Supporting Information for:
Charge Transfer Dissociation (CTD) Mass
Spectrometry of Peptide Cations Using High
Kinetic Energy Helium Cations
William D. Hoffmann1, Glen P. Jackson1,2
1
Department of Forensic and Investigative Science, West Virginia University, Morgantown, WV
26506
2
C. Eugene Bennett Department of Chemistry, West Virginia University, Morgantown, WV
26506
The experimental setup is shown schematically in Figure 1. A custom fabricated rear
cover mounts the saddle field source along the axis of the ETD source ion optics. Figure 1B
shows the routing of signals from the mass spectrometer to the saddle field ion/fast atom source
(VSW/Atomtech, Inc. Macclesfield, UK). The trigger source is taken from pin 14 on the J1
connector of the Digital PCB board. This TTL signal is used to trigger an arbitrary waveform
generator (AFG3252, Tektronix Beaverton, OR). The arbitrary waveform generator produced a 0
- 5 V square wave which was timed to coincide with the fragmentation portion of the scan
function. The alignment of these signals is shown schematically in Figure S1. The square wave
pulse is amplified by a factor of 2000 by a fast high voltage amplifier (ANT 10B10, Matsusada
Precision Inc., Shiga, Japan) in the range of 1-10 kV.
Correspondence to: Glen P. Jackson 1600 University Ave, 208 Oglebay Hall, Morgantown, WV
26506, Email: glen.jackson@mail.wvu.edu, Tel: 304-293-9236
1
Figure S1. Schematic representation of the alignment of the analytical RF waveform
from the mass spectrometer and the high voltage DC output from the high voltage
amplifier.
In the experiments presented here, ultra high purity helium (PHEN30050, 99.999%,
Matheson Gas, Basking Ridge NJ) was additionally purified through a GC triple filter (22020,
Restek, Bellefonte, PA) to remove any residual contamination. The purified helium was
introduced through a precision leak valve and the pressure was set from 6-8 x10-5 Torr, as
monitored by the on-board ion gauge in the differentially-pumped ETD source region.
Experiments were conducted after pressures had sufficient time to equilibrate, which was usually
~5 minutes.
Substance P (acetate salt, S6883) was acquired from Sigma Aldrich (St. Louis, MO).
LC/MS grade methanol and glacial acetic acid were obtained from Fisher Scientific (Waltham,
MA). Water was obtained from an in-house Milli-Q purification system with > 18 MΩ salt
content. All reagents were used without further purification. 60 µM solutions of substance P
2
were made in 1:1 (V:V) solutions of MeOH and H2O. All solutions were acidified to 1% with
HOAc.
All spectra were collected in positive mode with an ESI voltage of 4.5 kV, an ion transfer
capillary temperature of 250 °C, and a heated ESI source temperature of 60 °C. Solutions were
directly infused from the on-board syringe pump at a rate of 5 µL/min. A typical experimental
run consisted of collecting the full mass spectrum for a period of 30 seconds, then 30 seconds of
the isolated precursor, 2.5 minutes of CTD reaction (with the helium ion source pulsed on for ~1
s per scan) and 2.5 minutes with the electrospray source off and He+ source on, as background.
Fragmentation time was set to 1 s, and the saddle field ion/fast source was on for 980 ms of this
time. In all the spectra presented the saddle field source was operated at 6 kV. All spectra
presented are time-averaged over the 2.5-minute collection interval and subsequently
background subtracted. Initial experiments used an isolation window of m/z 4 to ensure adequate
precursor signal. The initial spectrum of substance P can be seen in Figure S2.
With the expanded isotopic envelope, the spectrum is very similar to that seen in Figure
2. A dominant product peak representing the [M+H]2+ radical species is observed along with a
near complete series of a-ions. The greater precursor intensity in Figure S2 relative to 2 (in the
manuscript) provides better signal to noise and more peak identifications, but the expanded
isotopic envelope somewhat obscures the identification of a+1 ions.
3
Figure S3. Appearance potential experiment using He+ charge transfer reaction with
Figure S2.
He+ CTDThe
spectrum
of substance
P with
anblank
expanded
isolationcharge
window of 4.0
neutral
chloroform.
first 3 minutes
represent
the
(background)
m/z. At least
one ion
from eachwas
major
fragmentation
identified.
isotopic
transfer
spectrum.
Chloroform
introduced
at ~3.2pathway
minutes.is The
bottomThe
panel
shows
envelop
afforded
by the expanded
window
helped
the identification
of
the
extracted
ion chromatogram
forisolation
m/z 35 and
37 while
theconfirm
top panels
show the timethe doublymass
charged
ions.for the blank and chloroform respectively. When chloroform is
averaged
spectra
introduced, signals representing the expected isotopic distribution of chloride ions are
produced.
To determine the amount of energy available for reaction, and confirm the presence of
low mass He+ ions, appearance potential experiments were conducted. Chloroform was
introduced by placing a small beaker of chloroform near the standard API capillary entrance. The
mass spectrometer was operated in MS/MS mode with a ‘ghost’ precursor of m/z 100, as there
was no sample was being infused or ionized in the electrospray source. Mass spectra were
collected using the low mass range mode from m/z 15 – 150. The extracted ion chromatogram
for m/z 35 and 37 is shown in the bottom panel of Figure S3. For the first 3 minutes show charge
transfer background signals, before chloroform was introduced. Between 3.2 and 6.4 minutes,
chloroform vapors were introduced to the high-pressure-trapping region via the API inlet and
several charge transfer reaction products are observed. The time-averaged mass spectra are seen
4
in the panels above each time domain. Clear signals representing the chlorine ions at m/z 34.83
and m/z 36.91 are seen when chloroform is introduced, but are noticeably absent from the
background. Previous work has shown the appearance potential of chlorine cations from
chloroform to be 22.0 ± 0.3 eV [1]. The metastable states of helium lie at ~20 eV [2], so the
presence of chlorine cations is good experimental evidence that the species emitting from the
saddle field source are helium cations. Beyond the appearance potential, the intensity of halogen
ions from the dissociation of halogenated hydrocarbons increases approximately linearly with
increasing electron energy [3, 4]. Based on a linear fit of Hobrock and Kiser’s EI data [1]—from
the appearance potential at 22 eV to the common EI energy at 70 eV—we calculate the average
energy of our system to be ~39 eV. Using the same method based on data for x-ray induced
dissociation of trichloromethane [5], the average CTD energy is approximately 33 eV. The
results of these two calculations give confidence that the typical activation energy using 6 keV
He+ ions lies between 22 and 40 eV, and is more likely between 30-40 eV. The maximum energy
that can be transferred from the lab frame to the center of mass frame for trichloromethane and 6
keV He+ ions is approximately 197 eV, but 100% conversion is very unlikely [6]. Conversion
efficiencies on the order of 1-10% are much more common, which would provide between 2-20
eV to the activation energy, in agreement with the observed results. When the mass of the
reacting partner (He+) is much less than the mass of the target ion (substance P), inelastic energy
transfer becomes a minor pathway and a much smaller fraction of the available center-of-mass
energy will be converted to internal energy [7].
The biggest factor limiting our acquisition rate is the effective flux of helium ions
through the trap. The saddle-field source has no onboard ion focusing elements, so the beam
exiting the source is divergent (5 degrees, according to the manufacturer) all the way through the
5
linear ion traps. This decreases the degree of overlap or effective flux between the trapped bioions and the transmitting helium ions and necessitates 1-s-long reaction times to achieve
reasonable signal-to-noise levels. At present, the effective ion cloud overlap between the He+
reagent ions and the stored precursor ions in the LIT is approximately 8%. The overlap estimate
is based on the approximate precursor ion cloud volume in the LIT (~100 mm3) [8] and the
divergent ion beam volume (~1200 mm3)—as calculated using the 5 degree divergent beam
passing through the 2 mm diameter center ion lens over the length of the high pressure LIT.
Based on this 8% overlap in ion clouds, the He+ flux passing through the high-pressure LIT (~10
nA), and the reduction in precursor abundance following CTD (2 x 105 counts), we can estimate
a CTD reaction efficiency of ~0.004% per helium ion. The measured CTD efficiencies at 1-s
reaction time with the current ion flux of ~10 nA through the high pressure LIT (where the
reactions take place) are on the order of 4%.
References
1. Hobrock, D. L.; Kiser, R. W. Electron impact studies of some trihalomethanes:
Trichloromethane, dichlorofluoromethane, chlorodifluoromethane, and trifluoromethane1. J.
Phys. Chem. 68, 575-579 (1964)
2. Siska, P. E. Molecular-beam studies of penning ionization. Rev. Mod. Phys. 65, 337-412
(1993)
3. Fiegele, T.; Hanel, G.; Torres, I.; Lezius, M.; Märk, T. D. Threshold electron impact
ionization of carbon tetrafluoride, trifluoromethane, methane and propane. J. Phys. B: At., Mol.
Opt. Phys. 33, 4263 (2000)
6
4. Torres, I.; Martínez, R.; Rayo, M. N. S.; Castaño, F. Electron impact dissociative ionization
of the ch2f2 molecule: Cross sections, appearance potentials, nascent kinetic energy distributions
and dissociation pathways. J. Phys. B: At., Mol. Opt. Phys. 33, 3615 (2000)
5. Lago, A. F.; Santos, A. C. F.; Souza, G. G. B. d. Mass spectrometry study of the
fragmentation of valence and core-shell (cl 2p) excited chcl3 and cdcl3 molecules. J. Chem.
Phys. 120, 9547-9555 (2004)
6. McLuckey, S. A. Principles of collisional activation in analytical mass spectrometry. J. Am.
Soc. Mass. Spectrom. 3, 599-614 (1992)
7. Goeringer, D. E.; McLuckey, S. A. Evolution of ion internal energy during collisional
excitation in the paul ion trap: A stochastic approach. J. Chem. Phys. 104, 2214-2221 (1996)
8. Schwartz, J. C.; Senko, M. W.; Syka, J. E. P. A two-dimensional quadrupole ion trap mass
spectrometer. J. Am. Soc. Mass. Spectrom. 13, 659-669 (2002)
7