Aerosol Science and Technology, 39:637–658, 2005
Copyright
c
American Association for Aerosol Research
ISSN: 0278-6826 print / 1521-7388 online
DOI: 10.1080/02786820500182040
A New Time-of-Flight Aerosol Mass Spectrometer
(TOF-AMS)—Instrument Description and First
Field Deployment
Frank Drewnick,
1
Silke S. Hings,
1
Peter DeCarlo,
2,3
John T. Jayne,
5
Marc Gonin,
6
Katrin Fuhrer,
6
Silke Weimer,
7,8
Jose L. Jimenez,
3,4
Kenneth L. Demerjian,
7
Stephan Borrmann,
1,9
and Douglas R. Worsnop
5
1
Max Planck Institute for Chemistry, Cloud Physics and Chemistry Department, Mainz, Germany
2
Program in Atmospheric and Oceanic Sciences, University of Colorado, Boulder, Colorado, USA
3
CIRES, University of Colorado, Boulder, Colorado, USA
4
Department of Chemistry & Biochemistry, Boulder, Colorado, USA
5
Aerodyne Research, Inc., Billerica, Massachusetts, USA
6
TOFWERK AG, Thun, Switzerland
7
State University of New York, Atmospheric Sciences Research Center, Albany, New York, USA
8
Now at: EMPA, D
¨
ubendorf, Switzerland, and Paul Scherrer Institute, Villigen, Switzerland
9
Institute for Atmospheric Physics, University of Mainz, Mainz, Germany
We report the development and first field deployment of a
new version of the Aerosol Mass Spectrometer (AMS), which
is capable of measuring non-refractory aerosol mass concentra-
tions, chemically speciated mass distributions and single particle
information. The instrument was constructed by interfacing the
well-characterized Aerodyne AMS vacuum system, particle focus-
ing, sizing, and evaporation/ionization components, with a com-
pact TOFWERK orthogonal acceleration reflectron time-of-flight
mass spectrometer. In this time-of-flight aerosol mass spectrome-
ter (TOF-AMS) aerosol particles are focused by an aerodynamic
lens assembly as a narrow beam into the vacuum chamber. Non-
Received 30 December 2004; accepted 13 May 2005.
The PMTACS-NY 2004 deployment of the TOF-AMS was sup-
ported in part by the New York State Energy Research and Devel-
opment Authority (NYSERDA), contract #4918ERTERE-S99 and the
U.S. Environmental Protection Agency (EPA), contract #R828060010.
Although the research described in this article has been funded in part
by the U.S. EPA, it has not been subjected to the Agency’s peer and
policy review and therefore does not necessarily reflect the views of the
Agency, and no official endorsement should be inferred. We thank Phil
Mortimer from Aerodyne Research, Inc., James Schwab from ASRC
at SUNY Albany, and Queens College for logistical support during
the PMTACS campaign. S. Hings thanks International Max Planck Re-
search School for funding for her participation in this work. J.L. Jimenez
and P. DeCarlo thank the U.S. Dept. of Energy (DE-FG02-03ER83599)
and the Office of Naval Research (grant N00244-04-P-0425) for fund-
ing participation in this work.
Address correspondence to Frank Drewnick, Max Planck In-
stitute for Chemistry, Cloud Physics and Chemistry Department,
J. J. Becherweg 27, D-55128 Mainz, Germany. E-mail: drewnick@
mpch-mainz.mpg.de
refractory particle components flash-vaporize after impaction onto
the vaporizer and are ionized by electron impact. The ions are con-
tinuously guided into the source region of the time-of-flight mass
spectrometer, where ions are extracted into the TOF section at a
repetition rate of 83.3 kHz. Each extraction generates a complete
mass spectrum, which is processed by a fast (sampling rate 1 Gs/s)
data acquisition board and a PC. Particle size information is ob-
tained by chopping the particle beam followed by time-resolved
detection of the particle evaporation events. Due to the capabil-
ity of the time-of-flight mass spectrometer of measuring complete
mass spectra for every extraction, complete single particle mass
spectra can be collected. This mode provides quantitative informa-
tion on single particle composition. The TOF-AMS allows a direct
measurement of internal and external mixture of non-refractory
particle components as well as sensitive ensemble average parti-
cle composition and chemically resolved size distribution measure-
ments. Here we describe for the first time the TOF-AMS and its
operation as well as results from its first field deployment during the
PM
2.5
Technology Assessment and Characterization Study—New
York (PMTACS-NY) Winter Intensive in January 2004 in Queens,
New York. These results show the capability of the TOF-AMS to
measure quantitative aerosol composition and chemically resolved
size distributions of the ambient aerosol. In addition it is shown that
the single particle information collected with the instrument gives
direct information about internal and external mixture of particle
components.
INTRODUCTION
Aerosol particles participate in many physical and chemical
processes in the atmosphere such as climate forcing through
direct and indirect effects, heterogeneous chemistry, or visibility
reduction (Andreae et al. 1997; Seinfeld and Pandis 1998; IPCC
637
638 F. DREWNICK ET AL.
2001; Warneck 1999; Watson 2002). In addition due to their
impact on human health, aerosols are increasingly recognized
as a major concern in urban air quality (Pope et al. 2002; Samet
et al. 2000; Wichmann et al. 2000). However, many questions
about particle formation, transport, transformation and impact
remain largely unanswered, in good part due to limitations of
available instrumentation.
In the last decade intensive research and development has
resulted in significant improvements in aerosol measurement
capabilities and a wealth of new fast and sensitive aerosol mea-
surement instruments. Among these, aerosol mass spectrometry
has proved to be a versatile and highly sensitive method for
aerosol analysis (McMurry 2000; Johnston 2000).
A milestone in the development of aerosol mass spectrome-
try has been the demonstration and development of online laser
vaporization/ionization aerosol mass spectrometers (McKeown
et al. 1991; Gard et al. 1997; Murphy and Thomson 1997;
Carson et al. 1997a,b). These instruments provide rapid informa-
tion on the size and chemical composition of single aerosol par-
ticles. However, due to the single-step vaporization/ionization
process these instruments suffer from biases in the particle-size
and chemical composition representation (Reilly et al. 2000;
Gross et al. 2000; Allen et al. 2000; Kane and Johnston 2000),
limiting their ability of providing quantitative analysis. In ad-
dition the optical detection techniques used to trigger the ab-
lation/ionization laser limit the analysis of particles to particle
sizes larger than 0.2 µminmost of these instruments. Oper-
ation of single-step laser ablation instruments with very high
laser power fluences has been shown to quantitatively convert
particles into atomic ions, leading to quantitative atomic com-
position analysis (Reents and Ge 2000; Mahadevan et al. 2002).
However, it has recently been shown that this approach does not
work for all particle compositions (Wang et al. 2004), and in any
case only elemental, not molecular information is obtained.
In recent years a number of aerosol mass spectrometers were
developed that allow quantitative analysis of aerosol composi-
tion by separating the particle evaporation and ionization pro-
cesses. In some of these instruments particles were evaporated
with an intense infrared laser pulse and subsequently ionized
by a second UV laser pulse (Zelenyuk et al. 1999; Morrical
et al. 1998). Other instruments evaporate collected particles
by heating the collection substrate, followed by electron im-
pact or chemical ionization (Tobias and Ziemann 1999; Smith
et al. 2004). The Aerosol Mass Spectrometer (AMS) devel-
oped by Aerodyne Research, Inc., flash-vaporizes particles, fo-
cused by means of an aerodynamic lens system onto a vaporizer
and analyses the evolving vapors by electron impact ionization
and quadrupole mass spectrometry (Jayne et al. 2000; Jimenez
et al. 2003a).
The AMS has been shown in multiple field and laboratory
experiments to have the potential for quantitatively measuring
chemical composition as well as chemically resolved size dis-
tributions (e.g., Drewnick et al. 2003; Schneider et al. 2004;
Jimenez et al. 2003a; Hogrefe et al. 2004; Allan et al. 2004;
Zhang et al. 2004a). In order to further improve sensitivity
and time resolution of the AMS and to extend its capabili-
ties to determine single particle information the Time-of-Flight
Aerosol Mass Spectrometer (TOF-AMS) was developed. This
instrument combines the demonstrated features of the AMS
particle collection, sizing, and evaporation/ionization technol-
ogy with state-of-the-art, time-of-flight mass spectrometry. Here
we report for the first time on the operation of the TOF-AMS
and present results from its first field deployment during the
PMTACS-NY 2004 campaign in Queens, New York. A system-
atic characterization study of the TOF-AMS and its capabilities
is currently in progress and will be the subject of a forthcoming
publication.
The PM
2.5
Technology Assessment and Characterization
Study—New York (PMTACS-NY) was one of several US EPA
“Supersites,” intended to provide enhanced measurement data
on chemical and physical properties of particulate matter and
its associated precursors. One of the primary objectives of this
study is to test and evaluate recently developed aerosol mea-
surement technologies like the TOF-AMS. The PMTACS-NY
2004 study was a winter study, performed at the same location
as the PMTACS-NY 2001 summer campaign (e.g., Drewnick
et al. 2004a,b) on the campus of Queens College in Queens,
New York.
INSTRUMENT DESCRIPTION
The Time-of-Flight Aerosol Mass Spectrometer (TOF-AMS)
is a combination of the well-characterized quadrupole mass
spectrometer-based Aerodyne AMS (from here on called “Q-
AMS”) aerosol sampling, sizing and evaporation/ionization
technology (e.g., Jayne et al. 2000; Jimenez et al. 2003a) and a
compact TOFWERK orthogonal extraction time-of-flight mass
spectrometer (TOF-MS, e.g., Steiner et al. 2001). An instrument
schematic of the TOF-AMS is given in Figure 1.
The TOF-AMS in its current version is mounted in a sin-
gle mobile rack with the vacuum system, the mass spectrometer
and the whole electronics, including data acquisition system in-
tegrated. The rack dimensions are 104 × 61 × 124 cm (rack vol-
ume ∼0.79 m
3
) and the TOF-AMS weighs about 200 kg. Under
sampling conditions the instrument has a power consumption of
approximately 600 W with about 1/3 of this power being used
by the data acquisition computer and the instrument electronics,
the remaining being consumed by the vacuum system (5 turbo
pumps and the single backing pump). During the PMTACS-NY
campaign the instrument was operated in a preliminary set-up
with the vacuum system and the mass spectrometer separated
from the electronics rack.
Here we describe the components of the TOF-AMS that are
different from those in the Q-AMS, and their most important
characteristics. Further information is given in Jayne et al. (2000)
about the AMS chamber and in Steiner et al. (2001) about the
time-of-flight mass spectrometer.
The TOF-AMS vacuum system consists of five individual,
differentially pumped chambers: the aerosol sampling chamber,
NEW TIME-OF-FLIGHT AEROSOL MASS SPECTROMETER 639
FIG. 1. Schematic of the Time-of-Flight Aerosol Mass Spectrometer (TOF-AMS). Aerosol is introduced into the instrument through an aerodynamic lens
focusing the particles through a skimmer and an orifice onto the vaporizer. Particle vapor is ionized and the ions are guided into the TOF-MS, which generates
mass spectra at ∼83.3 kHz repetition rate. For particle size measurement the particle beam is chopped with a mechanical chopper and the detection is synchronized
with the chopper opening time.
the particle-sizing chamber, the particle evaporation and ioniza-
tion chamber (two chambers), and the TOF-MS chamber. The
aerosol is sampled at a flow rate of approximately 1.4 cm
3
/s
into the aerosol-sampling chamber through a critical orifice
of 100 µmIDand an aerodynamic lens system (Zhang et al.
2002, 2004b) that focuses particles in the size range ∼50 nm to
∼600 nm into a narrow beam.
Behind the critical orifice the pressure drops to about 1.8 hPa,
depending on the mass flow rate into the inlet. This pressure is
monitored with a sensitive capacitance pressure gauge (MKS
Baratron, 0–10 Torr (0–13.33 hPa)), which provides a contin-
uous inlet flow rate measurement via external calibration with
abubble flow meter. The lens focuses particles in the above
mentioned size range with almost 100% efficiency into a nar-
row beam of ∼100 µm diameter a few cm behind the lens exit
(Heberlein et al. 2001). Smaller and larger particles are also
transmitted albeit with reduced efficiency.
The particle beam leaves the aerosol-sampling chamber
througha1mmIDchannel skimmer, which skims off most
of the air entering the inlet while transmitting the particle beam
into the particle-sizing chamber. The air is removed from the
aerosol-sampling chamber by a 280 l/s turbo molecular pump
(Varian V-301 NAV), backed by a diaphragm pump (Vacuubrand
MD1-Vario). The particle-sizing chamber is pumped by a 70 l/s
turbo molecular pump (Varian V-70LP). This pump as well as
the other downstream turbo pumps is backed by the inlet turbo
molecular pump so that only a single roughing pump is needed
for the whole system.
In the expansion of the air into the vacuum chamber at the
final nozzle of the aerodynamic lens the particles are acceler-
ated to a terminal velocity that depends on their aerodynamic
size (see next section), which is retained in the vacuum cham-
ber due to the lack of horizontal forces acting on the particles
under the high vacuum conditions. This effect is used to deter-
mine particle size by measuring particle velocity in the particle-
sizing chamber. For this purpose a mechanical chopper wheel is
mounted at the front end of this chamber with two radial slits,
each cutting through 0.5% of the chopper circumference, so that
the effective opening duty cycle is 1%. Particle size informa-
tion is obtained by chopping the particle beam and collecting
640 F. DREWNICK ET AL.
complete mass spectra as a function of particle time-of-flight
with the mass spectrometer synchronized with the chopper rota-
tion. The recording of complete mass spectra at each time step in
a chopper cycle is a major difference with the Q-AMS, in which
only one m/z is scanned for a given chopper cycle. This improve-
ment in measurement duty cycle results in a large improvement
in particle sampling statistics with respect to the Q-AMS, as
discussed below. Complete non-refractory particle composition
information (all m/z’s) without any sizing is collected with the
chopper completely removed from the particle beam in order to
maximize particle transmission, and alternatively blocking the
whole beam in order to record the background signals for sub-
traction. Again in this MS mode of the TOF-AMS a complete
mass spectrum is acquired every 12 µs, while in the Q-AMS
the mass spectrum is scanned over a mass range of m/z 1–300
within 300 ms, with only one m/z being detected at a given time.
Again this higher duty cycle results in greatly improved particle
sampling statistics with respect to the Q-AMS.
After traveling through the particle-sizing chamber, the par-
ticles enter the particle evaporation and ionization chamber
through a 3.8 mm orifice. This chamber consists of an outer
volume which is pumped by a 70 l/s turbo molecular pump
(Varian V-70LP), and the inner chamber which is pumped to a
pressure of approximately 2 × 10
−5
Pa by a 280 l/s turbo molec-
ular pump (Varian V-301 NAV). The particles impact onto the
vaporizer, located at the downstream end inside a compact cross
beam electron impact ion source. The particle vaporizer (as well
as the entire ionizer) used on the TOF-AMS is identical to that
used on the Q-AMS systems. The vaporizer has a diameter of
3.8 mm and is custom-built from porous tungsten, ∼20% void
volume with pore sizes of ∼100–200 µm. The front section of
the vaporizer where the particles impact has an inverted cone
shape, a 60-degree included angle. Both the cone shape and the
porosity are to enhance capture from bouncing particles. The
vaporizer is brazed onto a molybdenum heater body (containing
an embedded resistive wire potted in ceramic) and is therefore
heated by conduction. The vaporizer temperature is measured
with a micro thermocouple that is fixed on the front OD of the
vaporizer and can be adjusted in a range from about 250
◦
C (lim-
ited by radiative heating from the electron emission filament) up
to ∼1000
◦
C. Typical operating power of 2 W provides ∼600
◦
C
near the vaporizer surface.
Akey to this design (identical to the Q-AMS system) is
mounting the vaporizer in the center of the ionizer such that es-
sentially every molecule that leaves the vaporizer passes through
the ionization volume that is imaged into the mass spectrometer.
It is necessary to apply a voltage bias to the vaporizer to “re-
tune” the distorted electric field caused by placing the vaporizer
inside the ionizer. This voltage is typically within several volts of
the ion reference voltage. The ionizer used in the TOF-AMS is a
commercial cross-beam ionizer (Inficon/Balzers) with reduced
volume, compared to the standard cross-beam ionizer. The non-
refractory aerosol components flash-evaporate quickly (within
50–100 µs) after impaction of the particles on the vaporizer. The
resulting vapor molecules are ionized by 70 eV electrons emit-
ted from a tungsten filament located to the side of the ion source.
Ions are extracted from the ion source via a lens at a potential of
about –100V and focused into a beam with an Einzel lens, de-
celerated to about 50 eV. Instead of being directly injected into
the Q-AMS, in the TOF-AMS ions are transferred 96 mm to the
orthogonal TOF extractor through electrostatic lenses, which are
designed to keep the ion loss as small as possible. The electro-
static focusing should also keep mass discrimination effects at
minimum as in electrostatic fields the trajectories of particles
with equal E/q are identical. This means that those components
of the initial ion velocities which have a mass dependence (es-
sentially the plume velocity) become negligible compared to the
100 eV extraction energy. The main contribution of the initial
ion energy variability stems from the position of ionization in
the relatively strong extraction field within the ionizer, which
should not be mass dependent.
The ions enter the TOF-MS through a hole of approximately
6mmdiameter. The TOF-MS is housed in a compact vacuum
chamber of 265 × 155 × 75 mm, pumped by another 70 l/s turbo
molecular pump (Varian V-70LP). The ions are guided and col-
limated into the ion extractor. The open area of the extractor is
46-mm long and matches the active area of the MCP detector.
The ions drift through the TOF extractor at 50 eV before they
are orthogonally extracted into the TOF section by a pulsed high
voltage. After each extraction pulse the ions have to refill the ex-
tractor. This “fill up time” is m/z dependent and determines the
m/z dependent ion duty cycle. Lighter ions have a short fill up
time after which the extractor is “overfilled” and ions are lost.
This leads to a lower duty cycle compared to heavier ions whose
fill up time is comparable to the TOF extraction period. Hence,
the TOF-MS ion duty cycle increases with square root of mass
and reaches about 30% for the largest ion, which’s TOF just
matches the extraction period. Typically the extraction period is
12 µs, generating 83,300 complete mass spectra per second. The
ion drift energy is roughly 2 kV. The TOF-MS is equipped with
atwo-stage gridded ion reflector, resulting in an effective flight
path of 430 mm. After post acceleration the ions are collected by
a40mmchevron stack MCP detector (modified MCP 40/12/8
D EDR 46:1 CZ TC set, Burle Technologies, Inc., Sturbridge,
MA). All voltages of the TOF-MS, the ionizer, and the filament
current are generated using a custom-built power supply. The
time-of-flight mass spectrometer is described in more detail in
Steiner et al. (2001). The MCP output signal is detected in two
channels of a high-speed (1 Gs/s) analog-to-digital conversion
data acquisition card (AP240, Acqiris, Geneva, Switzerland) in
parallel. One channel records the mass spectral signal with an
amplification of 11 (Amplifier Model ACA-2-21-N, Becker &
Hickel GmbH, Germany), and the other without any amplifi-
cation, for extension of the dynamic range. Data collected by
the data acquisition card are transferred to a personal computer
and stored to disc. For high-duty cycle spectrum acquisition raw
mass spectra collected on the ADC card are averaged on the card
in real time before transfer to the PC every few seconds.
NEW TIME-OF-FLIGHT AEROSOL MASS SPECTROMETER 641
INSTRUMENT OPERATION DURING PMTACS-NY
The TOF-AMS records data in three different modes of op-
eration: The MS (Mass Spectrum) mode, the P-TOF (Particle
Time-of-Flight) mode, and the SP-TOF (Single Particle Time-
of-Flight) mode. The data acquisition software currently under
development will allow the instrument to shift back and forth
between the different operation modes every few seconds. How-
ever, the instrument was used with a first version of the data
acquisition software during the PMTACS-NY 2004 campaign,
which did not allow the rapid programmable alternation of all
three operating modes.
The MS or (averaged) Mass Spectrum mode is used to collect
averaged mass spectra of the non-refractory aerosol components
for an ensemble of particles. In order to maximize the duty cy-
cle in this mode the particle beam chopper is completely moved
out of the beam, enabling a maximum number of the particles
that are entering the instrument to impact on the vaporizer, and
then closed to allow the recording of background mass spectra.
The vapors evolving from the particles are continuously ion-
ized by 70 eV electron impact and the positive ions formed are
continuously transferred into the extractor of the TOF-MS.
During PMTACS-NY the orthogonal extraction voltage was
pulsed at 83.3 kHz, generating a complete mass spectrum every
12 µs. To keep up with this enormous data stream the mass spec-
tra were averaged in the memory of the data acquisition board in
real time. 312,000 spectra were averaged within 3.7 s. Then the
particle beam chopper was moved into the beam to completely
block it. Now the gas phase background spectrum was measured
for another 312,000 spectra or 3.7 s. Every 3.7 s the averaged
mass spectra were transferred to the PC RAM memory under
software control. The spectra were further processed and aver-
aged and were saved to a file on the computer hard drive every
five min. The average aerosol and gas-phase “difference” spec-
trum was calculated from the average mass spectrum with the
particle beam not blocked by the chopper (“beam open” spec-
trum) minus the average background mass spectrum (“beam
closed” spectrum). From these high-resolution mass spectra,
where the MCP signal was recorded with 1 ns time resolution,
unit mass resolution spectra (one signal intensity per m/z) were
also calculated. They were calculated from the raw mass spectra
by integration of the signal area at every m/z, which is propor-
tional to the ion current. The m/z ranges of interest were defined
by an m/z calibration, performed before each start of the data
acquisition. This ion mass calibration uses the flight times and
known m/z’s of two easy-to-identify background peaks. From
this information a calibration curve, relating ion TOF vs. m/z is
calculated for the whole mass spectrum ranging from m/z 4to
206 during the NYC campaign. In Figure 2, the raw (beam open,
beam closed and difference spectrum) and unit-resolution (dif-
ference) mass spectra are shown for one 5-min averaging inter-
val. The signal intensity is given in ions per spectrum. Since the
unit-resolution spectrum is the integral over the peak, the number
of ions per spectrum for each m/z is higher than the maximum
number of ions per spectrum for each individual channel of the
high-resolution peak. Note, that signals at m/z that are associ-
ated with organic species are much lower than at those associated
with inorganic species even though organics represents approxi-
mately 50% of the total non-refractory aerosol mass (see Figure
5). This is due to the fact that fragments of organic species are
found at a very large number of m/z, while the inorganic species
fragment only in a small number of major ions.
The P-TOF or (averaged) Particle Time-of-Flight mode is
used to collect averaged size distribution data for all non-
refractory aerosol components for an ensemble of particles. In
this mode the particle beam is chopped with the mechanical
chopper wheel at about 125 Hz. The chopper transmits particles
for 1% of the time (∼80 µs) and blocks them for the rest of
the time (∼7.92 ms). The open or closed position of the chop-
per is monitored with an LED whose reflection on the chopper
is measured by a photodiode, defining the opening time of the
chopper when particles are entering the particle sizing chamber
at the chopper position. The arrival times of the particles at the
vaporizer are determined by time-resolved detection of the mass
spectra for each chopper cycle. This is possible since the time
scale for evaporation, ionization and mass spectrometric analy-
sis is short (∼50 µs) compared to the flight time of the particles
through the particle-sizing chamber (∼3 ms).
During the expansion of the air into the vacuum the parti-
cles are accelerated to a velocity that depends on the inertia and
aerodynamic drag properties of the particles: the smaller and
less massive the particles are, the better they can follow the gas
molecules in the expansion and the faster their terminal veloc-
ity. To a first approximation for spheres the accelerating force
during the expansion (approximated by Stokes’ drag force) is
proportional to d
2
p
and the particle mass is proportional to d
3
p
,
the acceleration of the particles is proportional to d
−1
p
resulting
in a particle velocity proportional to d
−1/2
p
. The resulting equiv-
alent diameter is known as the vacuum aerodynamic diameter
(d
va
) (Jimenez et al. 2003b; DeCarlo et al. 2004). The particle
velocity to d
va
dependency was determined in the field using
monodisperse ammonium nitrate particles of known mobility
diameters (d
m
= 50 – 500 nm), generated with a Collison nebu-
lizer, diffusion dried, and size-selected with a DMA. The DMA
was calibrated with polystyrene latex spheres (PSLs, Duke Sci-
entific, Palo Alto, CA). This particle time-of-flight calibration
enables the transformation of particle flight times measured for
ambient particles into d
va
as described in the following section.
At 125 Hz chopping frequency each chopper cycle (chopper
opening and particle time-of-flight measurement cycle) is 8 ms
long. The TOF mass spectrometer was pulsed continuously at
83.3 kHz, producing a mass spectrum every 12 µs. To save data
acquisition on-board memory, during the first 100 µsofeach
chopper cycle no data were collected from the mass spectrom-
eter (“data delay”), since no particles or gases can fly down the
chamber at the speeds needed to arrive during that delay. Af-
ter the data delay 520 mass spectra were recorded (spaced by
12 µs) in every chopper cycle, covering 6.24 ms of the cycle. Due
to the limited on-board memory always two consecutive mass
