , v+ , N+ ) PFI-PI beam

The success in the preparation of an ion beam of reactant N2 + (X2 g + , v+ , N+ ) PFI-PIs with high intensity, high internal energy resolution, and high kinetic energy resolution mainly arises from the optimal design of an electric field pulsing scheme to be applied to ion lenses E1, I1, and I2 [see Figs. 2(a) and 3(a)]. The sequence of steps to be followed for the generation of state-selected N2 + PFI-PIs and their separation from N2 + prompt ions are described below. In order to optimize the operation of the VUV laser PFI-PI source, the first step is to tune the VUV laser sumfrequency (2ω1 + ω2 ) output to the IE(N2 ), where the formation of N2 + PFI-PIs is to be expected. Figure 2(b) shows a typical pulsing scheme (i.e., the potentials applied to E1, I1, and I2 as a function of time t) for VUV-PI measurements. Here, time t = 0 signifies the firing of the VUV laser. In this example, the dc potentials of 0.0 and 5.8 V are applied to I1 and I2, respectively, while a voltage pulse (amplitude = 60 V, width = 0.45 μs) was applied to E1 at t = 2.3 μs. This voltage pulse, that serves to produce PFI-PIs from N2 *(n) species as well as to extract PFI-PIs and prompt ions out of the PI/PEX

region, is referred here as the PFI/EX pulse. Since no separation electric field is applied in the PI/PEX region, N2 *(n) species and prompt ions formed by VUV laser excitation are expected to travel at the same velocity of the N2 MB until the ion voltage pulse was applied to E1. The 5.8 V dc potential applied to I2 was set to give the best N2 + signal at the ion MCP detector. The TOF spectrum [Fig. 2(c)] measured using the MCP detector on top of the dc quadrupole bender reveals a single N2 + TOF peak (FWHM ≈ 0.5 μs) at 36.5 μs, indicating that prompt ions and PFI-PIs were not separated and exited I1 at the same time as shown in Fig. 2(a). In order to separate the prompt ions from the PFI-PIs, it is necessary to apply a separation electric field to the PI/PEX region to retard the prompt ions. Figure 3(b) depicts a modified pulsing scheme of Fig. 2(b) by adding a retarding voltage pulse (amplitude = 2 V, width = 5.5 μs) to I1 at t = 150 ns. This in effect introduces a separation field of 2 V/cm for 5.5 μs. We found that in order to minimize the destruction of N2 *(n) species, it was necessary to keep the PI/PEX region field free during VUV laser excitation of the N2 beam. Thus, the separation electric field was only switched on at t = 150 ns.

FIG. 3. (a) Schematic diagram of lenses E1, I1, and I2, where the travel directions of the VUV laser, N2 MB, N2 + prompt ion beam, and N2 + MATI (or PFI-PI) beam are shown by arrows. (b) The pulsing scheme applied to lenses E1, I1, and I2: Time zero signifies the firing of the VUV laser. The PFI/EX field pulse with the amplitude of 60 V/cm and width of 0.45 μs (turned on at t = 2.30–2.75 μs) was applied to E1. The separation retarding field pulse with the amplitude of 2.0 V/cm and width of 5.5 μs (turned on at t = 0.15–5.65 μs) was applied on I1. I1 and I2 were kept at the dc voltages of 0 and 5.8 V, respectively. (c) The TOF spectrum for N2 + observed using the ion MCP detector on top of the dc quadrupole ion bender reveals two well-separated peaks of N2 + MATIs (or PFI-PIs) and N2 + prompt ions at 36 and 40 μs, respectively.


104202-5

Chang et al.

As shown in the schematic diagram of Fig. 3(a), the effect of the separation field is to move the prompt N2 + ions in the (–)z direction (opposite direction of the neutral N2 MB). High-n Rydberg N2 * (n) species formed by VUV laser excitation at the PI/PEX center are expected to continue traveling in the (+)z direction with the same velocity of the N2 MB. Switching on the voltage pulse (amplitude = 60 V, width = 0.45 μs) at t = 2.3 μs serves to produce PFI-PIs from N2 * (n) species and give all the PFI-PIs and prompt ions the same momentum in the (+)z direction. Due to this momentum gain, the prompt ions were turned around to travel toward I1. However, since the resultant momentum (or velocity) of the prompt ion beam is less than that of the PFI-PI beam, the prompt ions lag behind the PFI-PIs as confirmed by the TOF analysis. According to the pulsing scheme of Fig. 3(b), the PFI-PIs produced by the PFI/EX electric field pulse along with the prompt ions were under the decelerating influence of the separation electric field pulse for 2.3 μs, resulting in slowing the PFI-PI and prompt ion beams from exiting I1. As shown in Fig. 3(c), the TOF spectrum observed by the bender MCP ion detector exhibits two well-resolved ion peaks at 36 and 40 μs, which can be assigned as the PFIPI and prompt ion peaks, respectively. The TOF spectrum (not shown here) was also measured using the Daly-type MCP ion detector, which revealed the PFI-PI TOF peak at 135 μs and the prompt ion TOF peak at 160 μs. The fact that the temporal separation between the PFI-PI and prompt ion TOF peaks increases as the flight distance is increased indicates that the kinetic energies for the prompt ions are lower than those for the PFI-PIs. For a spectroscopic study, it is sufficient to measure the PFI-PI spectrum by gating the PFI-PI signal identified in the TOF spectrum as a function of VUV laser energy. However, since this experiment is concerned with state-selected ionmolecule collision studies, it is necessary to prepare rovibrationally selected reactant N2 + (X2 g+ , v+ , N+ ) PFI-PIs with high purity by rejecting the prompt ions, such that only stateselected reactant PFI-PI N2 + can enter the reaction gas cell to react with the neutral reactants. In order to obtain a quantitative measurement of the kinetic energies for the PFI-PIs and prompt ion beams, we have performed a retarding potential analysis of the Elab distributions for these ion beams. Figures 4(a) and 4(b) show the respective retarding potential curves for the PFI-PI and prompt ion beams, obtained by ramping up the dc potential applied to the short octopole and detecting the N2 + ions using the Daly-type ion MCP detector. The analysis of the falling steps observed in the retarding potential curves of Figs. 4(a) and 4(b) allows the determination of Elab = 8.25 ± 0.20 and 5.65 ± 0.5 eV for the PFI-PI and prompt ion beams, respectively. The fact that the PFI-PIs and prompt ion beams have well-resolved Elab distributions implies that we can reject the prompt ions from entering reaction gas cell by applying an appropriate potential barrier between 5.65 and 8.25 V along the ion TOF path. Figures 5(a) and 5(b) depict the N2 + TOF spectrum observed using the ion bender MCP detector and Daly-type MCP ion detector, under the pulsing scheme of Fig. 3(b) except that I2 is raised from a dc potential of 5.8 to 7.5 V. By applying this dc potential barrier at I2, the PFI-

J. Chem. Phys. 137, 104202 (2012)

FIG. 4. Here, the N2 + (v+ = 1; N+ = 0–8) PFI-PIs and N2 + prompt ions [prepared using the pulsing scheme of Fig. 3(b)] are observed in the TOF spectrum of Fig. 3(c). The retarding potential curve for of N2 + (v+ = 1, N+ = 0–8) PFI-PIs measured by ramping up the dc potential of the rf-octopole reaction gas cell and monitoring PFI-PIs intensity is shown in Fig. 4(a) and the retarding potential curve for the prompt ions measured using the same procedures is depicted in Fig. 4(b). The analysis of these retarding curves allow the determination of Elab = 8.25 ± 0.20 and 5.65 ± 0.50 eV for the PFI-PIs and prompt ions, respectively.

PI TOF (36.5 μs) observed [Fig. 5(a)] is only slightly longer than that (36 μs) found in Fig. 3(c). However, the prompt ion TOF peak observed in Fig. 3(c) was indiscernible in Figs. 5(a) and 5(b), indicating that the unwanted prompt ions have been completely blocked by the potential barrier of 7.5 V set at I2. Thus, we can conclude that by applying the pulsing scheme of Fig. 3(b) to E1 and I1, and a dc voltage of 7.5 V to I2, we have successfully prepared a pure ion beam of rovibrationally selected reactant N2 + PFI-PIs with Elab = 8.25 ± 0.20 eV. In all the measurements of N2 + PFI-PIs described below, the prompt ions were rejected by setting an appropriate potential barrier at I2, such that only a pure state-selected reactant N2 + PFI-PI beam with a selected Elab was allowed to enter

FIG. 5. The pulsing scheme used for the preparation of N2 + (v+ = 1; N+ = 0–8) PFI-PIs is the same as that shown in Fig. 3(b) except that I2 is set at 7.5 V to reject the prompt ions. The TOF spectra for N2 + PFI-PIs observed using (a) the ion bender MCP detector and (b) the Daly-type MCS ion detector. Both TOF spectra reveal a single TOF peak of N2 + PFI-PIs, indicating that prompt ions are rejected by the 7.5 V potential barrier set at I2.


104202-6

Chang et al.

J. Chem. Phys. 137, 104202 (2012)

the upper reaction gas cell and to pass through the DODQ mass spectrometer for detection by the Daly-type MCP ion detectors. 2. Preparation of a pure state-selected reactant N2 + PFI-PI beam with high Elab resolutions

Under supersonic expansion conditions, the pulsed N2 MB is expected to attain a Mach number greater than 20, and thus should have a Elab resolution of about 10% or a Elab spread of ≤5 meV. Since the N2 MB is the precursor of the N2 + PFI-PI beam, we may consider the limiting Elab value for the N2 + PFI-PI beam to be ≈5 meV. The need to apply a PFI/EX electric field at the PI/PEX region can cause a broadening of Elab or an increase of Elab for the N2 + PFI-PI beam. The greatest advantage of employing the pulsing scheme is its ability to avoid the Elab broadening effect for the reactant N2 + PFI-PI beam. Assuming that the VUV laser beam at the PI/PEX region has a finite thickness (z ≈ 3 mm) and that the PI/PEX region (i.e., the region between E1 and I1) is maintained at a dc field of V/z in V/cm for ion extraction, the N2 + PFI-PIs initially formed in the PI/PEX region with the z spread of 3 mm are expected to result in a Elab spread of (z/z)(V) when they exit the PI/PEX region, where V represents the voltage difference applied to I1 and E1 and z = 1 cm. For a V value of 1 V, the Elab spread resulting from the application of the dc field is calculated to be 0.3 eV, which is significantly larger than the limiting Elab spread of ≈5 meV for the neutral N2 MB. However, the Elab spread induced by the z spread of the PFI-PIs beam under the influence of a dc ion extraction field can be avoided by using a pulsed electric field with a pulsed width shorter than the ion transit times for exiting the PI/PEX region. If the pulsed electric field is turned off before the ions exit the PI/PEX region, all ions should gain the same momentum and thus the Elab spread should vanish. When applying the pulsing scheme presented in Fig. 3(b), we expect the Elab spread for PFI-PIs beam induced by the short PFI/EX electric field pulse to be zero because it is turned off before the PFI-PI beam exiting I1 electrode. However, the fact that after the short PFI/EX electric pulse was turned off, the PFI-PIs thus formed still experienced the separation field of 2 V/cm up to t = 5.5 μs, which can also cause the Elab to spread. In order to avoid this Elab broadening effect, the PI/PEX region should be field free before the PFI-PIs exit the PI/PE region. This requirement can be readily fulfilled by turning off the separation field prior to applying the ion PFI/EX electric field pulse. Under the latter conditions, once the PFI-PIs are formed, they only experience the PFI/EX electric field pulse, but not the separation field pulse. Since the pulse width of the PFI/EX electric field pulse is shorter than the PFI-PI transit times through the PI/PEX region, the Elab spread due to the z spread of the PFI-PIs can be avoided. Figure 6 shows that retarding potential curve for PFI-PIs beam generated by an electric pulsing scheme similar to that of Fig. 3(b), except that the PFI/EX electric field pulse has an amplitude of 15 V/cm and a width of 0.7 μs, and the separation field pulse is shortened to 2.2 μs, i.e., the separation field of 2.0 V/cm is switched off before the application of the

FIG. 6. Retarding potential curve for N2 + (v+ = 1; N+ = 0–8) PFI-PIs, showing the achievement of high kinetic energy resolution at Elab = 1.25 ± 0.05 eV. The N2 + (v+ = 1; N+ = 0–8) PFI-PI beam was prepared using the pulsing scheme that comprises a separation electric field pulse with amplitude 2.0 V/cm and width 2.05 μs (turned on at t = 0.15–2.20 μs) and a PFI/EX electric field pulse with amplitude 15.0 V/cm and width 0.7 μs (turned on at t = 2.3–3.0 μs). That is, the separation field was switched off prior to the application of the ion EX/PFI field.

ion PFI/EX electric field pulse. The analysis of the retarding potential curve of Fig. 6 indicates that the PFI-PI beam has Elab = 1.25 ± 0.05 eV. We note that the achieved Elab spread of ±0.05 eV is significantly better than that (Elab = ±0.2 eV) observed in the retarding potential curve of Fig. 4(a). The achievement of Elab = ±0.05 eV also makes possible that the absolute total cross section measurements can be made down to Elab = 0.05 eV or center-of-mass kinetic energy Ecm = 0.03 eV for the ion-molecule collision of N2 + + Ar. 3. High-resolution PFI-PI spectra for the N2 + (X2 g + ; v+ = 0, 1, and 2) vibrational bands

Although rotationally resolved VUV-PFI-PE spectra for the N2 + (X2 g + ; ν + = 0–9) vibrational bands have been reported previously,49, 50 the corresponding rovibrationally resolved VUV-PFI-PI measurements for N2 + (X2 g + ; v+ = 0–9, N+ ) have not been made. We note that Kostko et al. have recently succeeded in obtaining vibrationally resolved PFI-PI spectra for N2 + (X2 g + ; v+ = 0–4) and N2 + (A2 u + ; v+ = 0–5) using monochromatized synchrotron based VUV radiation.51 Figures 7(a)–7(c) depict the VUV laser PFI-PI spectra for the N2 + (X2 g + ; v+ = 0, 1, and 2) vibrational bands (upper spectra) obtained in the ranges of 125 520–125 850, 127 720–128 050, and 129 850–130 180 cm−1 , respectively, by using the Daly-type MCP ion detector. These spectra were recorded using the pulsing scheme as described in Sec. II C 2, where the amplitude and width of the PFI/EX electric field pulse were 15 V/cm and 0.7 μs, respectively. The experimental PFI-PI spectra thus obtained reveal well-resolved rotational photoionization transitions. The simulated spectra for


104202-7

Chang et al.

FIG. 7. (a) The PFI-PI spectrum of N2 + (X2 g + ; v+ = 0; N+ = 0–6) state (upper curve) along with the best simulation spectrum (lower curve) reveals the rotational transitions classified as O-, Q-, and S-branches. (b) The PFIPI spectrum of N2 + (X2 g + ; v+ = 1; N+ = 0–8) state (upper curve) along with the best simulation spectrum (lower curve) reveals the rotational transitions classified as O-, Q-, S-, and U-branches. (c) The PFI-PI spectrum of N2 + (X2 g + ; v+ = 0; N+ = 0–9) state (upper curve) along with the best simulation spectrum (lower curve) reveals the rotational transitions classified as O-, Q-, S-, U-, and W-branches. The PFI-PI spectra were collected by using a separation field pulse with the amplitude = 2 V/cm and width = 2.05 μs (i.e., turned on at t = 0.15–2.2 μs) and a PFI/EX electric field pulse with the amplitude = 15 V/cm and width = 0.70 μs (i.e., turned on at t = 2.30 –3.00 μs). The simulated spectra reveal the rotational temperature of 25 K for the supersonic N2 MB.

the PFI-PI N2 + (X2 g + ; v+ = 0, 1, and 2) vibrational bands are shown below the corresponding experimental spectra of Figs. 7(a)–7(c). The simulation used a Gaussian instrumental line profile of 4 cm−1 (FWHM) and the known rotational constants of B0 = 1.99824 cm−1 for the neutral N2 (X1 g + ; v = 0) ground state52 and B0 + = 1.9223, B1 + = 1.9034, and

J. Chem. Phys. 137, 104202 (2012)

B2 + = 1.8842 cm−1 for the ionic N2 + (X2 g + ; ν + = 0, 1, and 2)53, 54 vibrational states, respectively. Here, we assume that the relative intensities for the rotational branches are partly governed by the thermal rotational population of N2 (X 1 g + ; v = 0). The best simulated spectra yielded a rotational temperature of 25 K for N2 (X 1 g + ; v = 0, N ). The primary motivation of the simulation is to identify the rotational photoionization transitions and thus the N+ -rotational states of N2 + (v+ = 0–2) prepared in the PFI-PI measurements. The rotational assignments based on the simulation are marked on top of the experimental spectra. The unambiguous rotational assignments also allow the determination of the IE values of 125 659.5 ± 4.0, 127 843.7 ± 4.0, and 129 974.7 ± 4.0 cm−1 for the formation of N2 + (X2 g + ; ν + = 0, 1, and 2; N+ = 0), respectively, from N2 (X 1 g + ; v = 0; N = 0). These IE values are in excellent agreement with those determined in previous PFI-PE measurements.49, 50 As shown in Figs. 7(a)–7(c), all rotational transitions observed in the PFI-PI spectra for the N2 + (X2 g + ; ν + = 0, 1, 2) vibrational bands belong to the even rotational branches with N = N+ – N = –2, 0, 2, 4, and 6, which are labeled as the O-, Q-, S-, U-, and W-branches, respectively. The lack of N = odd rotational branches, along with the dominant Q-branches, observed in the PFI-PI spectra, is consistent with previous PFI-PE (PFI-ZEKE) studies.49, 50 This observation can be interpreted based on the Buckingham-Orr-Sichel (BOS) model, which assumes a single-electron picture.55, 56 The PFI-PI spectra of Figs. 7(a)–7(c) reveal enhanced intensities for the O-, S-, U-, and W-branches, especially for the v+ = 1, 2 bands, indicating the breaking down of the BOS model in predicting rotational transition intensities for N2 + (X2 g + ; ν + = 0, 1, 2) bands.55, 56 The intensity enhancements of the O-, S-, U-, and W-branches have been attributed to field-induced rotational autoionization and the complex resonance mechanisms, by which the N2 + (X2 g + ) vibrational bands, especially the v+ = 1 and 2 states, borrow intensities from near-degenerate interloper Rydberg states that converge to higher N2 + (A2 u + ; v+ ) ionization limits.49, 50 These rotational intensity enhancement mechanisms have made the PFIPI method more favorable for the preparation of state-selected reactant N2 + , particularly at higher rovibrational states. The rotational transitions for the formation of N+ = 0–6, 0–8, and 0–9 are well resolved in the v+ = 0, 1, 2 PFI-PI vibrational bands, respectively, indicating that the reactivity of N2 + in well-defined single rovibrational states (v+ = 0, N+ = 0–6), (v+ = 1, N+ = 0–8) and (v+ = 1, N+ = 0–9) can be examined individually.

III. RESULTS AND DISCUSSION

The present experimental focus is on the measurement of absolute total rovibrationally selected cross sections [σ (v+ = 0–2; N+ = 0–9)] for the N2 + (X2 g + ; v+ = 0–2; N+ = 0–9) + Ar CT reaction and their comparison with available cross sectional data in the literature.4, 36, 41–47 Preliminary measurement of σ (v+ = 1; N+ = 0–9) at Ecm = 0.04–10.0 eV and their comparison with previous experimental and theoretical results have been communicated.35


104202-8

Chang et al.

FIG. 8. The TOF spectrum of (a) reactant N2 + (X2 g + ; v+ = 1; N+ = 0–8) PFI-PIs and (b) product ion Ar+ accumulated for 5 and 15 min, respectively, using a multichannel scalar. The kinetic energy of the N2 + PFI-PI beam was Elab = 2 .65 ± 0.05 eV (Ecm = 1.56 ± 0.03 eV).

A. Vibrationally selected absolute total cross sections σ (v+ = 0, 1, and 2)

Figures 8(a) and 8(b) compare the TOF spectra for reactant N2 + (X2 g + ; v+ = 1; N+ ) PFI-PIs and product Ar+ observed at Elab = 2.65 ± 0.05 eV. In this experiment, the VUV laser frequency was set at the maximum (1277 842.95 cm−1 ) of the Q-branch of Fig. 7(b). Since the rotational transitions for the Q-branches are not resolved, the reactant N2 + (v+ = 1) PFI-PIs thus obtained comprises several rotational states (N+ = 0–9) with their population roughly governed by a rotational temperature of ≈25 K. Both the TOF spectra for reactant N2 + (X2 g + ; v+ = 1) and product Ar+ were measured using the Daly-type MCP ion detector and a multichannel scalar (MCS). The MCS accumulation times for reactant N2 + and product Ar+ were 5 and 15 min, respectively. The Ar pressure used in the upper reaction gas cell was 1.5 × 10−4 Torr. The TOF distribution of reactant N2 + is centered at 190 μs with a FWHM of ≈20 μs), whereas the TOF distribution for product Ar+ is considerably broader, covering the range of 200–450 μs. The TOF spectrum for product Ar+ exhibits two peaks. The dominant peak at 270 μs can be attributed to slow product Ar+ ions, which correspond to backward scattered Ar+ in the center-of-mass coordinate. The fast Ar+ peak at 225 μs is significantly weaker compared to the slow peak and can be assigned to forward scattered Ar+ . This observation is consistent with the expectation that electron transfer from Ar to N2 + mostly involves the long range electron jump mecha-

J. Chem. Phys. 137, 104202 (2012)

FIG. 9. Comparison of the σ (v+ = 0), σ (v+ = 1), and σ (v+ = 2) curves determined in the Ecm range of 0.04–10.00 eV, revealing that a significant enhancement of the total cross section as v+ is increased from 0 to 2. The σ (v+ = 0), σ (v+ = 1), and σ (v+ = 2) values were measured with reactant N2 + (v+ = 0, 1, and 2) PFI-PIs prepared by setting the VUV energy at the maximum of the Q-branch shown in Figs. 7(a)–7(c), respectively. The σ (v+ = 0) curve observed here exhibits a clear onset at Ecm = 0.17 ± 0.03 eV, which is consistent with the thermochemical threshold (marked by the red arrow) for the CT reaction from N2 + (v+ = 0) + Ar → Ar+ (2 P3/2 ) + N2 (v+ = 0). The error limits represent the reproducibility of the total cross sections. As in previous cross section measurements using the guided ion beam technique, the uncertainty of the absolute total cross sections obtained in this study is estimated to be 30%.

nism with only minor momentum transfers. By summing the ion counts measured in time-bins of the MCS and normalizing the reactant and product ion counts to the same accumulation time, along with the known Ar pressure in the reaction gas cell and the effective path length of the gas cell, the σ (v+ = 1) value at Elab = 2.65 ± 0.05 eV (Ecm. = 1.56 ± 0.03 eV) was determined. In order to measure σ (v+ = 1) at other Elab values, it is necessary to change the Elab of the reactant PFI-PI beam. In the present experiment, the Elab of the N2 + PFI-PIs at the upper reaction gas cell was changed by adjusting the dc potential applied to the octopoles relative to that at the PI/PEX region. This approach has the advantage of keeping the operation conditions of the molecular beam VUV laser PFI-PI source unchanged, and thus the Elab spread for the reactant PFI-PI beam can stay constant (Elab = ± 0.05 eV) as Elab is changed. The vibrationally selected absolute total cross sections σ (v+ = 0), σ (v+ = 1), and σ (v+ = 2) at Ecm = 0.04–10.00 eV observed in this study are compared in Fig. 9. These σ (v+ = 0), σ (v+ = 1), and σ (v+ = 2) values along with their error limits are also listed at Table I. The error limits given here represent the reproducibility of the measurements. As in similar absolute total cross sections measurements,7 we have assigned an error limit of ±30% for all the σ (v+ = 0), σ (v+ = 1), and σ (v+ = 2) values determined in the present study.


104202-9

Chang et al.

J. Chem. Phys. 137, 104202 (2012)

TABLE I. Absolute total cross sections σ (v+ ) in Å2 for the vibrationalstate-selected charge transfer reaction, N2 + (X; v+ ) + Ar → N2 + Ar+ . Absolute total cross sections: σ (v+ ) (Å2 ) Ecm (eV) 0.04 0.05 0.06 0.07 0.08 0.10 0.15 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.20 1.50 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00

v+ = 0

v+ = 1

v+ = 2

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.04 ± 0.02 0.05 ± 0.03 0.07 ± 0.04 0.07 ± 0.03 0.07 ± 0.01 0.09 ± 0.01 0.09 ± 0.03 0.09 ± 0.03 0.11 ± 0.01 0.15 ± 0.04 0.18 ± 0.04 0.24 ± 0.04 0.26 ± 0.02 0.25 ± 0.02 0.25 ± 0.06 0.24 ± 0.06 0.22 ± 0.07 0.24 ± 0.05 0.25 ± 0.06 0.20 ± 0.06

46.76 ± 4.02 42.30 ± 2.58 37.79 ± 3.73 36.37 ± 2.33

74.53 ± 3.54 70.08 ± 4.24 63.99 ± 4.32 62.05 ± 6.37

28.88 ± 1.91

54.93 ± 3.85

21.60 ± 1.89 19.18 ± 1.43 17.84 ± 2.81 17.82 ± 1.57 16.29 ± 1.94 15.78 ± 0.88 14.78 ± 1.02 14.02 ± 0.76 14.02 ± 0.76 13.56 ± 0.51 12.87 ± 1.13 13.30 ± 1.18 13.24 ± 0.56 14.01 ± 1.35 13.36 ± 0.60 13.47 ± 0.48 13.27 ± 0.94 13.21 ± 1.10 13.22 ± 1.04 13.46 ± 0.79

45.18 ± 2.07 40.84 ± 3.58 40.38 ± 3.68 40.38 ± 3.89 37.73 ± 2.31 36.38 ± 2.80 36.04 ± 3.83 35.65 ± 2.74 36.06 ± 1.38

35.98 ± 3.54 36.36 ± 3.21 37.38 ± 3.49 37.57 ± 4.94 37.38 ± 3.58 37.73 ± 3.58 37.53 ± 2.29 37.69 ± 1.00 37.69 ± 5.58

We note that the σ (v+ = 1) curve at Ecm = 0.04–10.00 eV reported here differs slightly from that reported recently due to a correction of the calibration factor used. Similar to the preparation of N2 + (v+ = 1), the N2 + (v+ = 0) and N2 + (v+ = 2) reactant ions were generated by setting the VUV laser frequencies at the maxima of the Q-branches of the PFIPI spectra shown in Figs. 7(a) and 7(c). The comparison in Fig. 9 shows that the total CT cross section is greatly enhanced as v+ is increased from 0 to 2. The present study provides new cross sectional data for v+ = 1 and 2 in the Ecm range of 0.04–1.00 eV, which also exhibit strong Ecm dependencies. The detailed comparison between the σ (v+ = 0), σ (v+ = 1), and σ (v+ = 2) values observed in this experiment and the cross sectional data reported in previous experimental4, 36, 41–44 and theoretical45–47 studies are shown in Figs. 10–12, respectively. The N2 + (X2 g + ; v+ = 0) + Ar → N2 (X1 g + ; v = 0) + Ar+ (3 P3/2 ) reaction is endothermic by 0.179 eV.41 As shown in Fig. 9, the σ (v+ = 0) curve observed here exhibits a distinct Ecm onset at 0.17 ± 0.03 eV, which is in excellent agreement with the thermochemical threshold for the N2 + (v+ = 0) + Ar CT collision.41 The small σ (v+ = 0) values of ≤0.25 Å2 for σ (v+ = 0) and the increasing trend of σ (v+ = 0) observed as Ecm is increased from the threshold are consistent with the behavior of an endothermic CT reaction. The observation of the Ecm onset for σ (v+ = 0) supports the conclusions that the Ecm spread achieved in the present experi-

FIG. 10. Comparison of σ (v+ = 0) values at Ecm = 0.04–10.00 eV obtained in the present study with cross sectional data at Ecm ≥ 1 eV reported in previous experimental measurements by Schultz and Armentrout,41 Shao et al.,36 and Govers et al.,4 and theoretical calculations by Parlant and Gislason.45 The total cross sections at Ecm ≈ 0.04 eV converted from the thermal rate coefficients of Smith and Adams42 and Kato et al.44 are also included in the figure. The converted total cross sections at Ecm ≥ 1 eV based on the rate coefficients obtained by Lindinger et al.43 are found to be significantly higher than the total cross sections obtained by other experiments. The high rate coefficients (or converted total cross sections) of Lindinger et al. were later attributed to the contamination of the reactant N2 + (v+ = 0) by excited N2 + (v+ = 1).

ment is ±0.03 eV, and that the contamination of the N2 + (v+ = 0) PFI-PI beam from excited N2 + (v+ ≥ 1), which are known to have significantly higher cross sections compared to σ (v+ = 0), is negligibly small.4, 36, 41–44 The comparison in Fig. 10 indicates that the agreement is lacking at Ecm = 3.0–10.00 eV, where the previous experimental and theoretical cross sections are higher than

FIG. 11. Comparison of σ (v+ = 1) values at Ecm = 0.04–10.00 eV obtained in the present study with absolute total cross sections at Ecm ≥ 1 eV reported in previous experimental measurements by Shao et al.36 and Govers et al.4 and theoretical calculations by Parlant and Gislason45 and Candori et al.46, 47 The total cross sections at Ecm ≈ 0.04 eV converted from the thermal rate coefficients of Kato et al.44 and Lindinger et al.43 are also included in the figure.


104202-10

Chang et al.

FIG. 12. Comparison of σ (v+ = 2) values at Ecm = 0.04–10.00 eV obtained in the present study with absolute total cross sections at Ecm ≥ 1 eV reported in previous experimental measurements by Shao et al.36 and Govers et al.4 and theoretical calculations by Parlant and Gislason45 and Candori et al.46, 47 The total cross section at Ecm ≈ 0.04 eV converted from the thermal rate coefficients of Kato et al.44 is also included in the figure. The dashed curve represents the prediction of Candori et al.46, 47

the present σ (v+ = 0) measurement. By combining a flowtube N2 + ion source with a guide-ion beam mass spectrometer, Schultz and Armentrout41 were able to measure σ (v+ = 0) at Ecm = 0.03–20 eV. After taking into account the experimental uncertainties, our results are consistent with the σ (v+ = 0) values measured by Schultz and Armentrout41 except at Ecm > 3eV, where their results are about 2.5 folds higher than our values. Shao et al.36 and Govers et al.4 have previously reported the measurements of σ (v+ = 0), σ (v+ = 1), and σ (v+ = 2) at Ecm > 1 eV using the VUV photoionization and the threshold photoelectron-photoion coincidence (TPEPICO) methods, respectively. Their σ (v+ = 0) values are higher than the results of Schultz and Armentrout41 and the present PFI-PI study. Parlant and Gislason have calculated the total cross sections for the N2 + (v+ = 0, 1, and 2) + Ar CT reactions based on the classical path method.45 These experimental and theoretical total cross sections are included in the comparisons shown in Figs. 10–12. At thermal energies (Ecm ≈ 0.04 eV), the rate coefficients (kT ) for the N2 + + Ar reaction have been measured for N2 + (v+ = 0) by Smith and Adams42 and for N2 + (v+ = 0, 1, and 2) by Kato et al.44 using the flow-tube method. The kT values for N2 + (v+ = 0 and 1) at Ecm > 1 eV have also been measured by Lindinger and co-workers.43 We have converted these kT values into cross section (σ ) by using the relation σ = kT (μ/2Ecm )1/2 , where μ is the reduced mass of N2 + and Ar. The σ values converted from the rate coefficient measurements are included in Figs. 10–12. The converted σ values of Lindinger et al.43 for N2 + (v+ = 1) are found to be significantly higher than the σ (v+ = 0) values obtained by this and other studies. The high rate coefficients for N2 + (v+ = 0) observed by Lindinger et al.43 were later realized due to the contamination of reactant N2 + (v+ = 0) by excited N2 + (v+

J. Chem. Phys. 137, 104202 (2012)

= 1). The comparison of the σ (v+ = 0) values in Fig. 10 also suggests that the thermal kT value obtained by Kato et al.44 is likely too high. Considering that the vibrationally excited N2 + (v+ = 1 and 2) states are 0.271 and 0.535 eV higher than the ground N2 + (v+ = 0) state, the N2 + (v+ = 1 and 2) + Ar CT reactions become exothermic by 0.09 and 0.356 eV, respectively. The significantly higher σ (v+ = 1) and σ (v+ = 2) compared to σ (v+ = 0) observed are consistent with the exothermic nature of the CT reactions. Furthermore, the Ecm -dependencies observed for σ (v+ = 1) and σ (v+ = 2) are also different from that of σ (v+ = 0). The σ (v+ = 1) [σ (v+ = 2)] is found to decrease rapidly from 46 Å2 (91 Å2 ) at Ecm = 0.04 to 14 Å2 (38 Å2 ) at Ecm = 1.00 eV. At Ecm = 1.00–10.00 eV, the σ (v+ = 1) [σ (v+ = 2)] remains nearly constant with the value of 14 Å2 (38 Å2 ). The σ (v+ = 1) and σ (v+ = 2) values reported previously are limited to Ecm > 1 eV. The studies of Shao et al.36 and Govers et al.4 show that the excited N2 + (v+ = 1) state reacts much more effectively than the ground N2 + (v+ = 0) state, and that additional excitation to N2 + (v+ = 2) only moderately increases the CT cross section.3, 42 However, based on our measurements shown in Fig. 9, the cross section for N2 + (v+ = 2) is about two times higher than that for N2 + (v+ = 1) in the Ecm range of 0.04–10 eV; but the σ (v+ = 1) and σ (v+ = 2) curves are found to have a similar profile for the Ecm -dependence. We find that the σ (v+ = 1) values obtained in this PFI-PI study are the lowest (Fig. 11), while those for σ (v+ = 2) are the highest (Fig. 12) in comparison with the previous available cross sectional data measured at the corresponding Ecm values. The converted cross sections for v+ = 1 obtained by Smith and Adams42 and Kato et al.44 are in excellent agreement (see Fig. 11). Although the latter converted cross sections are higher than the σ (v+ = 1) value measured at Ecm = 0.04 eV in the present study, they can still be considered as in agreement after taking into account the experimental uncertainty of ±30% for the absolute total cross sections obtained in the present experiment. Similarly, the converted cross section for v+ = 2 based on the thermal rate coefficient obtained by Kato et al.44 (see Fig. 12) can also be considered as consistent with the σ (v+ = 2) value determined in the present study after taking into account the experimental error limits. The dash curve in Fig. 11 represents the theoretical predictions for σ (v+ = 1) calculated recently by Candori et al.46 In the latter theoretical study, the potential energy surfaces (PES) leading to the entrance channels [N2 + (X,v+ ) + Ar) and exit channels [Ar+ (2 P3/2,1/2 ) + N2 (X,v )] were calculated.47 The CT was treated as non-adiabatic transitions between the entrance and exit channels, and the transition probability was calculated based on the Landau-Zener-Stückelberg (LZS) formulism.46 Considering that the absolute total cross sections determined in the present study have an uncertainty of about ±30%, we conclude that the σ (v+ = 1) values determined here at Ecm = 0.04–10.00 eV are in fair agreement with the LZS predictions. This conclusion can be taken as validation of the LZS model and for the PES used for the [N2 + Ar]+ reaction system. The dash curve shown in Fig. 12 represents the σ (v+ = 2) predictions based on the theoretical LZS calculation by


104202-11

Chang et al.

J. Chem. Phys. 137, 104202 (2012)

Candori et al.,46, 47 in which the theoretical σ (v+ = 2) values are obtained as the sum of the cross sections for the three product channels: [Ar+ (2 P3/2 ) + N2 (v = 1)], [Ar+ (2 P1/2 ) + N2 (v = 0)], and [Ar+ (2 P3/2 ) + N2 (v = 0)]. The lack of agreement between the LZS predictions and our measurements for σ (v+ = 2) is not surprising because the nearby product channels, such as [Ar+ (2 P1/2 ) + N2 (v = 1)] and [Ar+ (2 P3/2 ) + N2 (v = 2)], were not included in calculating the total cross sections. The agreement between the LZS predictions and the present measurement of σ (v+ = 2) is expected to improve when these product channels are taken into consideration. Furthermore, the product channel ArN+ + N was observed in the previous state-selected collision study of Ar+ (2 P3/2,1/2 ) + N2 at Ecm ≥ 8.24 eV,37 indicating that this product channel should also be included in the theoretical calculation of the total cross sections for the N2 + (X2 g + ; v+ = 0, 1 and 2; N+ ) + Ar collisions at Ecm ≥ 8.24 eV. B. Rotational dependence of the charge transfer cross section

We note that several previous studies have also examined the rotational effects in bimolecular ion-molecule reactions by using the REMPI and PFI-PI techniques.57–59 The cleanly resolved rotational transitions observed in the PFIPI N2 + (X2 g + , v+ = 0, 1, 2) vibrational bands depicted in Figs. 7(a)–7(c) indicate that it is feasible to examine the reactivity of N2 + (X2 g + , v+ = 0, 1, and 2) ions prepared in single rovibrational states. We have recently reported the absolute total cross sections σ (v+ = 1, N+ ) for N+ = 0–8 at Ecm = 1.56 ± 0.11 eV, and found that σ (v+ = 1) is independent of N+ .35 Here we present the absolute total cross sections σ (v+ = 2, N+ ) for N+ = 0–9 measured at Ecm = 0.05 and 1.00 eV. Since the total (v+ , N+ )-state-selected charge transfer cross section depends on the ratio of the intensity of product Ar+ to that of reactant N2 + (v+ , N+ ), the determination of σ (v+ = 2, N+ ) for N+ = 0–9 can be made by measuring the PFI-PI spectra for the corresponding reactant N2 + (v+ = 2, N+ ) and product Ar+ . Figure 13(a) shows the PFI-PI spectrum for reactant N2 + (X2 g + , v+ = 2, N+ = 0–9) (lower black curve) prepared at Ecm = 0.05 ± 0.03 eV, while the corresponding PFI-PI spectrum for product Ar+ is plotted in Fig. 13(b). The simulated spectrum (upper blue curve) along with the rotational assignments is also given in Fig. 13(a). The conditions and procedures for this experiment are the same as those described above except that the VUV laser frequency was scanned to prepare N2 + (v+ = 2) in different accessible N+ -rotational states. During the measurement, the product QMS was programmed to switch alternatively to mass 28 and 40 amu for the accumulations of ion counts for reactant N2 + (for 5 s) and product Ar+ (for 10 s), respectively. The observation that the spectra for reactant N2 + (v+ = 2, N+ = 0–9) and product Ar+ are nearly identical indicates that σ (v+ = 2) at Ecm = 0.05 ± 0.03 eV is essentially independent of N+ . The PFI-PI spectra for N2 + (v+ = 2, N+ = 0–9) and the corresponding product Ar+ were also recorded at Ecm = 1.00 eV. Based on the PFI-PI spectrum measurements, we have obtained the values for σ (v+ = 2, N+ ), N+ = 0–9, at

FIG. 13. Comparison of (a) the PFI-PI spectrum (lower curve) for reactant N2 + (v+ = 2; N+ = 0–9) prepared at Ecm = 0.05 ± 0.03 eV with (b) the corresponding spectrum observed by detecting product Ar+ . The simulated spectrum [upper curve shown in (a)] reveals the rotational transitions of the O-, Q-, S-, and U-branches as marked on the top of the simulated spectrum.

Ecm = 0.05 and 1.00 eV as depicted in Figs. 14(a) and 14(b), respectively, which are found to be essentially constant as N+ is varied in the range of 0–9. This together with the recently reported measurement for σ (v+ = 1, N+ ), N+ = 0–8, at Ecm = 1.56 ± 0.11 eV,35 we conclude that the total cross section

FIG. 14. The plot of σ (v+ = 2, N+ ) as a function of N+ for N+ = 0–9 measured at (a) Ecm = 0.05 ± 0.03 eV and (b) Ecm = 1.00 ± 0.03 eV.


104202-12

Chang et al.

for the CT collision of N2 + (v+ = 1 and 2) + Ar is independent of N+ . We note that the N+ = 9 level of N2 + contains about 200 cm−1 of internal energy. If increasing the rotational energy simply promotes an energy effect, the change in CT cross section within such a small energy range may be too small to be observed. IV. CONCLUSIONS AND FURTHER ADVANCES

This experiment demonstrates a successful implementation of a high-resolution VUV laser PFFI-PI ions source with a DQDO mass spectrometer for absolute total cross section measurements of state-selected ion-molecule reactions. We have shown that rovibrationally selected reactant N2 + can be prepared not only with high intensity and high quantum state purity, but also high collision energy resolution (Ecm = 0.03 eV). By employing this newly developed VUV laser PFI-PI DODQ apparatus, we have measured the absolute total cross sections for the N2 + (X2 g + ; v+ = 0, 1, and 2; N+ = 0–9) + Ar CT reaction at Ecm = 0.04–10.00 eV. This experiment shows that vibrational excitation of N2 + (X2 g + ) is significantly more effective than translational kinetic energy in promoting the N2 + (X2 g + ) + Ar CT reaction, whereas the total CT cross sections for the N2 + (X2 g + ; v+ = 1 and 2) + Ar collision are found to be independent of the rotational excitation of N2 + (X2 g + ; v+ = 1 and 2) in the range of N+ = 0–9. The total absolute cross section data obtained in the present study have allowed a detailed comparison of the theoretical predictions based on the LZS model. Although the present experiment is focused on the total cross section measurement of the N2 + (X2 g + ; v+ = 1 and 2; N+ = 0–9) + Ar CT reactions, the procedures described are generally applicable for the studies of other state-selected ion-molecule reactions. The success in the development of the high-resolution VUV laser PFI-PI method for the preparation of rovibrationally selected reactant ions with high intensity and high kinetic energy resolution promises to open new possibilities for the studies of unimolecular and biomolecular ion spectroscopy and dynamics. The present VUV laser PFI-PI scheme is shown to be able to avoid the Elab broadening effect of the reactant PFI-PI beam. This accomplishment along with the pulsed nature of the PFI-PI beam makes it logical to replace the reaction gas cell with a pulsed neutral supersonic beam to introduce the neutral reactants for crossed ionneutral beam measurements. In crossed ion-neutral beam collision studies, the pulsed neutral beam would be turned on only to time the arrival of the pulsed PFI-PI beam. This can reduce the gas load of the pumping system; as a result the background ions produced by the reactant PFI-PIs with ambient gaseous molecules can be significantly reduced. Furthermore, it is well known that the replacement of the reaction gas cell by using a neutral supersonic beam to conduct crossed ion-neutral beam studies can further improve the kinetic energy resolution,60 allowing cross section measurements to be made at Ecm values below thermal energies. Many ion-molecule reactions of relevance to plasma and planetary chemistry involve atoms and radicals. Thus, there is a great need to measure the reaction cross sections of ion-

J. Chem. Phys. 137, 104202 (2012)

atom and ion-radical reactions. Atoms and radicals in well defined internal states can be readily generated in the form of a molecular beam with well characterized kinetic energies by pyrolysis61 or by pulsed ultraviolet laser photodissociation.62 The pulsed nature of both the neutral atomic or radical beam as well as the reactant PFI-PI ion beam makes it ideal to conduct total cross section measurements of state-selected ionatom or ion-radical reactions by using the crossed ion-neutral beam arrangement.

ACKNOWLEDGMENTS

The experimental work was supported by the National Aeronautics and Space Administration (NASA) Planetary Atmospheres Program (Grant No. 07-PATM07-0012). The construction of the double-quadrupole-double-octopole apparatus used in this study was financed by the U.S. Department of Energy (DOE) Basic Energy Sciences (DE-FG0202ER15306). The VUV laser system used was constructed by the support of the National Science Foundation Instrumental Grant (CHE0342829). 1 W.

A. Chupka and M. E. Russell, J. Chem. Phys. 49, 5426 (1968). L. Anderson, Adv. Chem. Phys. 82, 177 (1992). 3 R. J. S. Morrison, W. E. Conway, T. Ebata, and R. N. Zare, J. Chem. Phys. 84, 5527 (1986). 4 T. R. Govers, P. M. Guyon, T. Baer, K. Cole, H. Fröhlich, and M. Lavollee, Chem. Phys. 87, 373 (1984). 5 I. Koyano and K. Tanaka, Adv. Chem. Phys. 82, 263 (1992). 6 C. Y. Ng, Adv. Chem. Phys. 82, 401 (1992). 7 C. Y. Ng, J. Phys. Chem. A 106, 5953 (2002). 8 R. A. Dressler, Y. Chiu, D. J. Levandier, X. N. Tang, Y. Hou, C. Chang, C. Houchins, H. Xu, and C. Y. Ng, J. Chem. Phys. 125, 132306 (2006). 9 X.-M. Qian, T. Zhang, C. Chang, P. Wang, C. Y. Ng, Y. Chiu, D. J. Levandier, J. S. Miller, R. A. Dressler, T. Baer, and D. S Peterka, Rev. Sci. Instrum. 74(9), 4096 (2003). 10 V. G. Anicich, Astrophys. J., Suppl. Ser. 84, 215–315 (1993); and references therein. 11 C. Y. Ng, Adv. Chem. Phys. 52, 263–362 (1983). 12 C. Y. Ng, Adv. Ser. Phys. Chem. 10A, 394 (2000). 13 C. Y. Ng, Int. J. Mass. Spectrom. 200, 357 (2000). 14 C. Y. Ng, Annu. Rev. Phys. Chem. 53, 101 (2002). 15 E. W. Schlag, ZEKE Spectroscopy (Cambridge University Press, Cambridge, 1998). 16 P. M. Johnson, Adv. Ser. Phys. Chem. 10A, 296 (2000). 17 Y. Song, M. Evans, C. Y. Ng, C.-W. Hsu, and G. K. Jarvis, J. Chem. Phys. 112, 1306–1315 (2000). 18 M. Evans and C. Y. Ng, J. Chem. Phys. 111, 8879–8892 (1999). 19 Y. Song, M. Evans, C. Y. Ng, C.-W. Hsu, and G. K. Jarvis, J. Chem. Phys. 111, 1905 (1999). 20 G. K. Jarvis, M. Evans, C. Y. Ng, and K Mitsuke, J. Chem. Phys. 111, 3058 (1999). 21 S. Stimson, M. Evans, C.-W. Hsu, and C. Y. Ng, J. Chem. Phys. 126, 164303 (2007). 22 C. Chang, C. Y. Ng, S. Stimson, M. Evans, and C.-W. Hsu, Chin. J. Chem. Phys. 20, 352 (2007). 23 G. K. Jarvis, Y. Song, and C. Y. Ng, Rev. Sci. Instrum. 70, 2615 (1999). 24 C. Y. Ng, J. Electron. Spectrosc. Relat. Phenom. 112, 31 (2000). 25 G. K. Jarvis, K.-M. Weitzel, M. Malow, T. Baer, Y. Song, and C. Y. Ng, Phys. Chem. Chem. Phys. (Commun.) 1, 5259 (1999). 26 K.-M. Weitzel, M. Malow, G. K. Jarvis, T. Baer, Y. Song, and C. Y. Ng, J. Chem. Phys. 111, 8267 (1999). 27 K.-C. Lau and C.-Y. Ng, Acc. Chem. Res. 39, 823–829 (2006). 28 G. K. Jarvis, K.-M. Weitzel, M. Malow, T. Baer, Y. Song, and C. Y. Ng, Rev. Sci. Instrum. 70, 3892 (1999). 29 T. Zhang, X.-M. Qian, P. Wang, C. Y. Ng, Y. Chiu, D. J. Levandier, J. S. Miller, and R. A. Dressler, J. Chem. Phys. 119, 10175 (2003). 2 S.


104202-13

Chang et al.

30 X. M. Qian, T. Zhang, Y. Chiu, D. J. Levandier, J. S. Miller, R. A. Dressler,

and C. Y. Ng, J. Chem. Phys. 118, 2455–2458 (2003). 31 X. N. Tang, H. F. Xu, C. Houchins, C. Y. Ng, Y. Chiu, R. A. Dressler, and D. J. Levandier, J. Chem. Phys. 126, 234305 (2007). 32 X. N. Tang, H. F. Xu, T. Zhang, Y. Hou, C. Chang, C. Y. Ng, Y. Chiu, R. A. Dressler, and D. J. Levandier, J. Chem. Phys. 122, 164301 (2005). 33 W. B. Tzeng and J. L. Lin, J. Phys. Chem. A 103, 8612 (1999). 34 S. T. Park, S. Y. Kim, and M. S. Kim, J. Chem. Phys. 114, 5568 (2001). 35 Y. C. Chang, H. Xu, Y. Xu, Z. Lu, Y.-H. Chiu, D. J. Levandier, and C. Y. Ng, J. Chem. Phys. 134, 201105 (2011). 36 J. D. Shao, Y. G. Li, G. D. Flesch, and C. Y. Ng, J. Chem. Phys. 86, 170 (1987). 37 G. D. Flesch and C. Y. Ng, J. Chem. Phys. 92, 2876 (1990). 38 C. L. Liao, J. D. Shao, R. Xu, G. D. Flesch, Y. G. Li, and C. Y. Ng, J. Chem. Phys. 85, 3874 (1986). 39 C. L. Liao, R. Xu, and C. Y. Ng, J. Chem. Phys. 85, 7136 (1986). 40 C. L. Liao, R. Xu, and C. Y. Ng, J. Chem. Phys. 84, 1948 (1986). 41 R. H. Schultz and P. B. Armentrout, Chem. Phys. Lett. 179, 429 (1991). 42 D. Smith and N. G. Adams, Phys. Rev. A 23, 2327 (1981). 43 W. Lindinger, F. Howorka, P. Lukac, S. Kuhn, H. Villinger, E. Alge, and H. Ramler, Phys. Rev. A 23, 2319 (1981). 44 S. Kato, J. A. de Gouw, C.-D. Lin, V. Bierbaum, and S. Leone, Chem. Phys. Lett. 256, 305 (1996). 45 G. Parlant and E. A. Gislason, J. Chem. Phys. 91, 5359 (1989). 46 R. Candori, F. Pirani, D. Cappellltti, P. Tosi, and D. Bassi, Int. J. Mass. Spectrom. 223–224, 499 (2003). 47 R. Candori, S. Canvalli, F. Pirani, A. Volpi, D. Cappelletti, P. Tosi, and D. Bassi, J. Chem. Phys. 115, 8888 (2001).

J. Chem. Phys. 137, 104202 (2012) 48 H.

K. Woo, K.-C. Lau, J. Zhan, C. Y. Ng, Y.-S. Cheung, W. K. Li, and P. M. Johnson, J. Chem. Phys. 119, 7789–7799 (2003). 49 J. W. Hepbun, J. Chem. Phys. 107, 7106 (1997). 50 F. Merkt and T. P. Softley, Phys. Rev. A 46(1), 302 (1992). 51 O. Kostko, S. K. Kim, S. R. Leone, and M. Ahmed, J. Phys. Chem. A 113(52), 14206 (2009). 52 K. P. Huber and G. Herzberg, Molecular Spectra and Molecular Structure. IV. Constants of Diatomic Molecules (Van Norstand, New York, 1974), p. 426. 53 T. A. Miller, T. Suzuki, and E. Hirota, J. Chem. Phys. 80, 4671 (1984). 54 D. W. Ferguson, K. N. Rao, P. A. Martin, and G. Guelachvili, J. Mol. Spectrosc. 153, 599 (1992). 55 A. D. Buckingham, B. J. Orr, and J. M. Sichel, Philos. Trans. R. Soc. London, Ser. A 268, 147 (1970). 56 Y. Itikawa, Chem. Phys. 28, 461 (1978). 57 R. J. Green, J. Qian, H.-T. Kim, and S. L. Anderson, J. Chem. Phys. 113, 3002 (2000). 58 S. R. Mackenzie and T. P. Softley, J. Chem. Phys. 101, 10609 (1994). 59 L. Paetow, F. Unger, B. Beutel, G. Frenking, and K. M. Weitzel, J. Chem. Phys. 132, 174305 (2010). 60 P. Tosi, O. Dmitrijev, and D. Bassi, Chem. Phys. Lett. 200, 483 (1992). 61 X. Zang, A. V. Friderichsen, J. T. Mckinnon, T. G. Lindeman, D. E. David, D. C. Dayton, S. Nandi, G. B. Ellison, and M. R. Nimlos, Rev. Sci. Instrum. 74, 3077 (2003). 62 S. Nourbakhsh, K. Norwood, G.-Z. He, and C. Y. Ng, J. Am. Chem. Soc. 113, 6311 (1991).
