Evaporation of Ammonium Nitrate

7/27/2019 Evaporation of Ammonium Nitrate

1/6

Evaporation of Ammonium NitrateAerosol in a Heated Nephelometer:Implications for Field Measurements

M I C H A E L H . B E R G I N , * , ,

J O H N A . O G R E N ,

S T E P H E N E . S C H W A R T Z , A N DL Y N N M . M C I N N E S

Climate Monitoring and Diagnostics Laboratory,National Oceanic and Atmospheric Administration,Boulder, Colorado, 80303, and Environmental ChemistryDivision, Brookhaven National Laboratory,Upon, New York, 11973-5000

Ammonium nitrate is a semivolatile aerosol componentunder typical ambientconditionsandthusdifficulttomeasure.In the field, the aersosol scattering coefficient is usuallymeasuredwitha nephelometer byheatingtheambientaerosolto a low reference relative humidity (40%) in order to

measure a light scattering coefficient that is intrinsic to anaerosol rather than dependent on atmospheric relativehumidity. In this paper, we examinethe decrease in the lightscattering coefficient of ammonium nitrate aerosol dueto evaporation in a heated nephelometer. Changes in thescatteringcoefficientof alaboratory-generatedammoniumnitrate aerosol aremeasuredas a functionof meanresidencetime and temperature within the nephelometer samplevolume. At the same time, the change in the aerosol sizedistributiondue toammoniumnitrateevaporationis directlymeasured with a laser particle counter. The change in theaerosol size distribution and scattering coefficient is mod-eled as a function of mean residence time andtemperature. Model results for the change in the aerosol

scattering coefficient due to evaporation agree withmeasurements to within 10%. Application of the theory toconditions typical of NOAA field sites suggests that thedecrease in the aerosol scattering coefficient due to theevaporation of ammoniumnitrate is generally less than20%.

In t roduct ionThe aerosol light scattering coefficient, typically measuredby an integratingnephelometer (1, 2),is a keyaerosol propertyaffecting atmospheric visibility as well as the extent to whichaerosols scatter shortwave radiation away from the earthssurface, termed direct shortwave aerosol radiative forcing

(3). Aerosols often contain hygroscopic inorganic salts (i.e.,ammonium sulfate, ammonium nitrate) which increase insize with increasing relative humidity (RH), resulting in anincreasein theaerosollight scattering coefficient (4). Inorderto measure a light scattering coefficient that is intrinsic to anaerosol rather than dependent on atmospheric relativehumidity, it is customary to perform measurements at arelatively lowRH (typically less than 40%). This is commonly

accomplished by heating the ambient airstream under theassumption that aerosol chemical species other than waterare not volatilized. The evaporation of relatively volatileaerosol chemical species, such as ammonium nitrate, be-comes a concern in sampling systems that heat ambient air.

Ammonium nitrate aerosol is common to urban areas andin some cases contributes substantially to the total aerosolextinction on a regional scale. In Los Angeles, a significantfraction of the aerosol mass in the accumulation mode (i.e.,having diameters between 0.1 m and 1.0 m) has beenattributed to particulate ammonium nitrate (5, 6). Similarly,it has been reported that ammonium nitrate can account forroughly 20% of the total aerosol extinction measured duringthe winter in Denver (7). Measurements in The Netherlandshave estimated that light scattering by nitrate aerosol iscomparable to that of sulfate in thesummer andeven greaterduring the winter (8). Although measurements indicateammonium nitrate is prevalent in urban plumes, there is agreat deal of uncertainty associated with measurement ofthe aerosol mass size distribution because of the volatilenature of this aerosol (9, 10).

The NOAAClimate Monitoring and Diagnostics Laboratory(CMDL) maintains a series of regional and base line moni-toring stations to addressissuesrelated toaerosols andclimate

change. Measurements include the total aerosol numberconcentration, total light scattering and backscattering coef-ficients, light absorption coefficient, ionic composition, andtotal aerosol mass (11). Stations impacted significantly byurban plumes are Bondville, IL, Sable Island, Nova Scotia,and Niwot Ridge, CO. At NOAA aerosolmonitoring stations,the inlet airstream is heated to maintain a reference RH of40% (11). When the ambient RHis less than40%,no heatingtakes place at the inlet. The heating generally results in aseveral degrees kelvin increase in the airstream temperature,with a maximum value of 314 K. In addition, the integratingnephelometer used to measure the aerosol scattering coef-ficients heats the airstream by several degrees kelvin.Therefore, the possibility arises of evaporating a volatileaerosolchemicalspecies, suchas ammonium nitrate, resulting

in a decrease in theaerosol scatteringcoefficient. Theextentof such evaporation andthe resultantdecrease in theaerosolscatteringcoefficientdepend on several factors includingtheresidence time of air in the heated region, the temperature,the relative humidity,and the initial particle size distribution.

The evaporation of ammonium nitrate has been inves-tigated under a variety of laboratory and field conditions.Larson andTaylor (12) studied theevaporation of ammoniumnitrate aerosol solution droplets (RH 60%, T 290 K) bymeasuring the change in the aerosol size distribution as afunction of residence time in a diffusion stripper designed toremove HNO3 and NH3. The observed changes were inagreement with theoretical predictions based on gas phasediffusion and vaporpressures calculatedfor pertinent solutionchemical compositions. Harrison et al.(13) measured aerosol

sizedistributions using an impactor downstream of a denuderdesigned to scrub NH3 and HNO3 from the airstream. Incontrast to the results of Larson and Taylor, evaporation ofammonium nitrate was found to be much slower thanpredicted by theory, suggesting the existence of an unknownkinetic constraint inhibiting mass transport from aerosolparticles. Richardsonand Hightower (14) levitated anhydrousammoniumnitrate crystallinesolid andsupersaturated liquidparticles in an electrodynamic balance at room temperatureandmeasured changes in theparticle radiuswithtime. Theirmeasurements agreed with theoretical predictions based onthermodynamic properties andkinetictheoryfor evaporationof particlesin a vacuum, under theassumption that themass

* Correspondence should be addressed to this author at NOAA,CMDL, 325 Broadway, R/E/CG1, Boulder, CO 80303. Email:[email protected].

National Oceanic and Atmospheric Administration. Brookhaven National Laboratory.

Environ. Sci. Technol. 1997, 31, 2878-2883

2878 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 31, NO. 10, 1997 S0013-936X(97)00089-8 CCC: $14.00 1997 American Chemical Society

BNL-64


2/6

accommodation coefficient was initially 0.02, decreasing to0.004 after 4 h. Indeed, recent modeling resultshave indicatedthat there are several factors that must be considered whenestimating the evaporation of ammonium nitrate based onlab conditions including the aerosol size distribution, anduse of an appropriate evaporation theory (15). In additionto laboratory studies, field studies have indicated that the

volatilization of ammoniumnitrate aerosol maybe significantfor filters and impactors undersome atmosphericconditions(9, 10).

Thefocusof this paper is to study theeffect of ammoniumnitrate aerosolevaporation in a heated nephelometer on theaerosol light scattering coefficient under conditions repre-sentative of field measurements. We focus on measurementandmodelingof theevaporationof ammonium nitrate aerosolat a low relative humidity (10%, in order to dry the aerosol)asafunctionoftemperature(300-320 K), nephelometermeanresidence time, and initial aerosol particle size distribution.We predict the decrease in the aerosol light scatteringcoefficient expected for a variety of fieldsampling conditions.For these estimates, we assume that the particles arecomposed entirely of ammonium nitrate andevaporate with

the maximum theoretical mass transfer rate. These assump-tions yield an upper limit estimate of the decrease in theaerosol light scattering coefficient due to evaporation fortypical field conditions.

Experimental Approach

Figure 1 shows the experimental setup used to measure theevaporation of ammonium nitrate aerosols within an inte-grating nephelometer as a function of temperature andresidence time. An atomizer generated a polydisperseammoniumnitrate aerosol that was dried andpassedthrougha TSI Model 3071 differential mobility analyzer (DMA). Theflow settingsof the DMA were 2.0 L min-1 for thesample flow

and 7.5 L min-1 for the sheath flow which allowed for thegeneration of a narrow aerosol sizedistributionhavingsingletparticles (particles of one elementary unit of charge) of 0.40m diameter. After passing through the DMA, the aerosolstreamwasmixedwithadryairstreamthathadbeenscrubbedto removeNH3 andHNO3 (with packedcolumns of citric acidand sodium bicarbonate), and filtered to remove aerosol

particles. The temperature and RH of the airstream weremeasured directly upstream and downstream of the neph-elometer using Vaisala Model 50Y temperature and relativehumidity sensors. The aerosol then passed through a TSI3563 three-wavelength integrating nephelometer that mea-sured the aerosol scattering and backscattering coefficientsat 450,550, and700 nm (with theaerosolscattering coefficientat 450 nm being reported in this paper). The nephelometerwas maintained at a constant temperature by a processcontroller that supplied power to heating tape wrappedaround the outside of thenephelometercontrolvolume.Afterpassing through the nephelometer, the aerosol was dilutedwith clean air in order to ensure a flow rate of 27 L min-1

through a Climet 6300 Laser Particle Counter (LPC), whichmeasured the diameters of particles greater than 0.3m.The

LPC was calibrated with known size ammonium sulfateaerosol particles generated with the DMA. In addition, thetotal aerosol number concentration of particles greater than0.01 m in diameter was measured with a TSI 3760 Con-densation Nucleus Counter (CNC) at the exit of the nephelom-eter.

Experiments were performed at temperatures from 300 Kto 320 K in increments of 5 K. At each temperature, thenephelometer flow rate was maintained at 5 L min-1, 10 Lmin-1,and18Lmin-1which corresponds to mean residencetimes of 26 s, 13 s, and 7.3 s, respectively. At each tempera-ture and flow rate, the scattering coefficients (sp) weremeasured, as well as the concentrations of particles greater

FIGURE 1. Experimental setup.

VOL. 31, NO. 10, 1997 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 2879


3/6

than 0.01 m in diameter (CNC) and the size distribution ofparticles greater than 0.3 m in diameter (LPC).

To examine whether the nephelometer is well mixed (i.e.,a completely stirred reactor), a polydisperse ammoniumsulfate aerosol was generated with an atomizer and mixedwith clean, dry air. The number concentrationwas measureddirectly downstream of the nephelometer with a CNC, usingan experimental setup similar to Figure 1, but without theDMA. At each flow rate, filtered air from a HEPA capsule

filter internal to the nephelometer was switched into theairstream in order to flush aerosol particles from the controlvolume, and the number concentration of particles at theoutlet was recorded as a function of time. The experimentwas conducted at three flow rates. Figure 2 shows the decayin the normalized particle number concentration (particleconcentration at a given time, Nt, divided by the initialconcentration, Ni) with time. Also shown are the theoreticalcurvesfor thechange in theparticleconcentration with time,assuming the nephelometer is well mixed (with a samplevolume of 2.49 L). The theoretical curves are in generalagreement with the measurements, confirming that for theflow rates usedin this study thenephelometer canbe modeledas a well-mixed system.

In addition to testing the nephelometer mixing assump-tions, possible temperature-dependent loss of particles to

the nephelometer walls was determined as a function ofresidence time using nonvolatile ammonium sulfate aerosolwith the same input aerosol size distribution as that used forthe ammonium nitrate experiments. Results showed thatfor a nonvolatile aerosol the size distribution was similar bothupstream and downstream of the nephelometer and wasunaffected as the nephelometer temperature was increased,demonstrating that wall losses were negligible up to at least320 K. In addition, the aerosol scattering coefficients variedbyless than 5% ateach temperature because thetotalaerosolnumber concentrations and size distributions remainedrelatively constant.

Modeling theEffect ofParticleEvaporationontheLightScatteringCoefficient. The ratioof the scatteringcoefficientof an aerosol in a nephelometer for a given temperature, T,

wavelength, , and residence time, , to that of the initialaerosol scattering coefficient (i.e., scattering coefficient ofthe aerosol size distribution at the nephelometer inlet) isdenoted

where T() is the aerosol scattering coefficient at a given T,and ,and T(0)is thescattering coefficient of theaerosol sizedistribution at the nephelometer inlet. From the measuredsize distribution, the initial aerosol scattering coefficient canbe estimated as

For our calculations, the size distributions are divided into17 size bins (m) 17) having midpoint diameters rangingfrom0.30mto0.64minincrementsof0.02 m. Theaerosolscattering efficiency, KSj, is estimated using the approach ofBohren and Huffmann (16) under the assumption that theparticles are spherical and have a refractive index of 1.47(14),and nTj(0)is thenumber concentration of aerosolparticlesin thebin centered at Dpj at a given temperature forthe initial

particle size distribution.The estimation ofT() for a size distribution of particleswithin the nephelometer (i.e., for > 0) is considerablymorecomplicated than for T(0) since the aerosolsize distributionthat initially enters the nephelometer is heated to a tem-perature, T, which causesthe particlesto decrease in size dueto evaporation. T() can be written as

where each value ofDpj is divided into Nintervals of particleresidencetime ranging from 0.1s to100 s. Inorder toestimatenTj,k(), we assume that the nephelometer control volume iswell mixed, as previously discussed, and that particles within

the nephelometer of initial diameter Dpj have a residencetime distributionthat is exponentially weighted based on themean residence time within the nephelometer. Therefore,the particle number size distribution in the nephelometer isthe sum of the exponentially weighted distributions havinginitialdiameters Dpj. Inorder toestimatenTj,k(),itisnecessaryto estimate thechange in diameter of each initial particle sizeat a given temperature due to evaporation. The change indiameter due to evaporation in the nephelometer at a giventemperature can be solved based on the approach of Fuchsand Sutugin (17) as follows:

where MNH4NO3 and FNH4NO3 are the molecular weight anddensity of ammonium nitrate, Dh is the root mean squarediffusivity of ammonia and nitrate away from the particlesurface as discussed by Wexler and Seinfeld (18), F(Kn) is acorrection for free molecular effects as described by Pandisand Pilinis (19), A corrects for interfacial mass transportlimitations [in this case, equal to 1 since we assume a massaccommodation coefficient of 1 (15)], and Kp is the vaporpressure product of ammonia and nitric acid directly abovethe particle surface (20). For dry ammonium nitrate, thereis a more than 20-fold rise in Kp1/2 from 300 to 320 K,demonstrating the sensitivity of the change in particlediameter with time to temperature (20). It should noted thatabove the deliquescence point(RH)63% for typical ambientconditions) the value of Kp1/2 depends on the RH as well asthe temperature. As the RH increases, Kp1/2decreases, andtherate of evaporationof ammoniumnitrate fromthe particleslows. Theexperiments wereconductedat RHvalues betweenroughly 5% and 15%; we therefore assume that the particlesare dry with no associated water. The possibility cannot beexcluded that the particles are supersaturated with respectto aqueous ammoniumnitrate, andexist in a metastable stateas concentrated ammonium nitrate droplets. This wouldresult in greater solutevapor pressures over theparticlesandhence increased evaporation rates compared tothe dryaerosol(21).

To estimate nTj,k() for an initial size distribution, we firstuse eq 4 to predict the change in particle diameter for each

FIGURE 2. Nt/Ni measured and theory vs time for nephelometermean residence times of26, 13, and 7.3s.

RT() )T()

T(0)(1)

T(0) )

4j)1

m

D2pjKSjnTj(0) (2)

T() )

4j)1

m

k)1

N

D2pj,kKSj,knTj,k() (3)

dDpdt

) (MNH4NO3FNH4NO3RgT

)F(Kn)ADh 1DpKp1/2 (4)

2880 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 31, NO. 10, 1997


4/6

of the m, Dpj size bins. The particles evaporate much morerapidly as thetemperatureis increasedfrom 300to 320 K dueto the dependence ofKp1/2 on T. The particle residence timedistribution function for a given mean residence time is thenestimated based on the cumulative density function [exp(-t/)]. The residence time distribution function gives the

probability of finding a particle that has been in thenephelometerfor a time tat a specified mean residencetime.The residence timedistribution functions are combined withestimates of the change in diameter with time (eq 4) todetermine the size distributions of particles having initialdiameters, Dpj, within the nephelometer. In orderto estimateT() from eq 3 for a given Tand , nTj,k() is determined foreachofthe Nparticle residence times for the17 initialparticlessizes.

ResultsFigure 3 shows the modeled and measured number sizedistributions for mean residence times of 7.3, 13.0, and 26.0

s at 300 and 320 K. There are two distinct modes in thedistributions at 300 K that are associated with singly anddoublycharged particles from the DMA. Both the measuredand modeled size distributions show that for a given meanresidence time the particle size distributions shift towardsmaller particle sizes as the temperature increases from 300to 320K. Thisis due to the sensitivity ofKp1/2 to temperaturein the range from 300 to 320 K.

At 300 K, as the residence time increases, the particle sizedistributions shift toward smaller particles. This is expectedsince the particles have a longer time to evaporate in thenephelometer. Overall the model predicts a shift toward

smaller particle sizes as the residence time and temperatureincrease, as seen in the measurements. At 320 K, there isgenerally a discrepancy between the measured andmodelednumber size distributions. The measurements suggest thatthe particles are evaporating faster than the model predicts.A possible explanation may be that the particles exist assupersaturated droplets, which would have a greater evapo-ration rate. In addition, the measuredsize distributionsoftenhave peaks in the larger mode (Dp 0.5 m) that are notduplicated by the model. It is possible that the observedpeaks occur due to bumps in the Mie scattering efficiency ataround 0.5 m.

Figure 4a,b,ccompares modeled andmeasuredvalues forT()/300() as a function of residence time and temperaturefor a wavelength of 450 nm. The experimental values for

T()/300() are evaluated by dividingthe measuredscatteringfor a mean residence time, , and temperature, T, by thescattering coefficient at 300 K for the same values of . Thedata points for the measured values are the means of fiveexperiments, and the error bars are the standard deviations,which range from 0.03 to 0.08. The model estimates ofT-()/300() are made by taking the ratio of RT() to R300(),which are estimated using eq 1 with the approach discussedin the previous section. The ratio of T()/300() must beestimated since RT() cannot be directly measured in ourexperiments. Figure 4 shows that the modeled T()/300()values generally agree with the measurements. Overall, themodeled T()/300() values are within 10% of the mea-sured values, suggesting that the theory can be used toestimate the fractional decrease in the aerosol scatteringcoefficient of ammonium nitrate in the nephelometer as afunction of temperature and residence time within 10%. Forall three residence times at 315 and 320 K the modeled T-()/300() values slightly exceed the measured values. Thismay be due to a greater evaporation rate as previouslydiscussed.

DiscussionThe results show that the theoretical model employed hereaccurately describes the decrease in the aerosol scatteringcoefficient as a function of aerosol physical properties,nephelometer temperatures, and residence times and there-fore can be used to predict the decrease in the aerosolscattering coefficient of ammonium nitrate under fieldmeasurement conditions. Here we estimate the ratio of thescatteringcoefficient at Tand to theinitial aerosol scattering

coefficient, RT(), for a range of possible field conditions anddiscuss the decrease in the scatteringcoefficientexpectedforthe NOAA aerosol sampling protocols.

Figure 5 shows the inlet configuration for a typical NOAAaerosol monitoring station. Aerosols are drawn into a mainstack at a flow rate of 1000 L min-1, which corresponds to amean residence time in the stack of roughly 7 s. Within thestack, the temperature andRH remain at ambient conditions.The aerosol then enters a section that is heated in order toachieve an RH of 40% . Depending on the ambient tem-perature and RH, the airstream can be heated up to 314 K(which is an upper limit set point value at the NOAAmonitoring stations). Theheatedair thenentersa nephelom-

FIGURE3. Modeledandmeasuredammoniumnitratesizedistribu-tions at 300K and 320K for (a)) 7.3s, (b) ) 13.0s, and (c)) 26.0s.



5/6

eterwhich measuresthe aerosolscatteringand backscatteringcoefficientsat 450, 550, and700 nm. Thenephelometer lamp

heats the airstream by an additional 2-5 K. At the Bondvillestation, the mean temperaturewithin the nephelometer overan entire year (based on hourly averages) is 301 K, withtemperatures exceeding 310 K 5% of the time and temper-atures above 320 K roughly 1%of the time. For the flow ratesemployed, the mean residence times in the heated sectionof the inlet and the nephelometer are 2 s and 4.8 s, respec-

tively. Because the residence time and temperature in thenephelometer are greater than in the heated section ofthe inlet, we assume that the evaporation of particles in theheated inlet section is negligible compared to that in thenephelometer.

In order to estimate the decrease in the aerosol scatteringcoefficientof ammonium nitrate for specific fieldconditions,it is necessary to measurethe initial particle size distributionof the ambient aerosol. Because of evaporation, as discussedpreviously it is difficult to measure the ammonium nitrateaerosol size distributions by traditional methods such asimpactors (9, 10). Therefore, there is not a great deal ofinformation on the size distributions of ammonium nitratereported in the literature. More importantly, there is littleinformation available on the contribution of ammonium

nitrate aerosol to the total aerosol extinction, with theexception of a study in Denver (7), and several studiesin TheNetherlands (8, 22). John et al. (6) measured the mass sizedistributions of ammonium nitrate aerosol in Los Angelesand report data in terms of lognormal moments. Based onthese measurements, we assume a similarlognormal ambientsize distribution at the NOAA sampling sites having ageometric number median diameter (Dp,g) of 0.4 m and ageometric standard deviation (GSD) of 1.6.

Model estimates of the decrease in the aerosol scatteringcoefficient at 450 nm for a given temperature with increasingmean residence time, , are shown in Figure 6. For a givenmean residence time, theratioof thefinalto initial scatteringratio, RT(), decreases with increasingtemperature, reflectingthe Kp1/2 dependence on temperature. As indicated in thefigure, for a mean residence time of 4.8 s typical of NOAAsites, the scattering coefficient decreases by at most 20% at320 K. Again we stress that the estimated RT() values areintended to be upper estimates of the change in the aerosolscattering coefficient due to evaporation, because of theassumption that the aerosol is composed entirely of am-monium nitrate and the assumption of maximum aerosolevaporation rate. There are several reasons that ammoniumnitrate aerosols may take longer to evaporate in the atmo-sphereincluding the existenceof ammonium nitrate solutiondroplets at RH values above the deliquescence point whichresult in lower NH3 and HNO3 vapor pressures above theparticle surfaces, as well as organic surface coatings whichinhibit gas phase transport (18).

FIGURE 4. T()/300() measured and theory vs temperature forammoniumnitrate particles at(a)) 7.3s, (b)) 13.0s, and(c)) 26.0s.

FIGURE 5. Typical NOAA CMDL aerosol inlet design.

FIGURE6. RatioofscatteringcoefficientatTandtoinitialscatteringcoefficient,RT(),vsmeanresidencetime()forammoniumnitratesizedistributionshavingageometric numbermediandiameter(Dp,g)of 0.4mand a GSD of 1.6.

2882 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 31, NO. 10, 1997


6/6

Figure 7 shows the effect of particle size and temperatureon the decrease in the aerosol scattering coefficient. In thefigure, RT() is plottedvs geometricnumbermedian diameter,Dp,g, fora geometricstandarddeviation of 1.6,for temperaturesranging from 300 to 320 K. As seen in the figure, for a given

temperature, RT() decreases as Dp,g decreases. This is duetothe inverse dependence ondiameterof therateof diameterchangeduetoevaporation,asshownineq4. Figure7suggeststhat the maximum decrease in the scattering due to evapora-tion of ammonium nitrate aerosol during field sampling is40% (for Dp,g ) 0.2 m and T) 320 K), although as notedabove it is difficult to verify that this is a realistic value forDp,g for the conditions at the aerosol monitoring stations.Figure8 shows theeffect ofGSD on RT() fora mean residencetime of 4.8 s and an assumed ammonium nitrate particlenumber meandiameter of 0.4m. Theeffect ofGSD is clearlyseen ata temperature of 320K for whichRT() increases from0.58 for a monodisperse aerosol of 0.4 m to 0.90 for a GSD) 2.0. This increase is due to the increase in the number ofparticles having diameters greater than Dp,g as the GSDincreases. As shown in eq 2, the scattering coefficient is afunctionofDp2 aswellasKs. Asthe GSDincreases,the relativenumber of particles greater than Dp,g increases as does Ks forthese particles, which results in an increase in the scatteringcoefficient as well as RT().

In summary, the model estimations ofRT() suggest thatfor a worst-case scenario at NOAA field sampling locations(i.e., T) 320 K, Dp,g ) 0.2 m) the decrease in the aerosolscattering coefficient due to the evaporation of ammoniumnitrate aerosol is roughly 40%. However, based on the few

aerosolsize distributionmeasurementsavailable, the decreasein the aerosol scattering coefficient is generally expected tobe less than 20%.

AcknowledgmentsWe thank J. Wendell at NOAA for his technical expertise infabricating the control systems used in the experiments, andA. Wexler at the University of Delaware for advice on themodeling. This research was supported in part by anappointmentto theGlobal Change DistinguishedPostdoctoralFellowships sponsored by the U.S. Department of Energy,Office of Health and Environmental Research, and admin-

isteredby theOak Ridge Institutefor Science andEducation.Work at BNL was supplied by the Environmental SciencesDivision of the U.S. Department of Energy (DOE) as part ofthe Atmospheric Radiation Measurement Program and wasperformed under the auspices of DOE Contract DE-AC02-76CH00016.

Literature Cited(1) Beuttell, R. G.; Brewer, A. W. J. Sci. Instrum. Phys. Ind. 1949, 26,

357-359.(2) Anderson, T. L.; Covert, D. S.; Marshall, S. F.; Laucks, M. L.;

Charlson, R. J.; Waggoner, R. J.; Ogren, J. A.; Caldow, R.; Holm,R. L.; Quant, F. R.; Sem, G. J.; Wiedensohler, A.; Ahlquist, N. A.;Bates, T. S. J. Atmos. Ocean. Technol. 1996 13, 967-986.

(3) Charlson, R. J.; Schwartz,S. E.;Hales, J. M.; Cess, R. D.; Coakley,J. A., Jr.; Hansen, J. E.; Hoffmann, D. J. Science1992, 225, 423-430.

(4) Tang, I. N.; Munkelwitz, H. R. Atmos. Environ. 1993, 27A, 467-473.

(5) Wall,S. M.; John,W.;Ondo,J. L.Atmos. Environ.1988, 22, 1649-1656.

(6) John,W.; Wall,S. M.;Ondo, J.L.; Winklmayr, W.Atmos. Environ.1990, 24A, 2349-2359.

(7) Groblicki, P.J.; Wolff,G. T.;Countess,R. J.Atmos. Environ.1981,15, 2473-2484.

(8) Ten Brink, H. M.; Veefkind, J. P.; Waijers-Ijpelaan, A.; van derHage, J. C. H. Atmos. Environ. 1996, 30, 4251-4261.

(9) Zhang, X.; McMurry, P. H. Atmos. Environ. 1992, 26A, 3305-3312.

(10) Appel,B. R.;Tokiwa,Y.;Haik,M.Atmos. Environ.1981, 15,283-289.

(11) Ogren, J. A.; Anthony, S.; Barnes, J.; Bergin, M. H.; Huang, W.;McInnes, L. M.; Myers, C.; Sheridan, P.; Thaxton, S.; Wendell,J. NOAA/Climate Monitoring and Diagnostics Laboratory Sum-mary Report 1994-19951996, 50-60.

(12) Larson, T.V.; Taylor, S. G.Atmos. Environ. 1983, 17, 2489-2495.(13) Harrison, R. M.; Sturges, W. T.; Kitto, A.-M. N.; Li, Y. Atmos.

Environ. 1990, 24A, -1888.(14) Richardson, C. B.; Hightower, R. L. Atmos. Environ. 1987, 21,

971-975.(15) Bai, H.; Lu, C.; Ling, Y. M. Atmos. Environ. 1995, 29, 313-321.(16) Bohren, C. F.; Huffman, D. R. Absorption and scattering of light

by small particles; Wiley and Sons: New York, 1983.(17) Fuchs, N. A.; Sutugin, A. G. International Reviews of Aerosol

Physics and Chemistry; Hidy, M., Brock, J. R., Eds.; Pergamon:New York, 1971; pp 1-60.

(18) Wexler, A. S.; Seinfeld, J. H. Atmos. Environ. 1990, 24A, 1231-1246.

(19) Pandis, S. M.; Pilinis, C. In The Handbook of EnvironmentalChemistry;Huntzinger, C.,Ed.; SpringerVerlag: Heidelberg, 1995;pp 35-68.

(20) Mozurkevich, M. Atmos. Environ. 1993, 27A, 261-270.(21) Clegg, S. L.; Brimblecombe, P. J. Aerosol Sci. 1995, 26, 19-38.(22) Veefkind, J. P.; van der Hage, J. C. H.; Ten Brink, H. M. Atmos.

Res. 1996, 41, 217-228.

Received for review February 3, 1997. Revised manuscriptreceived May 29, 1997. Accepted June 9, 1997.X

ES970089H

X Abstract published in Advance ACS Abstracts, August 1, 1997.

FIGURE7. RT()vsDp,gforanammoniumnitrateaerosol withaGSD) 1.6and) 4.8s.

FIGURE8. RT()vsTforanammoniumnitrate aerosol withDp,g)0.4m,) 4.8s,andGSDs rangingfrom1.0(monodisperse)to2.0.


Evaporation of Ammonium Nitrate

Documents

Transcript of Evaporation of Ammonium Nitrate