The ToF-ACSM: a portable aerosol chemical speciation monitor with TOFMS detection

17 Pages·2013·1.68 MB·English

Atmos.Meas.Tech.,6,3225–3241,2013 Atmospheric O
p
www.atmos-meas-tech.net/6/3225/2013/ e
n
doi:10.5194/amt-6-3225-2013 Measurement A
c
c
©Author(s)2013.CCAttribution3.0License. Techniquese
s
s
The ToF-ACSM: a portable aerosol chemical speciation monitor
with TOFMS detection
R.Fröhlich1,M.J.Cubison2,J.G.Slowik1,N.Bukowiecki1,A.S.H.Prévôt1,U.Baltensperger1,J.Schneider3,
J.R.Kimmel2,4,M.Gonin2,U.Rohner2,D.R.Worsnop4,andJ.T.Jayne4
1LaboratoryofAtmosphericChemistry,PaulScherrerInstitute,Villigen,Switzerland
2TofwerkAG,Thun,Switzerland
3MaxPlanckInstituteforChemistry,Mainz,Germany
4AerodyneResearch,Inc.,Billerica,Massachusetts,USA
Correspondenceto:A.S.H.Prévôt([email protected])
Received:4July2013–PublishedinAtmos.Meas.Tech.Discuss.:25July2013
Revised:21October2013–Accepted:24October2013–Published:26November2013
Abstract. We present a new instrument for monitoring accessibletotheQ-ACSM.Thisallowsforquantificationof
+
aerosolcomposition,thetime-of-flightaerosolchemicalspe- certain hydrocarbon and oxygenated fragments (e.g. C H
3 7
ciationmonitor(ToF-ACSM),combiningprecisionstate-of- andC H O+,bothoccurringatm/Q=43Th),aswellasim-
2 3
the-arttime-of-flightmassspectrometrywithstability,relia- provinginorganic/organicseparation.
bility,andeasyhandling,whicharenecessitiesforlong-term
monitoringoperationsonthescaleofmonthstoyears.Based
onAerodyneaerosolmassspectrometer(AMS)technology,
theToF-ACSMprovidescontinuousonlinemeasurementsof 1 Introduction
chemicalcompositionandmassofnon-refractorysubmicron
aerosolparticles.IncontrasttothelargerAMS,thecompact- Over the last decades, ongoing research efforts have solid-
sized and lower-priced ToF-ACSM does not feature parti- ified the knowledge base about the significant role aerosols
clesizing,similartothewidely-usedquadrupole-ACSM(Q- play in Earth’s ecosystem (Gu et al., 2003; Mercado et al.,
ACSM).ComparedtotheQ-ACSM,theToF-ACSMfeatures 2009;Mahowald,2011)andclimate(LohmannandFeichter,
a better mass resolution of M =600 and better detection 2005;Forsteretal.,2007;Carslawetal.,2010).Furthermore,
1M
limits on the order of <30ngm−3 for a time resolution of evidence for severe adverse effects of aerosols on human
health has been reported (Seaton et al., 1995; Laden et al.,
30min. With simple upgrades these limits can be brought
down by another factor of ∼8. This allows for operation at 2000;Cohenetal.,2005;PopeandDockery,2006),though
the mechanisms of action and effect of aerosol composi-
higher time resolutions and in low concentration environ-
tion remain largely unclear. To assess and address these is-
ments. The associated software packages (single packages
suesalargenumberofairqualitymonitoringendeavoursare
forintegratedoperationandcalibrationandanalysis)provide
needed.Essentialtothisareinstrumentscapableofgathering
ahighdegreeofautomationandremoteaccess,minimising
insituinformationaboutthechemicalpropertiesandcompo-
theneedfortrainedpersonnelonsite.Intercomparisonswith
sitionoftheambientparticlesonalong-termbasis.Suchin-
Q-ACSM, C-ToF-AMS, nephelometer and scanning mobil-
strumentscanprovidevaluableinsightsintomanyattributes
ityparticlesizer(SMPS)measurements,performedduringa
of the aerosol, e.g. source or toxicity, with higher time res-
first long-term deployment (>10 months) on the Jungfrau-
olution (minutes to hours) than conventional filter sampling
jochmountainridge(3580ma.s.l.)intheSwissAlps,agree
with subsequent post-processing. Effects on ecosystem and
quantitatively.Additionally,themassresolutionoftheToF-
climate mainly occur on large temporal and spatial scales,
ACSM is sufficient for basic mass defect resolved peak fit-
therefore it is similarly important to be able to collect these
ting of the recorded spectra, providing a data stream not
data over long-term periods. In addition, this facilitates the
PublishedbyCopernicusPublicationsonbehalfoftheEuropeanGeosciencesUnion.
3226 R.Fröhlichetal.:TheToF-ACSM:aportableaerosolchemicalspeciationmonitor
conductofepidemiologicalstudiesusefulinassessinglinks Jungfraujoch(JFJ,3580ma.s.l.),intheSwissAlps.Thisde-
betweenhealthandaerosols. ployment demonstrated the instrument stability, sensitivity,
The various types of the Aerodyne aerosol mass spec- andenabledquantitativecomparisonwithotheraerosolmass
trometer (hereafter denoted AMS; quadrupole-AMS: Jayne spectrometersandparticleinstruments.
etal.,2000,compacttime-of-flight(C-ToF)-AMS:Drewnick
et al., 2005, high resolution time-of-flight (HR-ToF)-AMS:
DeCarloetal.,2006)haveproventobeveryproductiveand 2 Apparatus
powerful tools in terms of recording aerosol mass spectra
with high sensitivity. An overview of numerous AMS cam- The components of the ToF-ACSM are mounted in a rect-
paigns demonstrating the importance of organics in the to- angular rack with edge lengths of 65cm×51cm×60cm,
tal ambient PM1 aerosol budget is shown in Canagaratna it weights 75kg and consumes approximately 330W
etal.(2007),andJimenezetal.(2009)usedAMSdatatoun- when the inlet valve is open (to compare Q-ACSM:
ravelthechemicalevolutionoforganicaerosolintheatmo- 48.3cm×53.3cm×83.8cm, 63.5kg, 300W; HR-ToF-
sphere. However, the monetary and manpower investments AMS:104cm×61cm×135cm,170kg,600W).Thiscom-
associatedwithAMSmeasurementsanddataanalysismake pact size facilitates transport and enables a simpler integra-
thisinstrumentimpracticalforlong-term,widespread,semi- tion into existing monitoring stations or places where space
autonomousmonitoringinitiatives. islimited,e.g.aeroplanes.
The Aerodyne quadrupole aerosol chemical speciation Figure1showsaschemeoftheToF-ACSMwiththemain
monitor (Q-ACSM, Ng et al., 2011) is built upon the same components. A primary difference between the ToF-ACSM
samplinganddetectiontechnologyastheAMSbutwithre- and the Q-ACSM and AMS is the different vacuum system
duced complexity (e.g. no particle size measurement) and design. The vacuum chamber, which has a total length of
performance. The ACSM is specially designed for unat- 43cm (Q-ACSM: same; AMS: 59cm) is divided into four
tended monitoring applications with minimal user interven- differentiallypumpedsections.APfeifferSF2704-stagetur-
tion to close the gap between AMS and filter sampling. It bomolecular pump (www.pfeiffer-vacuum.com) is mounted
is able to record mass spectra of ambient non-refractory directly to the vacuum chamber, and backed by the same
submicron aerosol with unit mass resolution (UMR) up to VacuubrandMD1diaphragmpump(www.vacuubrand.com)
mass to charge ratios of 200Th, although the region above utilised in all AMS systems. The analyser is evacuated
140Th is usually omitted because of its negligible contri- through a direct opening to the vacuum chamber and thus
bution to aerosol mass and a decreasing transmission of doesnotrequireanadditionalpump.Thepressureisreduced
the quadrupole. To date it has been used successfully by overthestagesfrom∼5×10−2mbarattheexitoftheaero-
more than 40 research groups all over the world (cf. Sun dynamiclensto∼10−7mbarintheionisationchambercon-
et al., 2013, 2012; Budisulistiorini et al., 2013; Seto et al., tainingvaporiserandioniser(foracloserdescriptionofinlet
2013; Takahama et al., 2013) and has inspired some very andvaporiser/ioniserseeSect.2.1).Theelectronicsrequired
fruitful international cooperations like the ACSM subgroup foroperationofthesystem,theacquisitionPCandthedata
of the European ACTRIS (Aerosols, Clouds, and Trace acquisition(DAQ)cardareallmountedwithintheinstrument
gases Research Infrastructure Network) project (www.psi. rack.
ch/acsm-stations).Besidestheindividualscientificoutputof
every instrument, the unique databases produced with the 2.1 Operationalprinciple
ACSM by such monitoring networks comprise a combina-
tion of chemical information, high time resolution, long- Aerosol enters the instrument over the inlet system on the
term measurements, and more, like the ability to measure frontal face of the vacuum chamber. This inlet system con-
semivolatilenitrateandorganicswithoutfilterartefactsorof- sistsofanautomatic3-wayvalveswitchsystem,theaerody-
flineanalysis.Thisprovidesinvaluableopportunitiesforthe namic lens and a critical orifice. Aerodynamic lens as well
modellingcommunity. asthevaporiser/ionisersystemareidenticaltothoseusedin
In this manuscript, we present a new instrument based both the Q-ACSM (Ng et al., 2011) and the research grade
on AMS and ACSM technology, the time-of-flight ACSM AMS (Jayne et al., 2000) instruments, except that vaporiser
(ToF-ACSM). This instrument retains the advantages of the and ioniser are divided into two parts to allow the filament
Q-ACSMsuchascompactdesign,semi-autonomousopera- flangetoberemovedeasily.
tion, and relatively low cost, while greatly improving mass With the valve switching system a filter is interposed pe-
resolution and detection limits. It is equipped with a Tofw- riodically into the flow of ambient air to the instrument in
erk ETOF (economy time-of-flight) ion mass spectrometer. order to measure the background signal. The particle signal
Upgradesarepossiblesincethehardwareiscompatiblewith is then obtained by taking the difference between the total
anyTofwerkTOFplatforms.Herewediscusstheoperation, signal measured without a filter (“sample mode”) and the
testing, and initial deployment of a ToF-ACSM for a pe- backgroundsignalmeasuredwithfilter(“filtermode”).This
riodof>10monthsatthehighaltituderesearchstation,the followstheprincipleappliedintheQ-ACSMandutilisesthe
Atmos.Meas.Tech.,6,3225–3241,2013 www.atmos-meas-tech.net/6/3225/2013/
R.Fröhlichetal.:TheToF-ACSM:aportableaerosolchemicalspeciationmonitor 3227
4-stage split flow turbomolecular pump on existing AMS systems, may be directly removed from
thevacuumchamber,allowingaquickandeasyreplacement
of the filaments. The same principle applies to the detector
flange.
100μm particle lens Numerous experiments with AMS and Q-ACSM have
ioniser
vaporiser shown that a fraction of the non-refractory particles does
not flash vaporise at the vaporiser but bounces off the oven
primary beam optics
and subsequently is not detected (Canagaratna et al., 2007;
TOF extractor
Matthewetal.,2008).Acollectionefficiency(CE)factorwas
Extraction
pulser introducedtocorrectforthiseffect.Inthemajorityofcases
theCEisapproximatelyCE=0.5.TheCEcausedbyparticle
preamp ADC
bounce increases with increasing particulate water content,
detector
PC acidity, and nitrate fraction (Middlebrook et al., 2012). Be-
TOFMS
causeitisnotpossibletoquantitativelydeterminethewater
Fig.1.ToF-ACSMschematic.Ioniser/vaporiseraswellasdetector content,adryingsystem(e.g.Nafionmembranedriersfrom
are easily accessible through two vacuum flanges at the backside
PermaPureMD,www.permapure.com)istypicallyinstalled
of the chambers. This enables an easy interchange of the ioniser
onthesamplinginlet.Thecampaignandinstrumentspecific
filamentsorbetweenthevarioussuitabledetectortypes.
CEcanbeassessedforexamplebycorrelationtoco-located
measurements (e.g. SMPS+aethalometer) or by theoretical
considerationsusingthemeasuredchemicalcomposition,as
same switching hardware. The sample flow into the instru- describedbyMiddlebrooketal.(2012).
ment is controlled by a critical orifice. The orifice diameter
for operation under normal pressure conditions is 100µm, 2.2 Time-of-flightmassanalyser
admittingaflowof1.4cm3s−1.
The Tofwerk ETOF mass analyser is based on the CTOF
Theaerodynamiclens(Liuetal.,1995a,b;Liuetal.,2007;
analyser used by the first-generation ToF-AMS systems
Zhang et al., 2004), which consists of a series of apertures
(Drewnick et al., 2005), and shares the same housing and
with decreasing diameter, focuses the submicron particles
interface connections. However, the fine grids used to de-
in the aerosol in a narrow beam into the vacuum chamber
velop the electric fields which guide the ions are replaced
whilethegasesdiverge.Thelighterairmoleculesareprefer-
with metallic plates, providing a more robust, economical
entiallystrippedfromtheaerosolbeamasitpassesthrough
solution at the expense of resolving power and sensitivity.
skimmers separating the 4 differentially-pumped chambers.
Fewervoltages(drifttube:∼3000V,reflectorbackplaneand
Positionanddesignoftheskimmersareparticularlyimpor-
grid:200–800V,pulser:∼800V,detector:∼3500V)arere-
tant for ACSM instruments because of the shorter vacuum
quired to operate the analyser, reducing the complexity of
chamber compared to the AMS. High signals caused by at-
the power supply and potential for failure during long-term
mospheric gases reduce the lifetime of the detector signif-
operation. The ions are detected using an SGE dynode de-
icantly and contribute to interferences in the aerosol mass
tector(www.sge.com),whicharemuchmorerobustthanthe
spectrum.
micro-channelplates(MCP)usedinstandardAMSsystems
At ambient pressure (1013mbar) the lens system has a
and can handle exposure to atmospheric pressure and hu-
close to 100% transmission at vacuum aerodynamic diam-
etersbetweend =150nmandd =450nmandanupper midity. The detector lifetime of the instrument deployed at
va va
cut-off(<15%transmission)aroundd =1µm(Liuetal., the Jungfraujoch for this study was 1yr. Together, the use
va
of the ETOF+SGE system gives an 8-fold (4-fold SGE ×
2007). The transmission for smaller particles (<100nm) is
2-fold ETOF) reduction in sensitivity as compared to the
somewhatreducedcomparedtotheQ-ACSM,whichisare-
CTOF+MCP(fordetectionlimitsseeSect.3.2),andamass
sult of the different vacuum systems and pumping speeds
resolvingpower( M )of∼600insteadof∼1000.Itisnoted
atthelensexitchamber.Arecentlydevelopedaerodynamic 1M
that a specially-designed flange, which bolts directly onto
lens extends the particle transmission to several microme-
the end of the vacuum chamber, is used to hold the detec-
tre (Williams et al., 2013) and is compatible with the ToF-
tor mount, providing unconstrained access and allowing for
ACSM.
fast replacement of an old detector or quick swap-out to a
At the end of the chamber, the particle beam impacts on
a resistively heated porous tungsten surface (T ≈600◦C). higher-sensitivity MCP detector for specific time periods of
interest.ThisisincontrasttotheQ-ACSMsystemwherethe
Therethenon-refractoryconstituentsintheparticlesflashva-
quadrupole head has to be disassembled to replace the de-
poriseandaresubsequentlyionisedbyelectronimpact.The
electrons used for the ionisation (E =70eV) are emitted tector or the AMS where the entire mass analyser must be
kin
removed.
byatungstenfilamentarrangedperpendiculartotheparticle
beam in the vaporisation region. The ioniser flange, unlike
www.atmos-meas-tech.net/6/3225/2013/ Atmos.Meas.Tech.,6,3225–3241,2013
3228 R.Fröhlichetal.:TheToF-ACSM:aportableaerosolchemicalspeciationmonitor
2.3 Dataacquisitionandanalysis start-upoperationisautomated,includingpumpingdownof
thevacuumchambers,makingitsuitableforoperationinthe
ThesignalsaredigitisedusingaUSB-basedacquisitioncard absenceoftrainedpersonnel.
connected to a compact PC. The 14-bit analogue-to-digital For further, more detailed analysis of the data, the HDF
converterdigitisestheanaloguesignalinto214 discretelev- filesmaybeanalysedusingtheIgor-based“Tofware”pack-
elsat0.8GSs−1withamaximummassspectrumacquisition age, which offers the usual analysis features employed by
rateof200Hzandmaximumaveragingdepthof65535ex- theatmosphericsciencecommunitysuchasmasscalibration,
tractions/spectrum.The64-foldincreaseintheresolutionof peakintegrationandhigh-resolutionpeakfitting.Adedicated
thedigitisationwithrespecttotheAcqirisAP240cardused plug-intotheTofwarebasepackageisemployedtodealwith
intheAMSallowsforamuchmoreaccuratesettingofboth theapplicationofACSM-specificcorrections,filtersubtrac-
theelectronicbaselineandnoisesuppressionthreshold.The tionandotherinstrument-specificrequirements.
increaseddynamicrangealsoaffordshigherfidelityrecord-
ing of single ion signals, and thus improved linearity in in-
3 Quantificationofaerosolmass
strumentresponse.
Data are recorded using the TofDaq data recorder from
The ToF-ACSM provides mass spectra of non-refractory
TofwerkAG,whicharefedintopresentationandanalysisus-
submicronparticulatematerialthatvaporisesat∼600◦Cand
ingtheTofwerk“IgorDAQ”packagerunningundertheIgor
∼10−7mbar. Further speciation (e.g. into nitrate, sulphate,
environment(WaveMetricsInc.,OR,USA).
chloride, ammonium, and organic compounds) is attained
Thebackgroundandtotalairmeasurementsarealternated
through analysis of fragmentation patterns (Allan et al.,
throughautomated,synchronisedswitchingofaninletvalve
2004).
suchthattheresultingtimeseriesdataresemblethatof“fast-
mode”MSoftheAMS(Kimmeletal.,2011).Inthismode
3.1 Calibrations
of operation, the background is assumed stable over the
timescale of a single data point (10min), and recorded for For the mass spectra to become quantitative, a number of
only1minbeforeswitchingtototalairsamplingforthere- calibrations are necessary to relate raw detector signals to
maining9min.Toignoretransientsintheinlet,thefirst20s quantitative mass spectra and to account for changes inside
ofdataafteravalveswitchmustbediscarded. (detectorsignaldecay)oroutside(pressure)theinstrument.
The software then interpolates and optionally smoothes
a running background trend between the filter data points, 3.1.1 Inletflow
which is then subtracted from the individual total measure-
ments. The ionisation efficiency (IE), m/Q and flow cali- The dependency of the pressure measured after the critical
brations,baselinesubtraction,sensitivitycorrectionandfrag- orifice on the flow has to be calibrated to detect and even-
mentationpatterns(Jayneetal.,2000;Jimenezetal.,2003; tuallycorrectforchangesintheinletflowduringoperation.
Allanetal.,2004;Drewnicketal.,2005)areapplied,yield- This can be done by connecting a sensitive needle valve to
ing the integrated species time trends at a maximum 20s theinlet andrecording thepressure andflow whileopening
temporalresolution.Thesaveddataproductsthusconsistof itstepwise.
Tofwerk-format HDF files, containing the full mass spectra
3.1.2 Baselineanddetector
and associated diagnostics, and of 10min and 1h timescale
tab-delimited text files of mass loading time trends and im-
IntheToF-ACSMtheretrievaloftheconversionfactorfrom
portant diagnostics, and organic species spectrum matrices
a signal amplitude at the detector measured in mV × ns to
suitable for analysis by tools such as the multilinear en-
ionss−1,theso-calledsingleioncalibration,isfullyautoma-
gine ME-2 (Paatero, 1999; Canonaco et al., 2013) and pos-
tised. The same applies to the determination of the spec-
itive matrix factorisation (PMF; Paatero, 1997; Paatero and
trum’sappropriatebaselineandthedetectorgain.Inregular
Tapper, 1994). In addition, IgorDAQ is designed to oper-
intervals the system checks and, if necessary, readjusts the
atewiththeapplicationprogramminginterfacefromGoogle
baselineandgainsettingsautonomously.
Inc. (www.google.com), allowing for the automated upload
Analogously to all AMS systems, variations of the nitro-
ofdatatoadedicated,password-protectedlocationandsub-
gensignalatm/Q=28Thwhichisassumedtobeconstant
sequentparsinganddisplayofdataproductsinWebbrowsers
due to its abundance in the atmosphere can be used to cor-
and on mobile devices using publicly-available JavaScript
rectforintrinsicchangesintheinstrumentlikeadecayofthe
widgets. The DAQ may also be configured to read param-
iondetectorsignalintheToFmoduleoccurringbetweenthe
eters from the server before the start of each measurement
automaticgainadjustments.
cycle,forexampleifachangeinthefilamentemissioncur-
rentweredesired.Thisalleviatestheneedforremote-desktop
applicationsforwhichtheInternetconnectionmay,particu-
larlyatremotesites,notbesufficient.Finally,theinstrument
Atmos.Meas.Tech.,6,3225–3241,2013 www.atmos-meas-tech.net/6/3225/2013/
R.Fröhlichetal.:TheToF-ACSM:aportableaerosolchemicalspeciationmonitor 3229
3.1.3 m/Q
1010 28.N00267 O2 CO2 184W
Easily identifiable ions from the chamber background are 109
usedtodeterminem/Qasafunctionofiontime-of-flight.As u.)108
highlightedinFig.2,forambientsampling,ionsselectedat al (a.107
thelowm/Qendofthespectrumtypicallyincludenitrogen Sign106
(N+: 28.0067Th), oxygen (O+: 31.9904Th), argon (Ar+: 105
2 2
39.9629Th) and carbon dioxide (CO+: 43.9904Th). Since 104
2
the calibration function is non-linear, one also needs a cali-
brationpointintheheavyendofthespectrum.Aslongasone
usestungstenfilamentsintheioniser,ionsofthefourstable Fig.2.RawmassspectrumoftheToF-ACSM(logarithmicscale).
isotopesoftungsten182W+,183W+,184W+ and186W+ will The x axis is interrupted between m/Q=50Th and 175Th to
show all peaks used in the mass-to-charge calibration. The cali-
alwaysbevisibleandcanbeusedforthem/Qcalibration.In
them/QcalibrationdepictedinFig.2,theisotopewiththe bration peaks are as follows: nitrogen (N2: 28.0067Th), oxygen
nucleon number A=184 (184W+: 183.9509Th) was used. (4O3.29:93014.T99h0),4aTnhd)t,uanrggsotnen(A(1r8:4W39:.9168239.9T5h0)9,Tchar)b.on dioxide (CO2:
The m/Q calibration is dynamically adjusted every 10min
by the software to account for potential drifts in instrument
performanceduringdeployment,e.g.inresponsetochanges For the mIE calibrations, ammonium nitrate (NH NO )
4 3
inroomtemperature. particles of known size and concentration are needed, sim-
ilar to the calibrations of the Q-ACSM (Ng et al., 2011).
3.1.4 Signal-to-mass
Hence,thesamecalibrationequipmentisrequired.NH NO
4 3
is mainly used because it is easily accessible and atomised,
To quantify the mass concentrations measured by the ToF-
vaporiseswith100%efficiencytoionsfromtheammonium
ACSM,thesignaltomassrelationofthedevicehastobede-
andnitratespecies.NH NO isalsowellfocusedbytheaero-
4 3
termined.Amass-basedcalibrationmethodusingthemass-
dynamiclensanddoesnotexperienceparticlebounceatthe
based ionisation efficiency mIE (Onasch et al., 2011) given
vaporiser. Particles of NH NO can be produced from an
4 3
in ions measured per picogram of aerosol particles entering
aqueous solution by a nebuliser, size selected with a DMA
theinstrumentisapplied.Equation(1)yieldsthemasscon-
afterbeingdriedby,e.g.asilicageldrierandthenfedsimul-
centrationγ ofaspeciesiderivedfromthemeasuredsignals
i taneouslytotheToF-ACSMandaCPCforcounting.
I ofitsmassfragmentsj.
i,j Withtheequipmentdescribedabove,afixedamount(be-
tween 300 and 1500cm−3) of NH NO calibration aerosol
1 X 4 3
γ = · I (1) particles with a uniform mobility diameter in the range be-
i (mIE ·q ) i,j
i V j tweendm=300–350nmareselectedwiththeDMAandsam-
pled by the instrument. This diameter and concentration
with γ in units of µgm−3, mIE in ions pg−1, I in ions range is recommended because there the lens still has unit
i i i,j
s−1 and the volumetric sample flow q in cm3s−1. As the transmission and the error caused by doubly charged parti-
V
mIE isdifferentforeveryambientspecies,itisconvenientto cles is minimised. Great care should be taken in the set-up
i
expressthedifferentmIE intermsofmIE (i.e.themIEof ofDMAandCPC.Uncertaintiesinnumberconcentrationor
i NO3
thesumofthemainnitratefragments:NO+ atm/Q=30Th particle size will obviously reduce the accuracy of the mIE
+
and NO m/Q=46Th) determined in the calibration (see calibration. The software then automatically calculates the
2
Eq.3belowandJimenezetal.,2003).Tothisend,arelative mIE onlyfromthesignaloftheNO+andNO+fragments
NO3 2
ionisationefficiencyRIEisdefined: ofthenitrateusingEq.(3):
mIE
RIEi = mIENOi3. (2) mIENO3 = INO3,m/nQ·=ρ3V0+·fIN·Oq3,m/Q=46. (3)
V
The RIE of a species i with respect to the mass-based ion-
i HereI aretheionsignalsinionss−1,nisthenumberof
isation efficiency of NO is unitless. Commonly used RIE i,j
3 particles measured with the CPC in cm−3, ρ the density of
values are 1.4 for organics and 1.3 for chloride. The RIE
NH NO ingcm−3,V thevolumeofoneNH NO particle
values of NH and SO should be calibrated at the begin- 4 3 4 3
4 4 incm3,f thefractionofNH NO thatisnitrateandq the
ning of each deployment and then be reviewed on a regular 4 3 V
sampling flow in cm3s−1. At the same time the RIE is
basis.Typicallytheyliebetween2.5–5and0.6–1.2,respec- NH4
determined from the signal of the ammonium fragments at
tively. During the long-term measurement at the Jungfrau-
joch, the ToF-ACSM RIEs were RIE =3.23±0.42 and m/Q=15, 16 and 17Th via the mass fraction of NH4 in
RIE =0.65±0.05. NH4 NH4NO3.
SO4
www.atmos-meas-tech.net/6/3225/2013/ Atmos.Meas.Tech.,6,3225–3241,2013
3230 R.Fröhlichetal.:TheToF-ACSM:aportableaerosolchemicalspeciationmonitor
Figure 3 shows the summed ToF-ACSM signal of nitrate
3
8x10
(blue,leftaxis)andofammonium(orange,rightaxis)atsev- 3
6x10
eralmassconcentrationsoftherespectivespeciessampledby s) 6 Slope = 773.7 ± 5.11 N
ttchoeenmtTraoastFsio-cAnosCn.ScNeMnot.treFaottihroabntotithsheoibohsnigesrh,veaerldianomevamerroraensiwupmoidnesseirgaonnfagtleh(eodsfeiscgponintae-l O ions (ions/3ToF-ACSM 42 R2 = 0.999825 RSl2ope== 205.989.7928 7±9 1.16 42MSCA-FoT4snoi( snHoi
lessmass)resultsfromthehighRIENH4 (2.99). N RIENH4 = 2.99 ± 0.03 )s/
The final RIENO3, which later is applied to the ambient 0 0
data, is slightly higher than one (RIENO3,ambient=1.1) be- 0 5 10 15 20 25 30
causethetwofragmentsofnitrateusedinthecalibrationonly NO or NH mass (µg/m3)
3 4
accountforabout90%ofthetotalnitratesignal.Thepartof
nitratethatfragmentsintoseparatenitrogenoroxygenatoms Fig. 3. NO3 (blue) and NH4 (orange) ionic signals from the mIE
isnotmonitoredinthecalibrationbecauseofthelowsignal calibrationplottedagainstthemassoftherespectivespecies,calcu-
latedfromtheoutputoftheCPC.Thedashedlinesrepresentlinear
tobackgroundratioatthecorrespondingm/Qratios.
fitstotheNH4(red)andNO3signals(green).Thecorresponding
Once RIE has been measured, the RIE of sulphate
NH4 slopesandcoefficientsofdeterminationaregivenintheboxes(red:
(RIE )caneasilybedeterminedbysampling(NH ) SO
particSlOe4sandadjustingRIE toyieldionbalanceb4et2ween4 NH4; green: NO3). The RIENH4 was calculated from the ratio of
SO4 thedeterminedslopes.
ammoniumandsulphate.
Itisrecommendedtorepeatthesignal-to-masscalibration
atleastevery8weeksduringnormaloperationandwithin-
creasedfrequencyfollowingaventingofthevacuumcham- ammoniumandorganicsignalsareevident,yieldingvariabil-
ber. ityof±0.2µgm−3,whilenitrate,sulphate,andchloridevary
only±0.025µgm−3.
3.2 Detectionlimits The table in Fig. 4b compares published 3-σ detection
limits of each species in ngm−3 for the Q-ACSM (Ng
ThechemicalspeciesdetectablewiththeToF-ACSM(organ-
etal.,2011;Sunetal.,2012)andToF-AMS(DeCarloetal.,
ics,ammonium,nitrate,sulphateandchloride)areretrieved
2006) instruments, scaled by Eq. (4), with those for the
by a recombination of ionic signals of the single fragments
ToF-ACSM equipped with the SGE detector. At 1min time
thatthespeciesbreakdowntoduringionisationandvapori- resolution they are 182ngm−3 for ammonium, 198ngm−3
sation(Allanetal.,2004).Inthefollowing,detectionlimits for organics, 18ngm−3 for sulphate, 21ngm−3 for nitrate,
aredefinedasthreetimesthestandarddeviation(3-σ)from and11ngm−3 forchloride.Inadditionwereportthedetec-
zeroata1mintimeresolution.Theycanbescaledtodiffer-
tion limits for a Q-ACSM measured simultaneously at the
entaveragingtimesusingtheformula
Jungfraujoch(formoreinformationwerefertoSect.4.1)to
r those reported for the ToF-ACSM. In this context it is im-
60s
DL =DL · , (4) portant to note that the detection limits are not absolute but
t 1min
t
willvaryasafunctionofthebackgroundmassspectralsignal
withthedetectionlimitof1minDL andDL beingthe (Drewnicketal.,2009).Aimoffutureworkistofurtherre-
1min t
detectionlimitforagivenaveragingtimet.Detectionlimits ducetheimpactoftheairbeamontherecordedsignals.First
arelargelygovernedbytheextenttowhichbackgroundsig- tests have demonstrated that an order of magnitude reduc-
nal originating from ions of ambient gases (mainly O , N , tionintheair-to-aerosolsignalratiocomparedtothecurrent
2 2
Ar,H O,andCO )interfereswithmeasurementofaselected configuration presented in this manuscript is feasible while
2 2
species. Species with significant signal at m/Q affected by keepingthe totalaerosol throughput highenough. Thiscor-
suchbackgroundsignalhavehigherdetectionlimits. responds to improvements of ∼3 times in organic species
To measure the detection limits, an additional filter was sensitivity.
placed on the ToF-ACSM inlet before the standard fil- An upgrade from the currently used SGE detector to the
tered/unfiltered switching valve, yielding a constant stream less-robustandshorter-livedMCPswilldecreasetheshown
of particle-free air. Figure 4a shows the time series of the detectionlimitsofallspeciesbyanotherfactorofabout4.
difference signal (difference between the signal at the two With the two modifications described above, the ToF-
positions of the switching valve) of this particle-free air. ACSM would advance into the domain of the C-ToF-AMS
This signal is centred at zero by definition, with the ob- whosedetectionlimitsarecurrentlystilllowerbyfactorsbe-
servedfluctuationsdeterminingspecies-dependentdetection tween 10 and 20, except for the chloride whose detection
limits. Each data point in Fig. 4 was obtained from 20s of limitwith11ngm−3alreadyliesatasimilarlylowlevellike
averaging, with the background signal recorded every 360s in the C-ToF-AMS (4ngm−3) and in the V-mode HR-ToF-
(visible as gaps in the data stream). The higher interfer- AMS (12ngm−3). The V-mode HR-ToF-AMS has a better
encesofwatervapourandotheratmosphericgaseswiththe massresolutionbutalsoslightlyhigherdetectionlimitsthan
Atmos.Meas.Tech.,6,3225–3241,2013 www.atmos-meas-tech.net/6/3225/2013/
R.Fröhlichetal.:TheToF-ACSM:aportableaerosolchemicalspeciationmonitor 3231
ThelowerdetectionlimitsoftheToF-ACSMallowforthe
µg/m)3 00..24 organic nitrate sulphate ammonium chloride One datapoint every 20s operationathighertimeresolutions(multipledatapointsper
(onc. -00..20 minute) while achieving detection limits needed for ambi-
mass c -0.4 0(a) 1000 2000 3000 4000 5000 6000 7000 efansttscahmanpgliensg.inTchoisncmeanktreastiiotnwleikllessuhitoerdt-tfeorrmsiptuluatmioenssawreheexre-
Elapsed me (seconds)
(b) QQ--AACCSSMM QQ--AACCSSMM QQ--AACCSSMM QQ--AACCSSMM TTooFF-- VV--mTooFd-AeMS CC--TTooFF--AAMMSS pected.TheToF-ACSMalsoprovidessignificantadvantages
((JNagy neet )al., ((SSuunn et ((TPhSi Is Mstuadyy ((TPhSi Is Astuugd y AACCSSMM (HDReTCoaFr-lAoMS ((DDeeCCaarrlloo et
2011) 2al0.,1 200)12) 2P0S1I 2L)ab 2JF0J122012) ((TPhSi Is Astuugd y 2(D0e0C6a)rlo et 2al0.,0 260)06) in measurement locations with slower temporal variability
2012) 2JF0J1220)12) al., 2006)
but lower concentrations due to its lower detection limits.
AAmmmmoonniiuumm 21655566 11336699 1704572 342530 117802 3388 1166
OOrrggaanniiccss 1831800 22995588 22143667 11250145 119988 2222 1199 For example, the ToF-ACSM detection limit for organics at
SSuullpphhaattee 213301 332299 11026334 1908661 1178 55 22 a15mintimeresolutionis51.1ngm−3.Forsuchalowcon-
NNiittrraattee 16063 338833 223660 17031 1291 33 11
CChhlloorriiddee 9670 116644 112304 17030 191 1122 44 centrationtofallabovetheQ-ACSMdetectionlimit,approx-
*DL are scaled to 1 min by√tmeas/60s imately 9h of integration would be required (based on the
Fig.4.(a)Signalsoffilteredair(inµgm−3)recordedwithanaver- JungfraujochdetectionlimitsscaledbyEq.4).
agingtimeof20s.Duringthegapsinthetimeseriesthebackground
signalwasmonitored.Green:organics,red:sulphate,blue:nitrate,
orange:ammonium,pink:chloride.(b)Overviewoverthe1minde- 4 Firstdatasetandintercomparisons
tectionlimitsinngm−3(3-σ)oftheToF-ACSMequippedwiththe
SGE detector (red box), Q-ACSM, HR-ToF-AMS in V-mode and
The initial deployment of the ToF-ACSM took place at the
C-ToF-AMS.
Sphinx Observatory of the High Altitude Research Station
Jungfraujoch. The (ongoing) measurements proved the in-
the C-ToF-AMS. Compared to the ToF-ACSM, the species strument’s stability and suitability for unattended long-term
dependentlimitsoftheV-modeHR-ToF-AMSarelowerby monitoringoperation.Todateithasproducedmorethanten
factorsbetween4and9. months of quantitative mass spectra of the alpine ambient
TheToF-ACSMimprovesonthedetectionlimitsreported aerosol. Here we discuss ToF-ACSM performance during
fortheQ-ACSMbyNgetal.(2011)byfactorsof8.5(NH ), thisinitialdeploymentandintercomparisonswithco-located
4
4.1 (organics), 7.3 (SO ), 3.1 (NO ), and 5.5 (Cl). Com- instrumentation.
4 3
paredtotheQ-ACSMdetectionlimitsreportedbySunetal.
(2012), the ToF-ACSM shows improvement by a species- 4.1 Generalinformation
dependent factor of 7.5 to 18.3. The difference between Ng
TheSphinxObservatoryoftheHighAltitudeResearchSta-
et al. (2011) and Sun et al. (2012) is probably due to the
tionJungfraujochislocatedinthehighAlpsoftheBernese
different environments the instruments were situated in, i.e.
Oberland(Switzerland)ontopofanexposedrockformation
higherpollutionlevelsalsocausehigherbackgroundsignals.
(3580ma.s.l.) in the ridge between the mountains Jungfrau
Thesamedifferencecanbeobservedinthedetectionlimits
andMönch(coordinates:07◦5900200E,46◦3205300N).Itisac-
measured with Q-ACSM used for the intercomparison pre-
cessiblebycograilroadyear-roundandconnectedtotheIn-
sented in Sect. 4.2.1. The average ambient mass concentra-
ternet, which facilitates the unattended operation of instru-
tionsandrelativehumidityaremuchhigheratPSIthanatthe
mentsbyenablingremoteaccess.
highalpinestationontheJungfraujoch.Thevaluesobserved
While the temperature inside the laboratory is kept more
in thePSI laboratory arecomparable to the values recorded
orlessconstantabove20◦Candtheinletisconstantlyheated
in Beijing by Sun et al. (2012), while the results from the
to 25◦C, the monthly average outside temperature on the
JungfraujocharecomparabletoNgetal.(2011).Thisistrue
Jungfraujochstaysbelowthefreezingpointofwaterallyear
for the organics, the nitrate and the chloride. The ammo-
(between−1.2◦CinJulyandAugustand−14.2◦CinFebru-
niumandsulphatedetectionlimits,however,areverydiffer-
ary). The temperature difference between inside and out-
ent in our instrument. A possible explanation is that the Q-
side and the generally low relative humidity at that altitude
ACSMsensitivitytothesespeciesdependsmorestronglyon
guaranteedryaerosolwithouttheneedforadditionaldrying
massspectrometertuningandthereforeismoreinstrument-
equipment.Becauseofthelowatmosphericpressure(onav-
specific. The side by side comparison of the Q-ACSM and
erage around 650mbar), a 130µm critical orifice was used
theToF-ACSMontheJungfraujochshowanimprovementof
in place of the standard 100µm orifice to retain the normal
thedetectionlimitswiththelatterbyfactorsof1.8(ammo-
samplemassflowrate.Tominimiselossesinthe5mmcop-
nium), 6.1 (organics), 58.9 (sulphate), 3.5 (nitrate), and 6.6
pertubingofthe∼2minletline,anadditionalflowofabout
(chloride).Theenhancementislowestforammonium,which
50cm3s−1(sampleflow∼2.4cm3s−1)isused.
ismostaffectedbyairandwaterinterference,andhighestfor
The aerosol concentrations at the Jungfraujoch are low
sulphate,havingasurprisinglyhighlimitofdetectioninour
(average PM concentration in 2012: ∼3µgm−3; BAFU,
Q-ACSMsystem. 10
2012), approximately 25% of which is mineral dust (Col-
laud Coen et al., 2004) originating to a considerable part
www.atmos-meas-tech.net/6/3225/2013/ Atmos.Meas.Tech.,6,3225–3241,2013
3232 R.Fröhlichetal.:TheToF-ACSM:aportableaerosolchemicalspeciationmonitor
from occasional Saharan dust events. These refractory par-
(a)
ticles are not detectable with the ACSM. These low con- 1
centrations occur because the station lies within the clean 2
continental free troposphere most of the year, especially in
3
winter,whileinsummeraerosolgeneratedatloweraltitudes
reachestheJungfraujochduetoverticalexchangeprocesses
(Lugaueretal.,1998;Henneetal.,2004).TheJungfraujoch
siteisafamoustouristdestinationandtheSphinxObserva-
torycanalsobevisitedbytourists,leadingtooccasionallo-
cal emission plumes from, e.g. helicopters, snow crawlers,
restaurantsorcigarettesmoke. (b)
3
6 ammonium
Local time on the Jungfraujoch is Central European
chloride
Time(UTC+01:00),usingdaylightsavingtimeinsummer
4 nitrate
(UTC+02:00).Alldatareportedinthismanuscriptisgiven organic
inUTC. 2 sulphate
0
4.2 Intercomparisons
09:30 09:40 09:50
The High Altitude Research Station Jungfraujoch is inte- 10.08.2012 UTC
gratedintoseveralmonitoringnetworks,includingtheSwiss
Fig. 5. (a) Organic mass concentrations measured on 10 Au-
national air pollution monitoring network NABEL (Na-
gust 2012 with the ToF-ACSM (red line, time resolution: 10min)
tionales Beobachtungsnetz für Luftfremdstoffe), the GAW
and the Q-ACSM (black line, time resolution: 30min). The aver-
(GlobalAtmosphereWatch)programmeoftheWMO(World
agespectrarecordedintheregionsmarkedinyellow(14:00–16:00,
Meteorological Organization) and the European ACTRIS
case 1, higher concentrations), green (04:00–06:00, case 2, lower
(Aerosols, Clouds, and Trace gases Research InfraStruc-
concentrations)andblue(10:00,case3,short-termpeakonlyseen
ture Network) network. As such, a large number of co- byToF-ACSM)arediscussedinFig.6.(b)Timeseriesduringpe-
located long-term measurements of both gas- and aerosol- riod3withthehighestavailabletimeresolutionof20s.Thegaps
species are conducted. For example, continuous long-term inthedataindicateperiodstheinstrumentwasacquiringtheback-
measurements of the aerosol scattering coefficient (with groundconcentrations.
a nephelometer) and the particle number size distribution
(with a scanning mobility particle sizer, SMPS) are per-
formed within the framework of the GAW program, and a (green,γ ≈0.35µgm−3),and(3)duringoneofthepeaks
org
Q-ACSM was operated at the JFJ as part of the ACTRIS at10:00(blue).
monitoring project during the second half of 2012. In ad- The Q-ACSM (black line) operated with a time resolu-
dition, during the intensive INUIT-JFJ/CLACE (Ice Nuclei tion of 30min and the ToF-ACSM (red line) with a 10min
Research Unit/CLoud and Aerosol Characterization Exper- time resolution. The two time traces follow the same trend
iment) campaign of February 2013, additional instruments andshowgoodagreementintermsofabsoluteconcentration
were deployed at the Jungfraujoch site, including a C-ToF- buttheToF-ACSMobservesseveralshort-termsignalpeaks.
AMS.Datafromthesewereusedtointercomparewiththose A closer look at one of the peaks (3, Fig. 5b) in the ToF-
recordedbytheToF-ACSM. ACSM at 20s time resolution demonstrates that this plume
actuallyconsistsoftwoplumes,eachoflessthan3mindura-
4.2.1 Q-ACSM tion.TherearetworeasonswhytheQ-ACSMdoesnotmea-
suretheseplumes.First,thetimeresolutionoftheQ-ACSM
The Q-ACSM was operated for five months alongside the measurements(30min)decreasesabilityoftheinstrumentto
ToF-ACSM at the Jungfraujoch site, utilising a shared inlet resolve rapidly changing concentrations. More importantly,
line. afundamentaldifferenceexistsbetweenQ-ACSMandToF-
In Fig. 5a, the mass concentration of the organic frac- ACSMmassspectralacquisition,inthattheToF-ACSMac-
tion measured by the two instruments over the course of quires the entire mass spectrum simultaneously (i.e. with
the 10 August 2012 is shown. This day provides the op- eachextraction,seeFig.1),whiletheQ-ACSMsequentially
portunity to study the output mass spectra at three different scans m/Q, with each scan cycle lasting several minutes.
interesting situations circled in the figure. These situations Thus,iftheQ-ACSMisnotscanningtherelevantpartofthe
are (1) concentrations well above the detection limit of the massspectrumduringtheplume,suchaplumemaygounde-
Q-ACSM between 14:00 and 16:00 (yellow, organic mass tectedoryieldonlyapartialmassspectrum,biasingthemea-
concentration γ ≈1.90µgm−3), (2) concentrations close surement.SuchabiasisevidentinFig.5ainthataveraging
org
to the Q-ACSM detection limit between 04:00 and 06:00 theToF-ACSMdatatothe30mintimescaleoftheQ-ACSM
Atmos.Meas.Tech.,6,3225–3241,2013 www.atmos-meas-tech.net/6/3225/2013/
R.Fröhlichetal.:TheToF-ACSM:aportableaerosolchemicalspeciationmonitor 3233
does not cause the instruments to agree, with the aerosol the quality of the Q-ACSM’s mass spectrum deteriorates.
massheresignificantlyunderestimatedbytheQ-ACSM.The Thequalityofthemassspectrumiscrucialfortheapplication
veracity of the short duration peaks measured by the ToF- ofstatisticalsourceapportionmentmethodslikepositivema-
ACSM was verified by comparison with nephelometer data trix factorisation (PMF; Paatero, 1997; Paatero and Tapper,
collectedat5mintimeresolution(seeSect.4.2.3andFig.8). 1994)andthemultilinearengine(ME-2;Paatero,1999).
Figure 6 shows Q-ACSM and ToF-ACSM mass spectra Theorganicstickspectrumofcase3shownatthebottom
averaged over the periods identified in Fig. 5a, arranged as ofFig.6lookssignificantlydifferenttothetwootherrather
follows. The top panel shows the raw difference spectrum similarToF-ACSMspectraincases1and2.Thepronounced
(calculatedfromthedifferencebetweentheunfilteredandfil- signalsatthemass-to-chargeratios41Th,55Th,57Thand
teredrawspectrum)fortheQ-ACSMduringperiod1,while 69Th point towards a significant fraction of hydrocarbon-
thesecondpanelshowstheorganicfractionofthisspectrum likeorganics(HOA)intheaerosol.Acomparisontoanambi-
integratedtointegerm/Q(“organicstickspectrum”).Panels entmassspectrumofHOAfromacampaigninParis(Crippa
3and4showthecorrespondingrawandorganicstickspectra et al., 2013) yields a good correlation with an R2 of 0.78
fortheToF-ACSM.Thesameplotsaredisplayedforperiod2 for case 3, while the ToF-ACSM spectra in cases 1 and 2
in panels 5 through 8, while panel 9 gives the organic stick do not correlate with the HOA spectrum (R2=0.42 and
1
spectrum for period 3. To make the spectral patterns more R2=0.22). This result illustrates the utility of highly time
2
visible,onlytheregionbetweenm/Q=40Thand100This resolvedmeasurementsevenformonitoringapplications,as
shownandinsomeplotstheyaxisisinterruptedtoshowthe herethelocalactivityofmachinerycanbeidentifiedanddis-
highsignalsatm/Q=43Thandm/Q=44Th.Coefficients tinguishedfromtherestofthedataset.
of determination (R2) given in the following are calculated The correlation in terms of absolute concentrations mea-
onlyforthepartoftheorganicstickspectrabetweenm/Q= suredwiththeToF-andQ-ACSMisillustratedinFig.7.On
45Thand100Th;otherwisethecorrelationisdominatedby thetop,thetimeseriesoforganics,nitrate,sulphate,ammo-
thehighsignalsat,e.g.m/Q=28Thorm/Q=44Th. nium,andchloridemeasuredwiththeToF-ACSMduringthe
A comparison of both total raw difference spectrum and first six months of the deployment are shown together with
stick organic spectrum between Q-ACSM and ToF-ACSM thecorrespondinginletpressuresandairbeamsignal.These
forthecasewiththehigherconcentration(1)showsreason- two parameters are used to account for changes in flow or
ableagreement(R2=0.62)exceptatthemasstochargera- detectorgainasdescribedinSect.3.1.Itisnotedthatmuch
tiosofnaphthalenefragments(m/Q=61–64Th,m/Q=74– of the variation in the airbeam signal observed in this time
78Th, m/Q=86–89Th and m/Q=98–102Th). This is ex- seriesarisesfrompurposefulchangesinsettingsfortheopti-
pected because the Q-ACSM (for calibration reasons) has misationofsignal.Thegapsinthetimeserieshavedifferent
a source of naphthalene (C10H8) incorporated inside the reasons: planned interruptions to install hardware upgrades,
vacuum chamber (Ng et al., 2011), constantly releasing software development, extended calibration and testing, or
moleculesandtherewithcausingalargerbackgroundsignal unplanned replacement of failed prototype power modules.
at the corresponding m/Q ratios. Additional differences in The prototype stage has now concluded and the system has
the spectra originate from two short-lived plumes observed beenrunningcontinuouslyforseveralmonthswithoutinter-
during the averaging period (cf. Fig. 5a), not being cap- ruption.
turedbytheQ-ACSMduetothescanningofthequadrupole Figure 7c shows a two-week period (3 to 15 August) se-
and lower temporal resolution. The correlation between Q- lectedforfurtheranalysisoftheToF-ACSM/Q-ACSMinter-
ACSMandToF-ACSMbecomesmuchworsewhenthelow comparison. The organic mass concentrations measured by
concentration case 2 is compared. While the ToF-ACSM the Q-ACSM (black line) and by the ToF-ACSM (red line,
spectrumofcase2remainssimilartotheToF-ACSMspec- averaged to 30min time resolution) are plotted on the same
trumofcase1(R2=0.81),theQ-ACSMspectrumofcase2 axis. Both instruments report approximately the same con-
nowlooksverydifferentandevenhassomenegativevalues centrations and exhibit good temporal correlation. The gap
atm/Q’sinfluencedbynaphthalenefragments.AR2of0.19 in this case was caused by a temporary interruption of the
betweenQ-andToF-ACSMatcase2confirmsthisobserva- ionisercurrentduetoaresolvedhardwareissue.
tion. Looking at the raw difference spectra, one can easily AtthebottomleftofFig.7,scatterplotsofthemasscon-
recognise that while the ToF-ACSM spectrum still exhibits centrations of the four main species (chloride is not plotted
nicelyseparated,Gaussian-shapedsignals,theQ-ACSMraw becausemeasuredconcentrationsalwayswerebelowdetec-
differencespectrumisnoisy.Thepeaksarenotaseasilysep- tionlimitattheJungfraujoch)areshown.Thedatarecorded
arable as in the raw difference spectrum of the Q-ACSM in with the ToF-ACSM are drawn on the y axis and the data
case 1. In summary these observations confirm that when from the Q-ACSM on the x axis. Slope and coefficient of
sampling from the same inlet, the mass spectra of both in- determination retrieved with a least orthogonal distance fit
strumentsshow a goodagreement aslong as theconcentra- to the scatter data for each species are given in the plots.
tionsarewellabovethedetectionlimitoftheQ-ACSM.Ap- There is a very good correlation for organics (R2=0.95,
proaching the detection limit (DL =0.136µgm−3),
org,30min
www.atmos-meas-tech.net/6/3225/2013/ Atmos.Meas.Tech.,6,3225–3241,2013
3234 R.Fröhlichetal.:TheToF-ACSM:aportableaerosolchemicalspeciationmonitor
TOTAL SIGNAL
1 ONLY ORGANICS SIGNAL
Quad
TOTAL SIGNAL
1 ONLY ORGANICS SIGNAL
ToF
TOTAL SIGNAL
2 Quad ONLY ORGANICS SIGNAL
TOTAL SIGNAL
2 ToF ONLY ORGANICS SIGNAL
3 ONLY ORGANICS SIGNAL
ToF R22 = 0.78
HOA
Fig.6.AveragespectraoftheregionshighlightedinFig.5.Thespectraaregroupedintofourpairsofrawdifferencespectrum(calculated
from the difference between the unfiltered and filtered raw spectrum) and the corresponding integrated UMR stick spectrum. The raw
differencespectrashowthetotalsignalwhiletheUMRstickspectraonlyshowtheorganicfraction.SpectrarecordedwiththeQ-ACSM
areshowninblueandToF-ACSMspectrainred.Thefirsttwopairsarefromtheafternoonperiodwiththehigherconcentrations(1),the
followingtwopairsarefromthemorningperiodwiththelowerconcentrations,(2)andthelast(9th)spectrumisfromthepeak,whichwas
onlyseenwiththeToF-ACSM(3).Coefficientsofcorrelationwerecalculatedforthespectralregionbetweenm/Q=45Thand100Th.
BetweenToF-ACSMspectrumandQ-ACSMspectrumtheyareR2=0.62forcase1andR2=0.19forcase2.TheR2giveninthelastplot
showsthecorrelationtoareferenceHOAspectrumfrom(Crippaetal.,2013).
slope=1.06), nitrate (R2=0.94, slope=0.95) and sulphate measures higher values. This may be caused by additional
(R2=0.87, slope=1.16) between ToF- and Q-ACSM. The noise in the scatter plot due to the low ammonium concen-
largest difference is found for ammonium, which still has trations(maximum0.5µgm−3)typicalofhighaltitudesites
a good correlation (R2=0.74, slope=1.30) but compared (Beig and Brasseur, 2000) and/or to the sensitivity of am-
to the other species the scatter is larger and the ToF-ACSM moniumRIEtoinstrumenttuning,wheresmallerrorsinthe
Atmos.Meas.Tech.,6,3225–3241,2013 www.atmos-meas-tech.net/6/3225/2013/