Reactive processing of formaldehyde and acetaldehyde in - ACPD

15 Pages·2011·1.94 MB·English

Atmos. Chem. Phys. Discuss.,11,19477–19506,2011 D
Atmospheric is
www.atmos-chem-phys-discuss.net/11/19477/2011/ Chemistry cu
s
doi:10.5194/acpd-11-19477-2011 andPhysics s
©Author(s)2011. CCAttribution3.0License. Discussions ion
P
a
p
e
Thisdiscussionpaperis/hasbeenunderreviewforthejournalAtmosphericChemistry r
andPhysics(ACP).PleaserefertothecorrespondingfinalpaperinACPifavailable. |
D
is
c
u
s
s
Reactive processing of formaldehyde and ion
P
a
acetaldehyde in aqueous aerosol mimics: p
e
r
surface tension depression and |
D
secondary organic products is
c
u
s
s
io
Z.Li,A.N.Schwier,N.Sareen,andV.F.McNeill n
P
a
DepartmentofChemicalEngineering,ColumbiaUniversity,NewYork,NY,10027,USA pe
r
Received:15June2011–Accepted:4July2011–Published:7July2011
|
Correspondenceto:V.F.McNeill([email protected]) D
is
PublishedbyCopernicusPublicationsonbehalfoftheEuropeanGeosciencesUnion. c
u
s
s
io
n
P
a
p
e
r
19477 |
Abstract D
is
c
u
The reactive uptake of carbonyl-containing volatile organic compounds (cVOCs) by ss
aqueous atmospheric aerosols is a likely source of particulate organic material. The ion
aqueous-phasesecondaryorganicproductsofsomecVOCsaresurface-active. There- P
a
p
5 fore, cVOCuptakecanleadtoorganicfilmformationatthegas-aerosolinterfaceand er
changesinaerosolsurfacetension. Weexaminedthechemicalreactionsoftwoabun-
|
dant cVOCs, formaldehyde and acetaldehyde, in water and aqueous ammonium sul-
fate(AS)solutionsmimickingtroposphericaerosols. Secondaryorganicproductswere D
is
identifiedusingAerosolChemicalIonizationMassSpectrometry(Aerosol-CIMS),and cu
s
10 changesinsurfacetensionweremonitoredusingpendantdroptensiometry. Hemiac- sio
etal oligomers and aldol condensation products were identified using Aerosol-CIMS. n
P
A hemiacetal sulfate ester was tentatively identified in the formaldehyde-AS system. a
p
Acetaldehydedepressessurfacetensionto65(±2)dyncm−1inpurewaterand62(±1) e
r
dyncm−1inASsolutions. Surfacetensiondepressionbyformaldehydeinpurewateris
|
negligible;inASsolutions,a9%reductioninsurfacetensionisobserved. Mixturesof
15
D
thesespecieswerealsostudiedincombinationwithmethylglyoxalinordertoevaluate is
c
the influence of cross-reactions on surface tension depression and product formation u
s
inthesesystems. Wefindthatsurfacetensiondepressioninthesolutionscontaining sio
n
mixedcVOCsexceedsthatpredictedbyanadditivemodelbasedonthesingle-species
P
20 isotherms. ap
e
r
|
1 Introduction
D
is
Organic material is a ubiquitous component of atmospheric aerosols, making up a c
u
major fraction of fine aerosol mass, but its sources and influence on aerosol proper- ss
io
ties are still poorly constrained (Kanakidou et al., 2005; Jimenez et al., 2009). Many n
P
25 common organic aerosol species are surface-active (Shulman et al., 1996; Facchini a
p
etal., 1999). Surface-activemoleculesinaqueoussolutionformstructuresthatallow e
r
19478 |
hydrophobicgroupstoavoidcontactwithwaterwhilehydrophilicgroupsremaininso- D
is
lution. Inanaqueousaerosolparticle,theymaypartitiontothegas-aerosolinterface, c
u
s
reducingaerosolsurfacetensionandpotentiallyactingasabarriertogas-aerosolmass s
io
transport(Folkersetal.,2003;McNeilletal.,2006). Depressedaerosolsurfacetension n
P
5 duetofilmformationmayleadtoadecreaseinthecriticalsupersaturationrequiredfor ap
the particle to activate and grow into a cloud droplet as described by Ko¨hler Theory e
r
(Kohler,1936). Thesurfacetensionofatmosphericaerosolsamplestendstobelower
|
thanthatpredictedbasedonthecombinedeffectsoftheindividualsurfactantsidenti-
D
fiedintheaerosol(Facchinietal.,1999). Thisisinpartbecausesomesurface-active is
10 aerosol organics remain unidentified. Additionally, the effects of interactions among cus
thesespeciesundertypicalaerosolconditions(i.e.supersaturatedsaltconcentrations, sio
acidic,multipleorganicspecies)aregenerallyunknown. n
P
Theadsorptionofvolatileorganiccompounds(VOCs)toaqueousaerosolandcloud ap
e
dropletsurfaceshasbeenproposedasaroutefortheformationoforganicsurfacefilms r
15 (DjikaevandTabazadeh,2003;DonaldsonandVaida,2006). Thereisalsogrowingev- |
idencethatthereactiveuptakeofthecarbonyl-containingVOCs(cVOCs)methylglyoxal
D
andglyoxalbyclouddropletsoraerosolwater, followedbyaqueous-phasechemistry is
c
to form low-volatility products, is a source of secondary organic aerosol material (Er- us
s
vens and Volkamer, 2010; Lim et al., 2010). We recently showed that methylglyoxal io
n
20 suppressessurfacetensioninaqueousaerosolmimics(Sareenetal.,2010). P
a
Formaldehyde and acetaldehyde, two abundant, highly volatile aldehydes, can be p
e
directlyemittedfromcombustionandindustrialsourcesorgeneratedinsituviatheoxi- r
dationofotherVOCs(SeinfeldandPandis,1998). Inaqueoussolution,bothformalde- |
hydeandacetaldehydebecomehydratedandformacetaloligomers,similartomethyl- D
25 glyoxalandglyoxal(Loudon,2009). Nozie`reandcoworkersshowedthatacetaldehyde isc
u
forms light-absorbing aldol condensation products in aqueous ammonium sulfate so- s
s
lutions (Nozie`re et al., 2010a). Formaldehyde was also recently suggested to react io
n
with amines to form organic salts in tropospheric aerosols (Wang et al., 2010). Due P
a
totheirprevalenceandknownaqueous-phaseoligomerizationchemistry,thereactive pe
r
19479 |
processingofthesespeciesinaqueousaerosolmimics,aloneandincombinationwith D
is
othercVOCs,isofinterest,buthasnotbeenthoroughlystudiedtodate. c
u
s
We investigated the chemical reactions of formaldehyde and acetaldehyde in pure s
io
water and concentrated ammonium sulfate (AS) solutions mimicking aerosol water. n
P
5 The potential of these species to alter aerosol surface tension was examined, and ap
secondary organic products were identified using Aerosol Chemical Ionization Mass e
r
Spectrometry(Aerosol-CIMS).
|
D
is
2 Experimentalmethods c
u
s
s
Aqueoussolutionscontainingvaryingconcentrationsoforganiccompounds(acetalde- io
n
hyde, formaldehyde and/or methylglyoxal) with near-saturation concentrations (3.1M) P
10 a
of AS were prepared in 100 mL Pyrex vessels using Millipore water. The concentra- pe
r
tion of acetaldehyde was 0.018M–0.54M. In the preparations, 5mL ampules of 99.9
wt%acetaldehyde(SigmaAldrich)weredilutedto1.78MusingMilliporewaterimme- |
diately after opening in order to minimize oxidization. Varying amounts of this stock D
is
15 solution were used to prepare the final solutions within 30 min of opening the am- cu
pule. Formaldehyde and methylglyoxal (MG) were introduced from 37 wt% and 40 ss
io
wt% aqueous solutions (Sigma Aldrich), respectively. The pH value of the reaction n
P
mixtures,measuredusingadigitalpHmeter(Accumet,FisherScientific),was2.7–3.1. a
p
The surface tension of each sample was measured 24 h after solution preparation e
r
usingpendantdroptensiometry(PDT).Pendantdropsweresuspendedfromthetipof
20
|
glass capillary tubes using a 100µl syringe. The images of the pendant drops were
capturedandanalyzedtodeterminetheshapefactor,H,andequatorialdiameter,d , D
e is
asdescribedpreviously(Sareenetal.,2010;Schwieretal.,2010). Theseparameters cu
s
wereusedtocalculatethesurfacetensionaccordingto: s
io
n
∆ρgd2 P
25 σ= H e (1) ape
r
19480 |
whereσissurfacetension,∆ρisthedifferenceindensitybetweenthesolutionandthe D
is
gas phase, and g is acceleration due to gravity (Adamson and Gast, 1997). Solution c
u
s
density was measured using an analytical balance (Denver Instruments). The drops s
io
were allowed to equilibrate for 2 min before image capture. Each measurement was n
P
5 repeated7times. ap
Aerosol-CIMS was used to detect the organic composition of the product mixtures e
r
asdescribedindetailpreviously(Sareenetal.,2010;Schwieretal.,2010). Mixturesof
|
formaldehyde, acetaldehyde-MG, and formaldehyde-MG in water and 3.1M AS were
D
prepared. Totalorganicconcentrationrangedfrom0.2–2M.AlltheASsolutionswere is
c
10 diluted at 24h with Millipore water until the salt concentration was 0.2M. The solu- us
tions were aerosolized in a stream of N2 using a constant output atomizer (TSI) and sio
flowedthroughaheated23cmlong, 1.25cmIDPTFEtube(maintainedat135◦C)at n
P
RH>50%beforeenteringtheCIMS,inordertovolatilizetheorganicspeciesintothe ap
e
gas phase for detection. The time between atomization and volatilization (≤3.5s) is r
15 tooshortfordetectablequantitiesoftheexpectedreactionproductstoform,therefore |
thedetectedmoleculesaremostlikelyformedinthebulkaqueoussolutions. Thesolu-
tionsweretestedinbothpositiveandnegativeionmode,usingH O+·(H O) andI−as Dis
3 2 n c
reagentions,respectively. Theapplicabilityofthisapproachtothedetectionofacetal us
s
oligomersandaldolcondensationproductsformedbydicarbonylsinaqueousaerosol io
n
20 mimicshasbeendemonstratedpreviously(Sareenetal.,2010;Schwieretal.,2010). P
Theaverageparticleconcentrationwas∼4×104cm−3 andthevolumeweightedgeo- ap
e
metricmeandiameterwas414(±14)nm. r
The Pyrex vessels shielded the reaction mixtures from UV light with wavelengths |
<280nm(Corning,Inc.),butthesampleswerenotfurtherprotectedfromvisiblelight. D
25 Wepreviouslyshowedthatexposuretovisiblelightinidenticalvesselsdoesnotimpact isc
u
chemistryintheglyoxal-ASorMG-ASreactivesystems(Sareenetal.,2010; Shapiro s
s
etal.,2009). io
n
P
a
p
e
r
19481 |
3 Results D
is
c
u
3.1 Surfacetensionmeasurements ss
io
n
3.1.1 Single-organicmixtures P
a
p
e
Results of the PDT experiments (Fig. 1) show that both formaldehyde and acetalde- r
5 hyde depress surface tension in 3.1M AS solution, but the formaldehyde mixture |
is less surface-active than that of acetaldehyde. The formaldehyde-AS solutions D
reach a minimum surface tension of 71.4±0.4dyncm−1 at 0.082molC (kg H O)−1. is
2 c
u
This represents a 9% reduction in surface tension from that of a 3.1M AS solution s
s
(78.5±0.3dyncm−1). The acetaldehyde-AS solutions showed more significant sur- io
n
face tension depression. The surface tension of the solutions reached a minimum P
10 a
of 62±1dyncm−1 when the acetaldehyde concentration exceeded 0.527molCkg pe
(H O)−1 (20.6% reduction compared to 3.1M AS solution). Compared to the sur- r
2
face tension of the acetaldehyde in 3.1M AS, the surface tension depression of ac- |
etaldehyde in water is less significant. The surface tension of acetaldehyde in water D
15 decreases rapidly and reaches a minimum value of 65±2dyncm−1 at 0.89molCkg iscu
(H2O)−1. Formaldehyde does not show any detectable surface tension depression in ssio
waterintheabsenceofAS. n
P
ThesurfacetensiondatacanbefitusingtheSzyszkowski-Langmuirequation: a
p
e
σ=σ −aTln(1+bC) (2) r
0
|
where σ and σ are surface tension of the solution with and without organics, T is
20 0 D
ambient temperature (298K), C is total organic concentration (moles carbon per kg is
c
H O), and a and b are fit parameters (Adamson and Gast, 1997). The parameters u
2 s
fromthefitstothedatainFig.1arelistedinTable1. sio
n
P
a
p
e
r
19482 |
3.1.2 Binarymixtures D
is
c
u
Surfacetensionresultsforaqueoussolutionscontainingamixtureoftwoorganiccom- ss
pounds (MG and formaldehyde or acetaldehyde) and 3.1M AS are shown in Fig. 2. ion
Foragiventotalorganicconcentration(0.5or0.05M),thesurfacetensiondecreased Pa
p
5 withincreasingMGconcentration. Re-plottingthedatafromFig.2asafunctionofMG er
concentration,itisapparentthatthesurfacetensionwasverysimilarformixtureswith
|
thesameMGconcentration,regardlessoftheidentityoramountoftheotherspecies
presentinthemixture(Fig.3). D
is
Henning and coworkers developed the following model based on the Szyszkowski- cu
s
Langmuir equation to predict the surface tension of complex, nonreacting mixtures of s
10 io
organics(Henningetal.,2005): n
P
σ=σ (T)−ΣχaTln(1+bC) (3) ap
0 i i i i i e
r
Here, C is the concentration of each organic species (moles carbon per kg H O), χ
i 2 i |
is the concentration (moles carbon per kg H O) of compound i divided by the total
2
D
15 soluble carbon concentration in solution, and ai and bi are the fit parameters from is
the Szyszkowski-Langmuir equation for compound i. The Henning model has been cu
s
showntodescribemixturesofnonreactiveorganics,suchassuccinicacid-adipicacid sio
ininorganicsaltsolution,well(Henningetal.,2005). Wealsofoundthatitwascapable n
P
ofdescribingsurfacetensiondepressioninreactiveaqueousmixturescontainingMG, a
p
e
20 glyoxal,andAS(Schwieretal.,2010). r
Thepredictedsurfacetensiondepressionforthebinarymixturesascalculatedwith
|
theHenningmodelisshowninFig.2asablackline,andtheconfidenceintervalsbased
D
on uncertainty in the Szyszkowski-Langmuir parameters are shown in grey. The ex- is
c
perimentallymeasuredsurfacetensionsare,ingeneral,lowerthantheHenningmodel u
s
25 prediction,indicatingasynergisticeffectbetweenMGandacetaldehyde/formaldehyde. sio
n
TheerrorofthepredictionforthemixturesofMGandacetaldehydeisbetween8–24%. P
a
The error tends to increase with the concentration of MG. However, the error is less p
e
than10%forformaldehyde-MGmixtures. r
19483 |
3.1.3 Ternarymixtures D
is
c
u
AsshowninFig.4,3.1MASsolutionscontainingternarymixturesofMG,acetaldehyde ss
andformaldehydealsoexhibitsurfacetensiondepressionlowerthanthatpredictedby ion
theHenning model. For theternary mixtureexperiments, the molarratio ofacetalde- Pa
p
5 hydetoformaldehydewaseither1:3(Fig.4aandb)or1:1(Fig.4candd)andtheMG er
concentrationwasvaried. Thetotalorganicconcentrationremainedconstantat0.05M.
|
RecastingthedataofFig.4asafunctionofMGconcentrationshowsasimilartrendas
whatwasobservedforthebinarymixtures;foraconstanttotalorganicconcentration, D
is
MGcontentlargelydeterminesthesurfacetension,regardlessoftherelativeamounts cu
s
ofacetaldehydeandformaldehydepresent(Fig.3). s
10 io
n
3.2 Aerosol-CIMScharacterization Pa
p
e
r
TheCIMSdatashowproductsofself-andcross-reactionsofformaldehyde,acetalde-
hydeandMGinpurewaterand3.1MAS.WedidnotperformAerosol-CIMSanalysis |
onacetaldehyde-ASoracetaldehyde-H Osolutionsbecausethesesystemshavebeen D
2 is
15 characterizedextensivelybyothers(Nozie`reetal.,2010a;Casaleetal.,2007). cu
s
s
3.2.1 Formaldehyde io
n
P
a
The mass spectra for formaldehyde in H O and in 3.1M AS obtained using negative p
2 e
ion detection with I− as the reagent ion is shown in Fig. 5. Possible structures are r
showninTable2. Thespectrumshowspeakswithmass-to-chargeratioscorrespond- |
20 ipnegatkosfocromnsicisatecnidtawtit8h1.h7e(mCiHacOe−2ta·2lHol2igOo)maenrds.20282.73.a3m, u29(1I−.1·C,Han2Od23·22H3.25Oa)maundarseevceorna-l Disc
sistent with clusters of hemiacetals with I−. A small amount of formic acid impurity us
s
exists in the 37% formaldehyde aqueous stock solution (Sigma Aldrich). The peaks io
n
at 95.6, 110.4, 273.8 and 304.7amu are consistent with clusters of ionized hemiac- P
a
25 etals with H2O. While ionization of alcohols by I− is normally not favorable, ionized per
19484 |
paraformaldehyde-typehemiacetalsarestabilizedbyinteractionsbetweentheionized D
−O− andtheotherterminalhydroxylgroup(s)onthemolecule(seetheSupplement). isc
u
s
Withinourinstrumentresolution,thepeaksatm/z176.7and193.8amupeakscould s
io
be consistent with the mass of methanol, which is added to commercial formalde- n
5 hyde solutions as a stabilizer. However, methanol is not predicted to cluster with I−. Pap
Furthermore, these peaks are not observed in the formaldehyde-H O spectrum and e
2 r
are present only with the addition of AS, implying that the species observed at those
|
masses are formed via reaction with AS. No favorable pathway for the reduction of
D
formaldehydetoformmethanolinanacidicmediumisknown. Thepeakat193.8amu is
c
10 is consistent with an organosulfate species formed from a formaldehyde hemiacetal us
dimer (C2H5O6S−·2H2O). The peak at 176.7amu matches an ion formula of C6H9O−6 sio
orC H O S−·H O,butthestructuresandformationmechanismsofthosespeciesare n
2 7 6 2 P
a
unknown. p
e
Thepositive-ionspectrumoftheformaldehydesolutionin3.1MAScorroboratesthe r
15 identificationofhemiacetaloligomers. Toourknowledge, organosulfatespecieshave |
notpreviouslybeenobservedusingproton-transfermassspectrometry(Sareenetal.,
D
2010). ThespectrumandpeakassignmentscanbefoundintheSupplement. is
c
u
s
s
3.2.2 Formaldehyde-methylglyoxalmixtures io
n
P
Thenegative-ionspectrum(detectedwithI−)foranaqueousmixtureofformaldehyde, a
p
e
20 MG,andASisshowninFig.6,withpossiblepeakassignmentslistedinTable3. Most r
of the peaks are consistent with formaldehyde hemiacetal oligomers, such as 186.7,
|
203.5, 230.3, 257.4, and 264.5amu. Formic acid was detected at 172.8amu and
D
208.7amu. Thepeakat288.1correspondstoMGself-reactionproductsformedeither is
c
viaaldolcondensationorhemiacetalmechanisms(Sareenetal.,2010;Schwieretal., u
s
s
25 2010). Several peaks could correspond to self-reaction products of either formalde- io
n
hydeorMG:216.5,252.4,324.5,and342.6amu. Thepeakat314.3amuisconsistent P
a
with a hemiacetal oligomer formed via cross-reaction of MG with two formaldehyde p
molecules, clustered with I− and two water molecules. The peak at 272.2 amu could er
19485 |
correspond to either a similar cross-reaction product (MG+2 formaldehyde) or a MG D
is
dimer. Formaldehydehemiacetalself-reactionproductsandformicacidweredetected cu
s
inthepositive-ionspectrum(Supplement). s
io
n
3.2.3 Acetaldehyde-methylglyoxalmixtures Pa
p
e
The H O+.(H O) spectrum for aqueous acetaldehyde-MG-AS mixtures is shown in r
5 3 2 n
Fig. 7, with peak assignments listed in Table 4. Hydrated acetaldehyde can be ob- |
servedat 98.4amu. Severalpeaks areconsistent withthecross-reaction productsof D
is
MGandacetaldehydeviaanaldolmechanism(126.0, 134.0, 206.7, and248.9amu). c
u
Formic, glyoxylic, and glycolic acids correspond to the peaks at 84.4, 93.5, and 95.5 ss
io
10 amu, respectively. Since no significant source of oxidants exists in the reaction mix- n
tures, the formation mechanisms for these species in this system are unknown. The Pa
p
peaks at 88.9 and 107.2 are consistent with either pyruvic acid or crotonaldehyde. e
r
Large aldol condensation products from the addition of 6–10 acetaldehydes are ob-
|
servedat192.9,289.6,and297amu. Thepeaksat145.1,162.9,164.7and235amu
areconsistentwithMGself-reactions,asdiscussedbySareenetal.(2010). Thepeak D
15 is
at 137.3amu is consistent with a species with molecular formula C5H12O3, but the cus
mechanismisunknown. s
The I− negative-ion spectrum for acetaldehyde-MG-AS mixtures shows similar re- ion
P
sultstothepositive-ionspectrum(seeFig.8andTable5),howeveraldolcondensation a
p
20 productsarenotdetectedbythismethodunlesstheycontainaterminalcarboxylicacid er
grouporneighboringhydroxylgroups(Sareenetal.,2010). Smallacidspecies,such
|
asformic,aceticandcrotonicacid(172.7(208.4),186.4and230.7amu,respectively),
D
were detected. Hydrated acetaldehyde (189.6 and 224.1amu) and MG (216.3amu), is
c
andhemiacetalself-dimersofacetaldehydeandMG(230.7,256.4,264.4,269.5,and u
s
25 342.3) were also observed. 256.4amu is consistent with a MG aldol condensation sio
n
dimerproduct,and272.2amucouldcorrespondeithertoaMGhemiacetaldimeroran
P
aldolcondensationproduct. 242.9amu,I−·C5H8O3,isconsistentwithanaldolconden- ape
sation cross product of MG and acetaldehyde. 194.6amu corresponds to both of the r
19486 |
following ion formulas, C6H9O−6 and C2H7O6S−·H2O, but the mechanisms and struc- Dis
turesareunknown. cu
s
Notethatseveralpeaksappearatsimilarmass-to-chargeratiosinthenegativemode s
io
massspectraofboththeformaldehyde-MGandacetaldehyde-MGmixtures. MGself- n
P
reactionproductsareexpectedtobepresentinbothsystems. Beyondthis,formalde- a
5 p
hyde and acetaldehyde are structurally similar small molecules which follow similar er
oligomerizationmechanismsaloneandwithMG.Inseveralcases,peaksinthemass
|
spectracorrespondingtostructurallydistinctexpectedreactionproductsforeachsys-
D
tem have similar mass-to-charge ratios. For example, the formaldehyde hemiacetal is
4-mer (I−.C H O ) and the acetaldehyde dimer (I−.C H O.2H O) are both apparent cu
10 4 10 5 4 6 3 2 ss
at264amu. io
n
P
a
p
4 Discussion e
r
|
Both formaldehyde and acetaldehyde, and their aqueous-phase reaction products,
were found to depress surface tension in AS solutions. However, surface tension de- D
is
pressionwasnotobservedinaqueousformaldehydesolutionscontainingnosalt. Net c
15 u
s
surface tension depression by acetaldehyde was greater in the AS solutions than in s
pure water. These differences are likely due to chemical and physical effects of the ion
P
salt. Thesaltpromotestheformationofsurface-activespecies: severalofthereaction a
p
products in the AS systems identified using Aerosol-CIMS are known or expected to e
r
be surface-active, such as organosulfates (Nozie`re et al., 2010b) and organic acids.
20 |
Saltscanalsoalterthepartitioningofthesevolatileyetwater-solubleorganicspecies
D
between the gas phase and aqueous solution. Formaldehyde has a small Henry’s is
Lawconstantof2.5Matm−1,althoughhydrationintheaqueousphaseleadstoanef- cu
s
fective Henry’s Law constant of 3×103Matm−1, similar to that of MG (Seinfeld and sio
Pandis,1998;BettertonandHoffmann,1988). TheeffectiveHenry’sLawconstantfor n
25 P
acetaldehyde in water at 25◦C was measured by Betterton and Hoffmann (1988) to ap
be11.4Matm−1. TheHenry’sLawconstantofformaldehydewasshownbyZhouand er
19487 |
Mopper to increase slightly in aqueous solutions containing an increasing proportion D
is
of seawater (up to 100%), but the opposite is true for acetaldehyde (Zhou and Mop- c
u
s
per,1990). Thereactionmixturesstudiedhereequilibratedwiththegasphasefor24h s
io
beforethesurfacetensionmeasurementswereperformed. Eachpendantdropequili- n
P
5 bratedfor2minbeforeimagecapture,afterwhichtimetherewasnodetectablechange ap
indropshape. Someoftheorganicsmaybelosttothegasphaseduringequilibration. e
r
However,thelowervolatilityoftheaqueous-phasereactionproducts,especiallythose
|
formed through oligomerization, leads to significant organic material remaining in the
D
condensed phase (enough to cause surface tension depression and be detected via is
c
10 Aerosol-CIMS). us
WhenformaldehydeandacetaldehydearepresentincombinationwithMG,aswould sio
likelyhappenintheatmosphere,thereisasynergisticeffect: surfacetensiondepres- n
P
sion in the solutions containing mixed organics exceeds that predicted by an additive ap
modelbasedonthesingle-speciesisotherms. Thiseffectcouldbeduetotheformation er
15 ofmoresurface-activereactionproductsinthemixedsystems. Thedeviationfromthe |
Henningmodelpredictionwaslessthan10%exceptinthecaseoftheacetaldehyde-
D
MG-AS mixtures. Between 21–30% of the detected product mass was identified as is
c
cross products in the Aerosol-CIMS positive mode analysis of the acetaldehyde-MG us
s
mixtures following Schwier et al. (2010). Most of the oligomers identified in this sys- io
n
20 temwere aldolcondensationproducts, whichhavefewerhydroxyl groupsthanacetal P
a
oligomers and are therefore expected to be more hydrophobic. A number of organic p
e
acidproducts,likelytobesurface-active,werealsoidentifiedintheacetaldehyde-MG- r
ASsystem. |
In contrast to the MG-glyoxal system (Schwier et al., 2010), the presence of D
25 formaldehyde and/or acetaldehyde in aqueous MG-AS solutions does influence sur- isc
u
face tension depression, in fact, to a greater extent than predicted by the Henning s
s
model. However, the results of the binary and ternary mixture experiments suggest io
n
that MG still plays a dominant role in these systems since the measured surface ten- P
a
sionwasremarkablysimilarineachmixtureforagivenMGconcentration. pe
r
19488 |
Theformaldehydehemiacetaldimer(C H O S)mayformviathereactionofC H O D
2 6 6 2 6 3 is
with H2SO4 (Deno and Newman, 1950). The equilibrium concentration of H2SO4 in cu
our bulk solutions (3.1M AS, pH=3) is small (2.8×10−7M). Minerath and coworkers ss
io
showed that alcohol sulfate ester formation is slow under tropospheric aerosol con- n
P
ditions (Minerath et al., 2008). Based on our observations, assuming a maximum a
5 p
Aerosol-CIMSsensitivityof100Hzppt−1 tothisspecies(Sareenetal.,2010)weinfer er
a concentration of ≥2×10−4M in the bulk solution after 24h of reaction. Using our
|
experimental conditions and the kinetics of ethylene glycol sulfate esterification from
Minerathetal.,wepredictamaximumconcentrationof7×10−8M.Thissuggeststhat Dis
c
either (a) the kinetics of sulfate esterification for paraformaldehyde are significantly u
10 s
faster than for alcohols (b) SO−42 or HSO−4 is the active reactant, contrary to the con- sion
clusionsofDenoandNewman,or(c)sulfateesterificationisenhancedbyatomization.
P
See the Supplement for details of these calculations. Photochemical production of ap
e
organosulfateshasalsobeenobserved(Gallowayetal., 2009; Nozie`reetal., 2010b; r
15 Perri et al., 2010). Our samples were protected from UV light by the Pyrex reaction |
vessels,andnosignificantOHsourcewaspresent,sowedon’texpectphotochemical
D
organosulfateproductiontobeefficientinthissystem. is
c
u
Nitrogen-containingcompoundscouldalsobeformedinthesereactionmixturesdue s
s
tothepresenceoftheammoniumion(Sareenetal., 2010; Nozie`reetal., 2009; Gal- io
n
20 loway et al., 2009). No unambiguous identifications of C-N containing products were Pa
madeinthisstudy,butanalysisusingamassspectrometrytechniquewithhighermass p
e
resolutioncouldrevealtheirpresence. r
The relatively low solubility of formaldehyde and acetaldehyde in water suggests |
thattheirpotentialtocontributetototalSOAmassislowascomparedtohighlysoluble D
25 speciessuchasglyoxal. ThisissupportedbytheobservationsofKrolletal.(2005)that iscu
ASaerosolsexposedtoformaldehydeinanaerosolreactionchamberdidnotresultin s
s
significant particle volume growth. However, formaldehyde and acetaldehyde in the ion
gasphasecouldadsorbattheaerosolsurface(vs.bulkaqueousabsorption),andthis P
a
p
mayalsoimpactaerosolsurfacetension(DonaldsonandVaida, 2006). Furthermore, e
r
19489 |
Romakkaniemi and coworkers recently showed significant enhancement of aqueous- D
is
phase SOA production by surface-active species beyond what would be predicted c
u
s
basedonHenry’sLawduetosurface-bulkpartitioning(Romakkaniemietal.,2011). s
io
n
P
a
5 Conclusions p
e
r
5 Twohighlyvolatileorganiccompounds,formaldehydeandacetaldehyde,werefoundto |
formsecondaryorganicproductsinaqueousammoniumsulfate(AS)solutionsmimick- D
ingtroposphericaerosols. Thesespecies,andtheiraqueous-phasereactionproducts, is
c
u
leadtodepressedsurfacetensionintheaqueoussolutions. Thisaddstothegrowing s
s
bodyofevidencethatVOCsareasecondarysourceofsurface-activeorganicmaterial io
n
inaerosols. P
10 a
p
e
Supplementarymaterialrelatedtothisarticleisavailableonlineat: r
http://www.atmos-chem-phys-discuss.net/11/19477/2011/ |
acpd-11-19477-2011-supplement.pdf. D
is
c
u
Acknowledgements. This work was funded by the NASA Tropospheric Chemistry program ss
15 (grantNNX09AF26G)andtheACSPetroleumResearchFund(Grant48788-DN14). Theau- ion
thorsgratefullyacknowledgetheKobersteingroupatColumbiaUniversityforuseofthependant P
a
droptensiometer. p
e
r
|
References
D
is
Adamson,A.W.andGast,A.P.:Physicalchemistryofsurfaces,Wiley,NewYork,1997. cu
20 Betterton,E.A.andHoffmann,M.R.: HenryLawConstantsofSomeEnvironmentallyImpor- ss
tantAldehydes,Environ.Sci.Technol.,22(12),1415–1418,1988. ion
Casale, M. T., Richman, A. R., Elrod, M. J., Garland, R. M., Beaver, M. R., and Tolbert, M. P
a
A.: Kinetics of acid-catalyzed aldol condensation reactions of aliphatic aldehydes, Atmos. p
e
Environ.,41(29),6212–6224,2007. r
19490 |
Deno, N. C. and Newman, M. S.: Mechanism of Sulfation of Alcohols, J. Am. Chem. Soc., D
is
72(9),3852–3856,1950. c
u
Djikaev, Y. S. and Tabazadeh, A.: Effect of adsorption on the uptake of organic trace gas s
s
by cloud droplets, J. Geophys. Res.-Atmos., 108(D22), 4869, doi:10.1029/2003JD003741, io
n
5 2003. P
a
Donaldson,D.J.andVaida,V.: Theinfluenceoforganicfilmsattheair-aqueousboundaryon p
e
atmosphericprocesses,Chem.Rev.,106(4),1445–1461,2006. r
Ervens, B. and Volkamer, R.: Glyoxal processing by aerosol multiphase chemistry: towards
|
akineticmodelingframeworkofsecondaryorganicaerosolformationinaqueousparticles,
D
10 Atmos.Chem.Phys.,10,8219–8244,doi:10.5194/acp-10-8219-2010,2010. is
Facchini, M. C., Mircea, M., Fuzzi, S., and Charlson, R. J.: Cloud albedo enhancement by cu
s
surface-activeorganicsolutesingrowingdroplets,Nature,401(6750),257–259,1999. s
Folkers,M.,Mentel,T.F.,andWahner,A.: Influenceofanorganiccoatingonthereactivityof ion
aqueous aerosols probed by the heterogeneous hydrolysis of N O , Geophys. Res. Lett., P
2 5 a
15 30(12),1644–1647,doi:10.1029/2003GL017168,2003. pe
Galloway,M.M.,Chhabra,P.S.,Chan,A.W.H.,Surratt,J.D.,Flagan,R.C.,Seinfeld,J.H., r
andKeutsch,F.N.: Glyoxaluptakeonammoniumsulphateseedaerosol: reactionproducts |
and reversibility of uptake under dark and irradiated conditions, Atmos. Chem. Phys., 9,
D
3331–3345,doi:10.5194/acp-9-3331-2009,2009. is
c
20 Henning, S., Rosenrn, T., D’Anna, B., Gola, A. A., Svenningsson, B., and Bilde, M.: Cloud us
dropletactivationandsurfacetensionofmixturesofslightlysolubleorganicsandinorganic s
io
salt,Atmos.Chem.Phys.,5,575–582,doi:10.5194/acp-5-575-2005,2005. n
P
Jimenez,J.L.,Canagaratna,M.R.,Donahue,N.M.,Prevot,A.S.H.,Zhang,Q.,Kroll,J.H., a
p
DeCarlo,P.F.,Allan,J.D.,Coe,H.,Ng,N.L.,Aiken,A.C.,Docherty,K.S.,Ulbrich,I.M., e
r
25 Grieshop,A.P.,Robinson,A.L.,Duplissy,J.,Smith,J.D.,Wilson,K.R.,Lanz,V.A.,Hueglin,
C.,Sun,Y.L.,Tian,J.,Laaksonen,A.,Raatikainen,T.,Rautiainen,J.,Vaattovaara,P.,Ehn, |
M.,Kulmala,M.,Tomlinson,J.M.,Collins,D.R.,Cubison,M.J.,Dunlea,E.J.,Huffman,J. D
A.,Onasch,T.B.,Alfarra,M.R.,Williams,P.I.,Bower,K.,Kondo,Y.,Schneider,J.,Drewnick, is
c
F.,Borrmann,S.,Weimer,S.,Demerjian,K.,Salcedo,D.,CottrellL.,Griffin,R.,Takami,A., u
s
30 Miyoshi, T., Hatakeyama, S., Shimono, A., Sun, J. Y., Zhang, Y. M., Dzepina, K., Kimmel, sio
J. R., Sueper, D., Jayne, J. T., Herndon, S. C., Trimborn, A. M., Williams, L. R., Wood, E. n
P
C.,Middlebrook,A.M.,Kolb,C.E.,Baltensperger,U.,Worsnop,D.R.: EvolutionofOrganic a
p
AerosolsintheAtmosphere,Science,326(5959),1525–1529,2009. e
r
19491 |
Kanakidou,M.,Seinfeld,J.H.,Pandis,S.N.,Barnes,I.,Dentener,F.J.,Facchini,M.C.,Van D
is
Dingenen,R.,Ervens,B.,Nenes,A.,Nielsen,C.J.,Swietlicki,E.,Putaud,J.P.,Balkanski, c
u
Y., Fuzzi, S., Horth, J., Moortgat, G. K., Winterhalter, R., Myhre, C. E. L., Tsigaridis, K., s
s
Vignati,E.,Stephanou,E.G.,andWilson,J.:Organicaerosolandglobalclimatemodelling: io
n
5 areview,Atmos.Chem.Phys.,5,1053–1123,doi:10.5194/acp-5-1053-2005,2005. P
a
Kohler,H.: Thenucleusinthegrowthofhygroscopicdroplets,Trans.FaradaySoc.,32,1152– p
e
1161,1936. r
Kroll,J.H.,Ng,N.L.,Murphy,S.M.,Varutbangkul,V.,Flagan,R.C.,andSeinfeld,J.H.:Cham-
|
berstudiesofsecondaryorganicaerosolgrowthbyreactiveuptakeofsimplecarbonylcom-
D
10 pounds,J.Geophys.Res.-Atmos.,110(D23),D23207,doi:10.1029/2005JD006004,2005. is
Lim,Y.B.,Tan,Y.,Perri,M.J.,Seitzinger,S.P.,andTurpin,B.J.: Aqueouschemistryandits cu
s
roleinsecondaryorganicaerosol(SOA)formation,Atmos.Chem.Phys.,10,10521–10539, s
doi:10.5194/acp-10-10521-2010,2010. ion
Loudon,G.M.:Organicchemistry,RobertsandCo.,GreenwoodVillage,CO,2009. P
a
15 McNeill, V. F., Patterson, J., Wolfe, G. M., and Thornton, J. A.: The effect of varying levels pe
of surfactant on the reactive uptake of N O to aqueous aerosol, Atmos. Chem. Phys., 6, r
2 5
1635–1644,doi:10.5194/acp-6-1635-2006,2006. |
Minerath, E. C., Casale, M. T., and Elrod, M. J.: Kinetics feasibility study of alcohol sulfate
D
esterificationreactionsintroposphericaerosols,Environ.Sci.Technol.,42(12),4410–4415, is
c
20 2008. us
Nozie`re,B.,Dziedzic,P.,andCordova,A.:ProductsandKineticsoftheLiquid-PhaseReaction s
+ io
ofGlyoxalCatalyzedbyAmmoniumIons(NH ),J.Phys.Chem.A,113(1),231–237,2009. n
Nozie`re,B.,Dziedzic,P.,andCordova,A.: Inorg4anicammoniumsaltsandcarbonatesaltsare Pa
efficientcatalystsforaldolcondensationinatmosphericaerosols,Phys.Chem.Chem.Phys., pe
r
25 12(15),3864–3872,2010a.
Nozie`re, B., Ekstrom, S., Alsberg, T., and Holmstrom, S.: Radical-initiated formation of |
organosulfatesand surfactantsin atmospheric aerosols, Geophys. Res. Lett., 37, L05806, D
doi:10.1029/2009GL041683,2010b. is
c
Perri,M.J.,Lim,Y.B.,Seitzinger,S.P.,andTurpin,B.J.:Organosulfatesfromglycolaldehydein u
s
30 aqueousaerosolsandclouds: Laboratorystudies,Atmos.Environ.,44(21–22),2658–2664, sio
2010. n
P
Romakkaniemi, S., Kokkola, H., Smith, J. N., Prisle, N. L., Schwier, A. N., McNeill, V. F., a
p
andLaaksonen, A.: PartitioningofSemivolatileSurface-ActiveCompoundsBetweenBulk, e
r
19492 |
Surface,andGas-Phase,Geophys.Res.Lett.,38(3),L03807,doi:10.1029/2010GL046147, D
is
2011. c
u
Sareen, N., Schwier, A. N., Shapiro, E. L., Mitroo, D., and McNeill, V. F.: Secondary organic s
s
materialformedbymethylglyoxalinaqueousaerosolmimics,Atmos.Chem.Phys.,10,997– io
n
5 1016,doi:10.5194/acp-10-997-2010,2010. P
a
Schwier, A. N., Sareen, N., Mitroo, D. M., Shapiro, E. L., and McNeill, V. F.: Glyoxal- p
e
MethylglyoxalCross-ReactionsinSecondaryOrganicAerosolFormation,Environ.Sci.Tech- r
nol.,44(16),6174–6182,2010.
|
Seinfeld, J. H. and Pandis, S. N.: Atmospheric Chemistry and Physics: From air pollution to
D
10 climatechange,Wiley,NewYork,1998. is
Shapiro,E.L.,Szprengiel,J.,Sareen,N.,Jen,C.N.,Giordano,M.R.,andMcNeill,V.F.:Light- cu
s
absorbingsecondaryorganicmaterialformedbyglyoxalinaqueousaerosolmimics,Atmos. s
Chem.Phys.,9,2289–2300,doi:10.5194/acp-9-2289-2009,2009. ion
Shulman,M.L.,Jacobson,M.C.,Carlson,R.J.,Synovec,R.E.,andYoung,T.E.:Dissolution P
a
15 behavior and surface tension effects of organic compounds in nucleating cloud droplets, pe
Geophys.Res.Lett.,23(3),277–280,1996. r
Wang, X. F., Gao, S., Yang, X., Chen, H., Chen, J. M., Zhuang, G. S., Surratt, J. D., Chan, |
M.N.,andSeinfeld,J.H.:EvidenceforHighMolecularWeightNitrogen-ContainingOrganic
D
SaltsinUrbanAerosols,Environ.Sci.Technol.,44(12),4441–4446,2010. is
20 Zhou, X. L. and Mopper, K.: Apparent Partition-Coefficients of 15 Carbonyl-Compounds Be- cus
tween Air and Seawater and Between Air and Fresh-Water – Implications for Air Sea Ex- s
io
change,Environ.Sci.Technol.,24(12),1864–1869,1990. n
P
a
p
e
r
|
D
is
c
u
s
s
io
n
P
a
p
e
r
19493 |
D
is
c
u
s
s
io
n
P
a
p
e
r
|
D
is
Table1.Szyszkowski-LangmuirFitParametersaccordingtoEq.(2). cu
s
s
io
Mixture σ (dyncm−1) a(dyncm−1K−1) b(kgH O(molC)−1) n
0 2 P
a
Methylglyoxal+ 78.5 0.0185±0.0008 140±34 p
e
3.1M(NH ) SO r
4 2 4
(Sareenetal.,2010) |
Acetaldehyde+ 78.5 0.0008±0.0046 9.53±3.86
D
3.1M(NH4)2SO4 is
Formaldehyde+ 78.5 0.0119±0.0043 50.23±44.8 cu
s
3.1M(NH ) SO s
Acetaldehy4d2e+4 72.0 0.0037±0.0011 491.64±689 ion
H O P
2 a
p
e
r
|
D
is
c
u
s
s
io
n
P
a
p
e
r
19494 |
440 Table 1. Szyszkowski-Langmuir Fit Parameters according to Eq. (2)
Mixture σ (dyn cm-1) a (dyn cm-1 K-1) b (kg HO (mol C)-1)
0 2
Methylglyoxal +
3.1 M (NH)SO 78.5 0.0185±0.0008 140±34
42 4
(Sareen et al. 2010)
Acetaldehyde +
78.5 0.0008±0.0046 9.53±3.86
3.1 M (NH)SO
42 4
Formaldehyde +
78.5 0.0119±0.0043 50.23±44.8
3.1 M (NH)SO
42 4
Acetaldehyde + D
H2O 72.0 0.0037±0.0011 491.64±689 iscu
s
441 s
io
Ta4b4l2e 2. ProposedpeakassignmentsforAerosol-CIMSmassspectrawithI−ofatomizedsolu- nP
tio4n4s3 of0T.a2bMle f2o. rPmroapldoesehdy dpeeakin a3ss.1igMnmAenSt.s for Aerosol-CIMS mass spectra with I- of atom- ap
444 ized solutions of 0.2 M formaldehyde in 3.1 M AS. e
r
m/z (amu) Molecular Possible
Ion Formula Mechanism |
± 1.0 amu Formula Structures
D
81.7 CHO2-·2H2O CH2O2 Formic Acid is
c
95.6 C2H5O3-·H2O C2H6O3 n=2 hemiacetal uss
110.4 C2H3O3-·2H2O C2H4O3 n=2 hemiacetal ion
P
176.7 C6H9O6- C6H10O6 Unknown Unknown ap
CHOS-·HO CHOS e
2 7 6 2 2 8 6 r
193.8 CHOS-·2HO CHOS Hemiacetal sulfate
2 5 6 2 2 6 6 |
208.7 I-·CH2O2·2H2O CH2O2 Formic Acid D
is
223.3 I-·C2H6O3·H2O C2H6O3 n=2 hemiacetal cu
s
sio
273.8 C8CH81H5O179O-·H10-2 O CC88HH1186OO190 n=8 hemiacetal nPa
p
e
r
291.1 I-·CHO CHO n=5 hemiacetal
5 8 6 5 8 6 |
304.7 C9H19O10-·H2O C9H20O10 n=9 hemiacetal Dis
c
u
323.5 I-·C6H14O7 C6H14O7 n=6 hemiacetal ssio
445 n
P
a
p
e
r
19495 |
17

D
Table3.Proposedpeakassignments446forAerosol-CIMSmassspectrawithI− ofatomized is
c
solutionsof2M44f6o rmTaabllde 3e. hPryodpoese/dM peGak (a1ss:ig1n)meinnts3 fo.r1 AMerosAolS-C.IMS mass spectra with I- of atom- us
s
447 ized solutions of 2 M formaldehyde/MG (1:1) in 3.1 M AS. io
m/z (amu) Molecular n
± 1.0 amu Ion Formula Formula Possible Structures Mechanism P
a
172.8 I-·CH2O2 CH2O2 Formic Acid pe
186.7 I-·C2H4O2 C2H4O2 cyclic F acetal r
203.5 C5H11O6-·2H2O C5H12O6 n=5 F hemiacetal |
208.7 I-·CH2O2·2H2O CH2O2 Formic Acid Dis
216.5 I-·C3H6O3 C3H6O3 Hcyydcrlaicte Fd aMceGta ol r cus
s
230.3 I-·C3H4O4 C3H4O4 n = 3 F hemiacetal io
n
252.4 II-·-·CCI3-3H·HC668OHO364·O·2HH32 2OO CCC633HHH686OOO343 nH =cy y3dMc rFlaiG tche Fedaml MdaicoaGelc,t ,ea tloa rl , Paper
257.4 C8H17O9- C8H18O9 n=8 F hemiacetal |
264.5 I-·C4H10O5 C4H10O5 n=4 F hemiacetal D
I-·C6H10O4 C6H10O4 MG aldaocle atanld hemi- iscus
272.2 I-·C5H6O5 C5H6O5 hMemGi a+c e2tFa l sionP
a
288.1 II-·-·CCI-6·6HCH668OHO314·0·2OHH52 2OO CCC666HHH1860OOO435 MG aldaocle atanld hemi- per|

314.3 I-·C5H12O5·2H2O C5H12O5 hMemGi a+c e2tFa l Dis
c
324.5 II-·-·CCI6-6H·HC1610H2OO1546·O·2HH72 2OO CCC666HHH111420OOO765 n=M6G F hheemmiiaacceettaall , ussionP
342.6 II-·-·CC66HH1124OO67··2HH22OO CC66HH1142OO76 n=M6G F hheemmiiaacceettaall , aper
448 19496 |
449
18