We present here the design of a new tandem mass spectrometer that combines an electrospray ion ... molecules and to test the ability of theory to predict these.
REVIEW OF SCIENTIFIC INSTRUMENTS 81, 073107 共2010兲
A new tandem mass spectrometer for photofragment spectroscopy of cold, gas-phase molecular ions Annette Svendsen, Ulrich J. Lorenz, Oleg V. Boyarkin, and Thomas R. Rizzo Laboratoire de Chimie Physique Moléculaire, Ecole Polytechnique Fédérale de Lausanne, EPFL SB ISIC LCPM, Station 6, CH-1015 Lausanne, Switzerland
共Received 1 May 2010; accepted 8 June 2010; published online 20 July 2010兲 We present here the design of a new tandem mass spectrometer that combines an electrospray ion source with a cryogenically cooled ion trap for spectroscopic studies of cold, gas-phase ions. The ability to generate large ions in the gas phase without fragmentation, cool them to ⬃10 K in an ion trap, and perform photofragment spectroscopy opens up new possibilities for spectroscopic characterization of large biomolecular ions. The incorporation of an ion funnel, together with a number of small enhancements, significantly improves the sensitivity, signal stability, and ease of use compared with the previous instrument built in our laboratory. © 2010 American Institute of Physics. 关doi:10.1063/1.3458014兴
I. INTRODUCTION
Since the mid-1980s, considerable interest has been focused on spectroscopic investigations of isolated biological molecules in the gas phase.1–13 The motivation for many of these studies has been to understand the intrinsic spectroscopic and conformational properties of biologically related molecules and to test the ability of theory to predict these properties accurately. While amino acids and small peptides can be volatilized thermally and studied in the cold environment of a supersonic expansion, the extension of this approach to larger molecules is made difficult by their thermal lability. During this same period, the extension of mass spectrometric techniques to biological molecules has literally exploded with the advent of ionization techniques such as electrospray14 and matrix assisted laser desorption ionization,15 which can put molecules of virtually any mass into the gas phase with little or no fragmentation. Many groups have recently begun to capitalize on these developments by performing spectroscopic studies of electrosprayed ions.9,12,13,16–24 The major advantage of supersonic molecular beam techniques for spectroscopic studies is the low internal temperatures that can be achieved, which greatly simplifies the spectra of large molecules and reveals sharp spectral features related to their structure. The challenge for spectroscopic studies of larger biological molecules is thus to find ways to cool the ions produced by electrospray to similar internal temperatures. Gerlich25,26 has pioneered the use of highorder multipole rf ion traps for cooling molecules of astrophysical interest, and we have recently shown that these techniques can be coupled with electrospray ionization and employed for spectroscopic studies of biological molecules.13,22–24,27 Building on this initial work, we report here a second-generation apparatus for spectroscopic studies of electrospray-generated ions, which greatly improves on the performance of the original machine. While our focus is 0034-6748/2010/81共7兲/073107/7/$30.00
on applications to biological molecules, this apparatus may also be well suited for studies of synthetic polymers, transition-metal complexes, or nanoparticles. II. INSTRUMENTATION
Figure 1 shows an overview of the tandem mass spectrometer designed for photofragment spectroscopy of cold, biomolecular ions. The major components of this apparatus are a nanospray-based mass-selective ion source, a cryogenic multipole rf ion trap, and a mass-selective ion detection system. The ion optical elements are housed in a vacuum chamber divided into seven differentially pumped stages. The following sections describe this apparatus in detail. A. Nanospray source
Figure 2 shows the details of the front end of the apparatus comprising a nanospray source, an ion funnel, and a hexapole ion guide. The nanospray source 共Proxeon, ES070, Odense, Denmark兲 consists of a metalized borosilicate needle 关1 m inner diameter 共ID兲 tip兴, which serves as the ion emitter 共not shown in Fig. 2兲. The needle is connected to a sample reservoir 共a microcentrifuge tube兲 through a metal holder such that the liquid is contained in a closed volume. In order to facilitate flow through the needle, this volume can be pressurized. The sample holder is mounted on a xyz translational stage but electrically insulated such that the needle tip can be biased at 500–1000 V with respect to ground. Ions produced by nanospray are transferred into vacuum through a capillary made from an 11.5 cm long stainless steel tube 关0.5 mm ID, 2 mm outer diameter 共OD兲兴 inserted into a copper block, which can be heated to 250 ° C by two 100 W cartridge heaters 共Prang⫹Partner AG, Pfungen, Switzerland兲. The copper block is surrounded by a polyether ether ketone jacket that isolates it thermally and electrically from the source flange into which the parts are mounted. The capillary is typically biased at 100–200 V.
81, 073107-1
© 2010 American Institute of Physics
073107-2
Rev. Sci. Instrum. 81, 073107 共2010兲
Svendsen et al.
Laser port
Quadrupole mass filter
VRF (+) Channeltron detector
Quadrupole bender
10 nF
10 nF
Vent
Quadrupole bender
Channeltron detector
10 nF
Vexit 1 MΩ
1 MΩ
0.5 MΩ
22pole ion trap
Quadrupole bender
Quadrupole mass filter Octopole Gate valve
Channeltron detector
Hexapole
Laser port
Vent
Ion funnel
VRF (-)
Heated capillary
FIG. 1. Overview of the tandem photofragment mass spectrometer.
B. Ion funnel
The capillary is followed by an ion funnel, which is a rf ion guide that effectively collects and focuses the divergent stream of ions exiting the capillary. The design of the funnel is based on those of Smith and co-workers.28,29 It consists of 100 stainless steel plates, 0.5 mm thick, separated by 0.5 mm thick Teflon spacers. The first section of the funnel has 57 plates, all having a central aperture of 25.4 mm diameter. The 20th plate of this section 共located ⬃2 cm from the funnel entrance兲 has been modified slightly to disperse the jet of neutrals exiting the capillary. This jet-disrupter plate has been cut such that it has a disk of diameter 6.5 mm located in the center of the 25.4 mm aperture and connected to the remainder of the plate by four 0.5 mm wide stripes forming a cross. The second section consists of 42 plates with an aperture diameter that decreases uniformly from 25.4 to 2.5 mm. The last plate has a 1.5 mm hole that serves as the conductance limit between two consecutive differential pumping stages. The electrical connections to the funnel are made by way Ion funnel Gate valve
Hexapole
Capillary
FIG. 2. Detailed view of the ion source region. For clarity, only every fourth plate of the ion funnel is shown.
0.5 MΩ
1 MΩ 10 nF
1 MΩ 10 nF
Vexit 10 nF
FIG. 3. Schematic drawing of the electrical circuits for connecting the ion funnel.
of a custom-made zero-insertion-force 共ZIF兲 socket 共Tactic Electronics, Plano, TX兲, into which a small tab on each plate inserts. Onto the opposite side of the ZIF socket is soldered a circuit board containing two networks of capacitors and resistors, shown schematically in Fig. 3. One network connects all odd-numbered plates to one phase of a rf potential with a dc potential gradient superimposed, while the second network connects all even-numbered plates to the other phase of the rf potential combined with the dc potential gradient. The peak-to-peak rf potential is typically 35 V. The dc potential for the first electrode is 190 V and that for the last electrode in the series is 10 V, resulting in a dc gradient of 18 V/cm. The jet disrupter and the conductance limit plate 共i.e., the last plate of the funnel兲 are connected separately and carry only a dc potential. The jet disrupter is normally held at 150 V while the conductance limit plate is biased at 14 V. The pressure in the ion funnel region is 1.5 mbar and is maintained by a rotary vane pump 共Alcatel, Annecy, France兲 with a pumping speed of 60 m3 / h. C. Hexapole ion guide or trap
Ions leaving the funnel enter a 17 cm long hexapole ion guide 共Analytica of Branford, Branford, CT兲 made of six gold-coated ceramic rods of 2 mm diameter positioned on an inscribed diameter of 4 mm. Two 4 MHz rf signals of opposite phase are used to drive the device; typical values for the amplitude and the dc offset are 100–200 V and 10 V, respectively. A dc lens placed at the exit of the hexapole guide allows the guide to also be used as a trap, depending on how the lens is biased. By applying a sufficiently high voltage to this lens, ions are reflected at the exit and lose enough kinetic energy through collisions with the background gas that they cannot escape over the potential barrier at the hexapole entrance. To trap ions, the exit electrode is biased at 10 V above the dc offset of the ion guide whereas the applied voltage to transmit ions is 5 V below the dc offset.
073107-3
Svendsen et al.
The hexapole guide spans two differential pumping sections and has a 3.5 cm long closed section in the middle pumped only from the ends. This construction is made in order to limit the conductance between the two pumping stages. A 60 l/s turbomolecular pump 共Pfeiffer, TMU 071 P, Nashua, NH兲 backed by a 5 m3 / h rotary vane pump 共Pfeiffer, DUO 5兲 maintains a pressure of 7 ⫻ 10−3 mbar at the entrance side of the ion guide. The terminal end of the hexapole is evacuated by a 500 l/s turbomolecular pump 共Pfeiffer TMU 521 P兲 backed by a 0.9 m3 / h membrane pump 共Pfeiffer, MVP 015–4兲, maintaining a pressure of 1 ⫻ 10−5 mbar. The capillary, ion funnel, and hexapole guide are all mounted on a homemade flange that is bolted to a conflat flange on the ion source vacuum chamber. All electrical connections from the source elements are made to gold-coated pins mounted on this assembly, and inside the vacuum chamber the pins contact homemade spring-loaded connectors, which in turn are connected to electrical feedthroughs on the chamber flanges. With this design, the source components can easily be removed from the chamber as a single unit without disengaging electrical connections, thereby greatly facilitating the cleaning of the source.
D. Gate valve
A homebuilt gate valve separates the source chamber from the rest of the machine. This valve, which is modeled after the design of Pittman and O’Connor,30 allows venting the source region to atmospheric pressure while maintaining the rest of the mass spectrometer under high vacuum conditions. Consequently, the gate valve speeds up the source cleaning procedure, as only the high-pressure regions of the apparatus need to be vented. An advantage of this valve compared to those commercially available is that its blade is electrically isolated and hence can be used as an electrostatic ion lens. Moreover, the assembly is made such that the separation of ion optical elements is kept at a minimum 共2 mm兲. The combination of these two features leads to a high ion transmission through the valve.30
E. Mass selection and octupole ion guide
The next element along the ion path is a quadrupole mass filter with a m/z range of 2–2000 amu 共Extrel, Pittsburgh, PA兲. After mass selection, the ions enter an electrostatic quadrupole ion deflector 共Extrel兲, which deflects the ions 90° in one of two opposite directions, depending on how it is biased. When traveling in one direction, ions are detected by a channel electron multiplier combined with a conversion dynode 共DeTech, 402-A-H, Palmer, MA兲. This detector is mainly used for diagnostic purposes and to obtain a mass spectrum of the sprayed solution. When deflected in the opposite direction, ions enter a 40 cm long octupole ion guide 共Extrel兲, which directs them further downstream. The quadrupole mass filter, quadrupole deflector, and channeltron detector are all housed in the same differential pumping stage. The region is pumped by a 500 l/s turbomo-
Rev. Sci. Instrum. 81, 073107 共2010兲
lecular pump 共Pfeiffer, TMU 521 P兲 backed by a 0.9 m3 / h membrane pump 共Pfeiffer, MVP 015–4兲, resulting in a pressure of 2 ⫻ 10−7 mbar. The octupole guide is installed in a pumping stage of its own evacuated by a 500 l/s turbomolecular pump backed by a 0.9 m3 / h membrane pump. The octupole device can be operated either as a guide or an ion trap. In trapping mode, the octupole entrance and exit lenses are held at potentials somewhat higher than the potential on the octupole axis. To trap the continuous stream of incoming ions, the ion kinetic energy must be dissipated, and this is achieved through collisions with a He buffer gas introduced into the chamber through a leak valve mounted externally on a flange. Under trapping conditions, the pressure is raised to 2 ⫻ 10−4 mbar, which requires lowering the rotation speed of the turbo pump in order to maintain an acceptable backing pressure. If the rf device serves as a guide only, no buffer gas is present, and the pressure is 10−8 mbar. F. Cryogenic 22-pole ion trap
A second quadrupole deflector placed after the octupole guide diverts the ions 90° in one of two opposite directions. Along one direction, the deflector is followed by an off-axis channel electron multiplier equipped with a conversion dynode 共DeTech, 402-A-H兲 for ion detection. In the opposite direction, a stack of five tube lenses, mounted on the quadrupole bender, serve to focus and decelerate the ions before they enter a linear, 22-pole ion trap. The design of the 22-pole rf ion trap is based on those of Gerlich.26,31 It consists of 22 calibrated stainless steel rods of 1 ⫾ 0.001 mm diameter and 41 mm long positioned on a circle with an inscribed diameter of 11 mm. One end of each rod is pressed into a hole in a supporting copper wall, and the other end 共reduced to 0.5 mm兲 is loosely supported by a ceramic sleeve 共OD= 1 mm, ID= 0.5 mm兲, inserted to a similar hole in the opposite supporting wall. This is done in an alternating fashion such that each wall electrically contacts one set of 11 rods. These walls each have an 8 mm hole centered on the trap axis for injection and extraction of ions. Ion trapping in the radial direction is then achieved by applying rf potentials of opposite phase to the two end walls. The trap is equipped with two tube electrodes at each end, and by applying appropriate potentials to the four electrodes the ions are confined along the trap axis. To extract ions from the trap, the potentials on the two electrodes at the exit side are transiently lowered. Five 0.15 mm thick ring electrodes of 13 mm ID spaced 6 mm apart, surround the cylindrical cage of rf electrodes and allow shaping the potential along the trap axis for manipulation of the ion cloud. The 22-pole trap is housed in a copper box, two opposing walls of which support the 22 rods. As these walls conduct the rf potential applied to the rods, they are electrically isolated from the remainder of the box by 1 mm thick sapphire plates. To ensure good thermal contact, two 0.1 mm thick indium foil gaskets are put on each surface of the sapphire plates, which are inserted between the supporting walls and the base plate. The trap assembly is bolted onto a closedcycle helium cryostat 共Sumitomo, SRDK-408D-W71D, Tokyo兲, which enables cooling the trap to 4 K, as measured by
073107-4
Rev. Sci. Instrum. 81, 073107 共2010兲
Svendsen et al.
a silicon diode 共Lakeshore, DT-670B-CU, Westerville, OH兲. An indium gasket is used to make good thermal contact between the trap and the cryostat. Moreover, the lowtemperature parts of the assembly are shielded against roomtemperature blackbody radiation by an aluminum housing connected to the first stage of the cryostat, which is at 70 K. All electrical connections to the trap are precooled to this temperature. To provide continuous control of the trap temperature in the range from 4 to 320 K, a 50 W cartridge heater 共Lakeshore, HTR-50兲 is installed on the top of the copper housing. Ions in the trap are cooled both internally and translationally to low temperature through collisions with cold He buffer gas. The He gas is injected into the trap volume using a pulsed valve 共Parker Hannifin, Series 9, Cleveland, OH兲 connected to the trap via a 25 mm long Teflon tube. The buffer gas is thermalized to the temperature of the trap as the He atoms undergo many collisions with the trap walls and rods before escaping the trap volume through the entrance or exit electrodes. After ejection from the trap, the ions are focused by a lens stack consisting of three tube electrodes. These lenses are mounted on a third quadrupole bender 共Extrel兲 that deflects the ion beam 90° into the last section of the apparatus: the detection chamber. The two quadrupole benders, lens stacks, and the 22pole trap are located in the same differential pumping stage, which is evacuated by a 500 l/s turbomolecular pump 共Pfeiffer TMU 521 P兲 backed by a 5 m3 / h scroll pump 共Edwards, XDS5, Crawley, England兲. The pumps maintain a background pressure below 2 ⫻ 10−9 mbar while the average pressure is ⬃5 ⫻ 10−6 mbar when He buffer gas is pulsed into the trapping volume. G. Mass-selected ion detection
The last differentially pumped vacuum stage houses a quadrupole mass filter 共Extrel兲 and channel electron multiplier combined with a conversion dynode 共DeTech, 402A-H兲 for mass-selective ion detection. The section is pumped by a 260 l/s turbomolecular pump 共Pfeiffer TMU 261 P兲 backed by the same 5 m3 / h scroll pump used for the previous stage. The background pressure is below 2 ⫻ 10−9 mbar but rises to ⬃10−8 mbar when helium is pulsed into the 22-pole trap in the neighboring vacuum stage. H. Electronics
Most of the dc and rf power supplies used for this apparatus are commercially available. Three 20-channel bipolar 共⫾400 VDC兲 power supplies 共Spectrum Solutions, TD 1400, Russellton, PA兲 provide the voltages for all electrostatic ion optical elements such as lenses, quadrupole benders, the ion funnel dc gradient, and the capillary. To eject ions from the traps, the exit lenses of the hexapole, octupole, and the 22pole ion traps are switched between two static outputs from the bipolar supplies. The switching is done in 100 ns with a home-built metal-oxide-semiconductor field-effect transistor switch in push-pull configuration.
The rf signals for the ion funnel are generated by a 0.5 MHz generator 共CGC Instruments, RFG50–10, Chemnitz, Germany兲 and has a variable rf amplitude determined by the magnitude of the voltage delivered to the generator by an external dc power supply 共FuG Elektronik GmbH, NTN 35–35, Rosenheim, Germany兲. The rf signal supplied to the hexapole ion guide is generated in a similar way but at a frequency of 4 MHz. In the case of the hexapole guide, a second external dc voltage, supplied by one of the Spectrum Solutions power supplies, is fed to the rf generator to control the dc offset of the rf signal. The rf power supplies driving the quadrupole mass filters and the octupole guide/trap were purchased from Extrel and have frequencies of 1.2 and 2.1 MHz, respectively. The rf signals for the 22-pole trap are generated by a homemade 4 MHz rf generator,32,33 and the rf amplitude is controlled by an external dc power supply 共FuG Elektronik GmbH, NLN 140M-500兲. The voltages for the conversion dynode and the channel electron multiplier are delivered by three high voltage power supplies 共Schulz-Electronic GmbH, AK0072 and AK0002, Baden-Baden, Germany兲, while the high voltage for the spray needle is supplied by a fourth high voltage supply 共EMCO High Voltage Corporation, Sutter Creek, CA兲. LABVIEW programs were written to control all the electronics. The computer interface uses a National Instruments PXI-chassis equipped with several cards providing digital and analog inputs and outputs, counters, and timers. Analog ⫾10 V outputs of either 13- or 16-bit resolution are used to set the output voltages of the different power supplies, whereas the mass commands for the quadrupole mass filters are controlled uniquely by 16-bit analog outputs. The analog inputs are used for readbacks from the power supplies to monitor their state and to check for faults. For ion detection, the output signal from one of the channeltrons is sent through a preamplifier 共Advanced Research Instruments, COMBO-100, Golden, CO兲 that has both digital and analog output signals. In digital mode, the transistor-transistor logic 共TTL兲 pulses generated by the preamplifier are sent directly to a counter input on one of the PXI cards, where those arriving within a certain time window are recorded. In analog mode, the output of the preamplifier is sent to a digitizer card in the PXI chassis, and the signal within certain a time span is integrated in software. The accurate timings needed for extracting ions from the traps and for synchronizing trap emptying with the ion detection time gate are produced by timer outputs from the PXI chassis. However, all the timings necessary to trigger lasers in laser experiments are generated by an external delay generator 共Berkeley Nucleonics, Model 565, San Rafael, CA兲, which also serves as the master clock for running the apparatus. III. EXPERIMENTAL
The machine was characterized in two ways. To assess the transmitted ion currents and the efficiency of cooling in the 22-pole ion trap, we used protonated tyrosine as a benchmark, since its UV spectrum is sharp, well understood, and provides a good monitor of the internal ion temperature.13 A
073107-5
Rev. Sci. Instrum. 81, 073107 共2010兲
Svendsen et al.
10000
C6H16N+
TyrH+ ion signal [kcounts/s]
Intensity [kcounts/s]
8000 6000 4000
-(H2O + CO)
2000
-NH3
Tyr.Na+
1000 CH6N+
C4H12N+ C3H10N+
500
+
C2H8N
x 100
0
0 50
75
100
125
150
175
200
225
20
250
40
60
100
120
FIG. 5. Mass spectrum of a mixture of primary amines.
FIG. 4. Mass spectrum of the ions produced when spraying a tyrosine solution.
spray solution of tyrosine 共Sigma Aldrich兲 at a concentration of ⬃0.2 mM was prepared in a 1:1 mixture of water and methanol. To characterize the low mass cutoff of the ion funnel, a solution of primary amines was prepared. It contained a sequence of primary amines from methylamine up to hexylamine, with the exception of pentylamine, dissolved in methanol. The mass spectra shown in the following section were acquired by guiding the continuous ion stream through the hexapole ion guide and scanning the first quadrupole mass filter while detecting ions on the channeltron behind the first quadrupole deflector. For experiments involving lasers, the ions of interest were mass-selected in the first quadrupole mass filter and collected in the octupole ion trap. An intense pulse of He buffer gas was injected into the cold 22-pole ion trap, and 1 ms later, a fraction of the ions were transferred from the octupole to the 22-pole trap. Here, they were cooled to ⬃10 K through collisions with the cold buffer gas and interrogated with an UV laser pulse. The ions were then released from the trap, mass-selected in the last quadrupole mass filter and detected by the channeltron detector. The repetition rate of the described cycle was 20 Hz. To obtain an UV excitation spectrum, the signal of any UV-induced fragment ion can be recorded as a function of the wavelength.
80
m/z [amu]
m/z [amu]
parameters, such as the ratio of the electrode spacing to the aperture radius, the rf frequency, and the dc gradient. To investigate the low-mass cutoff of our ion funnel under normal operating conditions, we recorded a mass spectrum of a mixture of primary amines. The spectrum shown in Fig. 5 demonstrates that ions with m/z values of ⬃40 amu are easily transmitted through the funnel, while ions down to m / z = 30 amu are transmitted with sufficient efficiency to allow for detection.
B. UV excitation spectrum
The presented instrument was designed for spectroscopic investigations of cold gas-phase ions, and protonated tyrosine ions are used to demonstrate how UV excitation spectra are measured and to diagnose the internal ion temperature. Figure 6共a兲 shows a mass spectrum of buffer-gas cooled TyrH+ ions that have been trapped in the 22-pole for 12 ms. Clearly, the spectrum is dominated by TyrH+ at m / z = 182 amu, but also a small peak at m / z = 186 amu is present, which is attributed to the formation of TyrH+ · He complexes in the ion trap. The small peak at 165 amu stems from elimination of NH3 from protonated tyrosine, and among all collision-induced dissociation 共CID兲 channels this fragmentation channel has the lowest energy threshold.34 The small peak at 173 amu could not be identified.
IV. RESULTS 400
A. Mass spectra and ion currents
200
counts
A typical mass spectrum obtained when spraying the tyrosine solution is shown in Fig. 4. The most intense peak is located at a m/z value of 182 amu, corresponding to protonated tyrosine, whereas the smaller peaks at 165 and 136 amu indicate the presence of fragment ions resulting from loss of neutral NH3 or H2O + CO, respectively. These types of fragment ions were also observed in studies of collisioninduced dissociation of protonated tyrosine.34 The scale of Fig. 4 is indicative of the continuous ion currents transmitted to the first channeltron detector, however since these count rates are close to the saturated regime of the detector, they should be considered as lower limits. Simulations and experiments have shown that ion funnels have a low-mass cutoff,35–37 which depends on several
(a) Laser off
0
(b) Laser on
108
200 136 147 0 60
80
100
120
140
160
180
m/z [amu]
FIG. 6. Mass spectra of trapped, buffer-gas cooled TyrH+ ions 共m / z = 182 amu兲: 共a兲 UV laser is off; 共b兲 Same as 共a兲, but ions are irradiated by an UV laser pulse resonant with an electronic transition of TyrH+.
073107-6
Rev. Sci. Instrum. 81, 073107 共2010兲
Svendsen et al.
V. COMPARISON WITH THE PREVIOUS INSTRUMENT B00
Counts
1500
0 0
D
1000
A00 500
A01
C00 B
0 1
x100
0 35000
35100
35200
35300
Wavenumber [cm-1]
FIG. 7. UV excitation spectrum of protonated tyrosine recorded by monitoring the number of fragment ions at m / z = 108 amu.
Figure 6共b兲 shows a mass spectrum recorded under the same conditions as those of Fig. 6共a兲, but here the ions were irradiated by an UV laser pulse resonant with an electronic transition of TyrH+ at 35 081 cm−1 prior to extraction of ions from the trap.24 Several fragment ions are clearly produced upon UV excitation, of which only the major one at m / z = 108 amu, which is the protonated side chain, is not observed in CID experiments.34 It should be noted that a previous study on protonated tyrosine by Stearns et al.24 reported the mass of the major fragment to be 107 amu, but in fact the mass resolution of that data was not high enough to distinguish between 107 and 108 amu. The 108 amu fragment was also observed in photofragmentation of roomtemperature protonated tyrosine at 263 nm,38 although it is less intense than the 107 fragment. This may be due to the higher excitation energy in these experiments. As demonstrated in Fig. 7, an UV spectrum of the trapped ions is obtained by recording the number of fragment ions at m / z = 108 amu as a function of the UV wavenumber. The spectrum shows sharp, well-resolved features in agreement with that previously reported by Stearns et al.24 In this previous study, the two intense lowest energy transitions at 35 081 and 35 111 cm−1 were assigned to the band origins of two different conformers: A and B. A hot band transition is clearly evident to the red of each of these electronic band origins: at ⫺40 and −46 cm−1 for A and B, respectively. The intensity of these hot bands implies a vibrational temperature of ⬃10 K for TyrH+.13 This is consistent with the appearance of the TyrH+ · He peak in the mass spectrum of Fig. 6共a兲, which suggests an internal temperature below 20 K.39 The UV spectrum of Fig. 7 was recorded with ⬃105 ions in the cold trap, which is an order of magnitude higher than possible with our previous apparatus.13 Storage of a large number of ions in a cold, 22-pole ion trap can lead to space charge effects that cause heating of the ion ensemble.40 As estimated from the spectrum, the vibrational temperature that we achieve with ⬃105 ions in the trap is still quite low, suggesting that rf heating due to space charge is not yet significant at this loading level.
The apparatus presented here resembles the instrument previously built in our laboratory,13,41 but the new apparatus has some design enhancements which improves the performance. The most important change is in the ion source region where an ion funnel has been incorporated instead of the traditional skimmer setup, which is employed in the older apparatus. According to Smith et al.,42 the ion funnel should improve the transmitted ion current by an order of magnitude, and we observe such an increase in the number of cold, trapped ions: from ⬃104 to ⬃105. This increased signal level allows us to increase the quadrupole mass resolution when necessary at the expense of lowering the ion signal. Other changes to the design include the addition of a gate valve that allows venting the ion source region for cleaning purposes without bringing the entire machine up to atmosphere, and the addition of a third quadrupole bender, which provides better optical access to the cold trap by making the laser axis shorter. We have also used a more powerful closed cycle refrigerator, which cools the 22-pole ion trap to 4 K rather than 6 K as in our previous apparatus. We find that this translates into colder internal ion temperatures by approximately the same increment. Finally, having complete software control of the potentials applied to all ion optical elements provides much greater flexibility and ease for optimizing the ion signals.
VI. SUMMARY
A new tandem mass spectrometer designed for photofragment spectroscopy of cold, biomolecular ions has been described. The instrument comprises a nanospray ion source, a cryogenically cooled 22-pole trap, and a mass-selective detection system. An ion funnel in the ion source region efficiently transmits ions through the high-pressure section of the apparatus leading to large ion currents downstream. Furthermore, the design of the source and the addition of a gate valve greatly facilitate cleaning procedures. The capabilities of the instrument were demonstrated by recording a photofragmentation spectrum and an UV excitation spectrum of protonated tyrosine. The UV spectrum showed sharp, well-resolved features indicative of internally cold ions. Even with a large number of ions 共⬃105兲 stored in the cold trap, the vibrational temperature estimated from the small intensity of hot bands of low-frequency vibrations in the electronic spectrum was found to be ⬃10 K. ACKNOWLEDGMENTS
We gratefully acknowledge financial support for this work from the Fonds National Suisse 共through Grant Nos. 20-120065 and 206021-117416兲 and the Ecole Polytechnique Fédérale de Lausanne 共EPFL兲. U.L. thanks the support of a doctoral fellowship from the “Fonds der Chemischen Industrie” of Germany. A.S. acknowledges financial support from the Danish Research Agency 共Grant No. 272-06-0312兲. The authors would like to thank K. Asmis and M. P. Gorshkov for helpful discussions.
073107-7 1
T. R. Rizzo, Y. D. Park, L. A. Peteanu, and D. H. Levy, J. Chem. Phys. 83, 4819 共1985兲. J. R. Cable, M. J. Tubergen, and D. H. Levy, J. Am. Chem. Soc. 110, 7349 共1988兲. 3 C. K. Teh and M. Sulkes, J. Chem. Phys. 94, 5826 共1991兲. 4 E. Nir, L. Grace, B. Brauer, and M. S. de Vries, J. Am. Chem. Soc. 121, 4896 共1999兲. 5 J. R. Carney and T. S. Zwier, J. Phys. Chem. A 104, 8677 共2000兲. 6 L. C. Snoek, R. T. Kroemer, M. R. Hockridge, and J. P. Simons, Phys. Chem. Chem. Phys. 3, 1819 共2001兲. 7 L. C. Snoek, E. G. Robertson, R. T. Kroemer, and J. P. Simons, Chem. Phys. Lett. 321, 49 共2000兲. 8 T. S. Zwier, J. Phys. Chem. A 105, 8827 共2001兲. 9 J. M. Bakker, L. M. Aleese, G. Meijer, and G. von Helden, Phys. Rev. Lett. 91, 203003 共2003兲. 10 M. S. de Vries and P. Hobza, Annu. Rev. Phys. Chem. 58, 585 共2007兲. 11 H. Fricke, A. Funk, T. Schrader, and M. Gerhards, J. Am. Chem. Soc. 130, 4692 共2008兲. 12 K. Pagel, P. Kupser, F. Bierau, N. C. Polfer, J. D. Steill, J. Oomens, G. Meijer, B. Koksch, and G. von Helden, Int. J. Mass. Spectrom. 283, 161 共2009兲. 13 T. R. Rizzo, J. A. Stearns, and O. V. Boyarkin, Int. Rev. Phys. Chem. 28, 481 共2009兲. 14 J. B. Fenn, M. Mann, C. K. Meng, S. F. Wong, and C. M. Whitehouse, Science 246, 64 共1989兲. 15 F. Hillenkamp, M. Karas, D. Holtkamp, and P. Klusener, Int. J. Mass Spectrom. Ion Process. 69, 265 共1986兲. 16 H. Oh, K. Breuker, S. K. Sze, Y. Ge, B. K. Carpenter, and F. W. McLafferty, Proc. Natl. Acad. Sci. U.S.A. 99, 15863 共2002兲. 17 H. B. Oh, C. Lin, H. Y. Hwang, H. L. Zhai, K. Breuker, V. Zabrouskov, B. K. Carpenter, and F. W. McLafferty, J. Am. Chem. Soc. 127, 4076 共2005兲. 18 S. B. Nielsen, A. Lapierre, J. U. Andersen, U. V. Pedersen, S. Tomita, and L. H. Andersen, Phys. Rev. Lett. 87, 228102 共2001兲. 19 D. Nolting, C. Marian, and R. Weinkauf, Phys. Chem. Chem. Phys. 6, 2633 共2004兲. 20 J. Oomens, N. Polfer, D. T. Moore, L. van der Meer, A. G. Marshall, J. R. Eyler, G. Meijer, and G. von Helden, Phys. Chem. Chem. Phys. 7, 1345 共2005兲. 21 A. Kamariotis, O. V. Boyarkin, S. R. Mercier, R. D. Beck, M. F. Bush, E. 2
Rev. Sci. Instrum. 81, 073107 共2010兲
Svendsen et al.
R. Williams, and T. R. Rizzo, J. Am. Chem. Soc. 128, 905 共2006兲. J. A. Stearns, O. V. Boyarkin, and T. R. Rizzo, J. Am. Chem. Soc. 129, 13820 共2007兲. 23 J. A. Stearns, M. Guidi, O. V. Boyarkin, and T. R. Rizzo, J. Chem. Phys. 127, 154322 共2007兲. 24 J. A. Stearns, S. Mercier, C. Seaiby, M. Guidi, O. V. Boyarkin, and T. R. Rizzo, J. Am. Chem. Soc. 129, 11814 共2007兲. 25 D. Gerlich, Adv. Chem. Phys. 82, 1 共1992兲. 26 D. Gerlich, Phys. Scr. T59, 256 共1995兲. 27 J. A. Stearns, C. Seaiby, O. V. Boyarkin, and T. R. Rizzo, Phys. Chem. Chem. Phys. 11, 125 共2009兲. 28 T. Kim, A. V. Tolmachev, R. Harkewicz, D. C. Prior, G. Anderson, H. R. Udseth, R. D. Smith, T. H. Bailey, S. Rakov, and J. H. Futrell, Anal. Chem. 72, 2247 共2000兲. 29 T. Kim, K. Tang, H. R. Udseth, and R. D. Smith, Anal. Chem. 73, 4162 共2001兲. 30 J. L. Pittman and P. B. O’Connor, J. Am. Soc. Mass Spectrom. 16, 441 共2005兲. 31 D. Gerlich, J. Chem. Soc., Faraday Trans. 89, 2199 共1993兲. 32 R. M. Jones and S. L. Anderson, Rev. Sci. Instrum. 71, 4335 共2000兲. 33 R. M. Jones, D. Gerlich, and S. L. Anderson, Rev. Sci. Instrum. 68, 3357 共1997兲. 34 H. El Aribi, G. Orlova, A. C. Hopkinson, and K. W. M. Siu, J. Phys. Chem. A 108, 3844 共2004兲. 35 J. S. Page, A. V. Tolmachev, K. Tang, and R. D. Smith, J. Mass Spectrom. 40, 1215 共2005兲. 36 J. S. Page, A. V. Tolmachev, K. Tang, and R. D. Smith, J. Am. Soc. Mass Spectrom. 17, 586 共2006兲. 37 A. V. Tolmachev, T. Kim, H. R. Udseth, R. D. Smith, T. H. Bailey, and J. H. Futrell, Int. J. Mass. Spectrom. 203, 31 共2000兲. 38 B. Lucas, M. Barat, J. A. Fayeton, M. Perot, C. Jouvet, G. Gregoire, and S. B. Nielsen, J. Chem. Phys. 128, 164302 共2008兲. 39 S. Mercier, Ph.D. thesis, Ecole Polytechnique Fédérale de Lausanne, 2008. 40 O. Asvany and S. Schlemmer, Int. J. Mass. Spectrom. 279, 147 共2009兲. 41 O. V. Boyarkin, S. R. Mercier, A. Kamariotis, and T. R. Rizzo, J. Am. Chem. Soc. 128, 2816 共2006兲. 42 S. A. Shaffer, K. Tang, G. A. Anderson, D. C. Prior, H. R. Udseth, and R. D. Smith, Rapid Commun. Mass Spectrom. 11, 1813 共1997兲. 22
