Open Collections

UBC Faculty Research and Publications

Ice nucleation on mineral dust particles: onset conditions, nucleation rates and contact angles. Eastwood, Michael L.; Cremel, Sebastien; Gehrke, Clemens; Girard, Eric; Bertram, Allan K. 2008-11-30

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


Bertram_2008JD010639.pdf [ 391.73kB ]
JSON: 1.0041868.json
JSON-LD: 1.0041868+ld.json
RDF/XML (Pretty): 1.0041868.xml
RDF/JSON: 1.0041868+rdf.json
Turtle: 1.0041868+rdf-turtle.txt
N-Triples: 1.0041868+rdf-ntriples.txt
Original Record: 1.0041868 +original-record.json
Full Text

Full Text

Ice nucleation on mineral dust particles: Onset conditions, nucleationrates and contact anglesMichael L. Eastwood,1Sebastien Cremel,1Clemens Gehrke,1Eric Girard,2and Allan K. Bertram1Received 20 June 2008; revised 4 September 2008; accepted 18 September 2008; published 20 November 2008.[1] An optical microscope coupled to a flow cell was used to investigate the onsetconditions for ice nucleation on five atmospherically relevant minerals at temperaturesranging from 233 to 246 K. Here we define the onset conditions as the humidity andtemperature at which the first ice nucleation event was observed. Kaolinite andmuscovite were found to be efficient ice nuclei in the deposition mode, requiring relativehumidities with respect to ice (RHi) below 112% in order to initiate ice crystalformation. Quartz and calcite, by contrast, were poor ice nuclei, requiring relativehumidities close to water saturation before ice crystals would form. Montmorilloniteparticles were efficient ice nuclei at temperatures below 241 K but were poor ice nuclei athigher temperatures. In several cases, there was a lack of quantitative agreement betweenour data and previously published work. This can be explained by several factorsincluding the mineral source, the particle sizes, the surface area available for nucleation,and observation time. Heterogeneous nucleation rates (Jhet) were calculated from themeasurements of the onset conditions (temperature and RHi) required from ice nucleation.The Jhetvalues were then used to calculate contact angles (q) between the mineralsubstrates and an ice embryo using classical nucleation theory. The contact anglesmeasured for kaolinite and muscovite ranged from 6C176 to 12C176, whereas for quartz andcalcite, the contact angles ranged from 25C176 to 27C176. The reported Jhetand q values mayallow for a more direct comparison between laboratory studies and can be used whenmodeling ice cloud formation in the atmosphere.Citation: Eastwood, M. L., S. Cremel, C. Gehrke, E. Girard, and A. K. Bertram (2008), Ice nucleation on mineral dust particles:Onset conditions, nucleation rates and contact angles, J. Geophys. Res., 113, D22203, doi:10.1029/2008JD010639.1. Introduction[2] Ice nucleation can occur in the atmosphere by eitherhomogeneous nucleation or heterogeneous nucleation. Het-erogeneous nucleation typically involves solid substrates,which are often called ice nuclei (IN). These ice nuclei havethe potential to modify climate by changing the formationconditions and properties of ice and mixed-phase clouds.[Ka¨rcher, 2004; Haag and Ka¨rcher, 2004; Lohmann andFeichter, 2005; Lohmann and Diehl, 2006]. Our lack ofknowledge on ice nucleation is a major obstacle for thesimulation of the complex interactions between aerosols andcold clouds. It has been shown in various investigations thatmixed-phase clouds often cannot be properly simulated withthe existing IN parameterizations, such as those of Fletcher[1962] and Meyers et al. [1992] [Girard and Curry, 2001;Girard et al., 2005]. More physically based parameteriza-tions are needed to simulate heterogeneous nucleation inclimate models.[3] Mineral dust particles are abundant in the atmosphere,and both laboratory [Pruppacher and Klett, 1997; Baileyand Hallett, 2002; DeMott, 2002; Zuberi et al., 2002; Hunget al., 2003; Archuleta et al., 2005; Mangold et al., 2005;Mo¨hler et al., 2006; Field et al., 2006; Kanji and Abbatt,2006; Knopf and Koop, 2006; Marcolli et al., 2007;Zimmerman et al., 2007] and field measurements [Sassen,2002, 2005; DeMott et al., 2003a, 2003b; Sassen et al.,2003; Toon, 2003; Cziczo et al., 2004; Twohy and Poellot,2005; Kanji and Abbatt, 2006] have shown that mineraldust particles are effective ice nuclei. Laboratory data haveshown that mineral dust particles can lower the supersatu-rations required for ice formation compared to homoge-neous nucleation. At the same time, field measurementshave shown that mineral dust particles can have a signifi-cant impact on cloud formation, cloud properties andprecipitation [Sassen, 2002, 2005; DeMott et al., 2003a;Sassen et al., 2003]. Measurements have also shown thatthe cores of ice crystals often contain mineral dust particles,suggesting that ice nucleation is often initiated by mineraldust aerosols in the atmosphere [Heintzenberg et al., 1996;Cziczo et al., 2004; Twohy and Poellot, 2005]. Furthermore,JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 113, D22203, doi:10.1029/2008JD010639, 2008ClickHereforFullArticle1Department of Chemistry, University of British Columbia, Vancouver,British Columbia, Canada.2Department of Earth and Atmospheric Sciences, University of Quebecat Montreal, Montreal, Quebec, Canada.Copyright 2008 by the American Geophysical Union.0148-0227/08/2008JD010639$09.00D22203 1of9measurements of the chemical composition of ice nuclei inthe atmosphere show that mineral dust is composed of asignificant fraction of atmospheric IN [Chen et al., 1998;Rogers et al., 2001; DeMott et al., 2003b; Richardson et al.,2007].[4] Although there have been numerous studies on the icenucleating properties of mineral dust particles, more work isstill needed for a complete understanding of the ice nucle-ation properties of these particles [Vali, 1996; Martin, 2000;Demott, 2002; Cantrell and Heymsfield, 2005]. For exam-ple, only a few of the most abundant types of mineralsfound in the atmosphere have been studied in detail. Inaddition, in many of the previous studies, only the onsetconditions for ice nucleation were reported and only a fewstudies considered ice nucleation rates (a key parameter fordescribing ice nucleation). Measurements of ice nucleationrates are needed to more accurately compare laboratoryresults and to extrapolate laboratory data to the atmosphereand for climate modeling.[5] Our studies focus on the ice nucleating properties ofmineral dust particles at temperatures between 247 and 233K, a temperature range relevant for the lower troposphere.As pointed out by Vali [1996] in a review on ice nucleation,the origin of ice in lower tropospheric clouds is notresolved, and it remains a question of great importanceand in need of new efforts. Also, there have been very fewmeasurements in the range of 247 to 233 K. In the followingstudy we focus on the ice nucleation properties of muscovite(a mineral in the mica group), kaolinite, montmorillonite,quartz and calcite particles. These minerals were chosensince they are major components of aerosolized mineraldust found in the atmosphere. To illustrate this point, themineralogies of Saharan dust collected in Sal Island, Bar-bados and Miami after three large Saharan dust outbreaksare shown in Table 1 [Glaccum and Prospero, 1980]. Theminerals chosen for our studies represent over 90% of thetotal mass during these outbreaks. Montmorillonite compo-sition was below the detection limit (5%) in the Saharandust studies illustrated in Table 1, but several studies havecited montmorillonite as one of the dominant clay mineralspresent in African and Asian dusts [Prospero, 1999;Hanisch and Crowley, 2001].[6] This paper is organized as follows: first, we presentmeasurements of the onset conditions for ice nucleation onmuscovite, kaolinite, montmorillonite, quartz and calcite.Here we define the onset conditions as the humidity andtemperature at which the first ice nucleation event wasobserved. Second, we compare this data with existing datain the literature. Third, from the onset conditions wedetermine heterogeneous ice nucleation rates (number ofnucleation events per unit surface area of solid material perunit time) for each mineral type. Fourth, we parameterizedthe heterogeneous nucleation rates using classical nucle-ation theory. Classical nucleation theory is a reasonablestarting point for analyzing laboratory data, as this theoryhas been used in the past for describing ice nucleation inatmospheric models [Ka¨rcher, 1996, 1998; Jensen andToon, 1997; Jensen et al., 1998; Ka¨rcher et al., 1998;Martin, 2000; Demott, 2002; Morrison et al., 2005]. In thislast step we determined the contact angle between an icenucleus and the mineral surfaces.2. Experimental[7] Micron-sized muscovite particles were generated bygrinding muscovite flakes (1–10 mm in diameter, <1 mm inthickness, purchased from Alfa Aesar), and micron-sizedquartz particles were produced by grinding a sample ofquartz tubing (purchased from United Silica). Kaolinite(purchased from Fluka), montmorillonite K10 (purchasedfrom Fluka) and calcite (purchased from Puratronic1, AlfaAesar) particles were used as supplied. Table 2 lists thechemical formulae of the minerals studied.[8] The apparatus used in these studies consisted of anoptical microscope coupled to a flow cell (see Figure 1) inwhich the humidity and temperature could be accuratelycontrolled [Dymarska et al., 2006; Parsons et al., 2004a,2004b]. Mineral dust particles were deposited on the bottomsurface of the flow cell; the relative humidity with respect toice (RHi) inside the cell was increased, and the conditionsunder which ice crystals formed were determined with areflected-light microscope (Zeiss Axiotech 100) equippedwith a 10C2 objective lens. From these measurements wedetermine the onset conditions (temperature and relativehumidity) for ice nucleation.Table 1. Mineralogy of Saharan Dust Collected at Sal Island (inthe Cape Verde Islands), Barabados, and Miami, Florida, AfterThree Large Saharan Dust Outbreaks [Glaccum and Prospero,1980]aMineralSal Island(%)Barbados(%)Miami(%)Mica 53.8 64.3 62.1Kaolinite 6.6 8.3 7.1Chlorite 4.3 4.1 4.2Quartz 19.6 13.8 14.2Microcline 2.2 1.5 1.1Plagioclase 5.4 4.1 4.5Calcite 8.2 3.9 6.9Montmorillonite C205 C205 C205aPercentages shown are averages of the three outbreaks (assuming 100%crystalline material).Table 2. Chemical Formulae of the Five Minerals Studied[Anthony et al., 1995]Mineral FormulaMuscovite KAl2(Si3Al)O10(OH,F)2Kaolinite Al4Si4O10(OH)8Quartz SiO2Calcite CaCO3Montmorillonite (Na,Ca)0.3(Al,Mg)2Si4O10(OH)2nH2O Figure 1. Schematic of the flow cell used for this study.D22203 EASTWOOD ET AL.: ICE NUCLEATION ON MINERAL DUST PARTICLES2of9D22203[9] The temperature of the cooling stage and hence theflow cell was regulated with a refrigerating circulator(Thermo Neslab ULT-95). A hydrophobic slide (whichsupported the particles) was positioned inside the aluminumcell body. An insulating spacer, made from polychloro-trifluoroethylene (PCTFE), was placed between the hydro-phobic glass slide and the flow cell body. This ensured thatthe coldest portion of the flow cell was the glass substrate(by C2410 K), thus preventing unwanted ice nucleation onother components of the cell. All seals within the cell weremade with Viton O-rings with the exception of the sealbetween the glass slide and the PCTFE spacer, which wasmade with low vapor pressure chlorotrifluoroethylenegrease (Series 28LT from Halocarbon Products, vaporpressure < 0.1 Torr at 293 K). The grease ensured that nospace remained between the glass slide and the PCTFEspacer, where ice could nucleate without being detected bythe microscope. For kaolinite and montmorillonite we alsoperformed experiments using a lower vapor pressure grease(Krytox LVP from DuPont, vapor pressure <10C013Torr at293 K) and the same results were obtained. Also, forkaolinite we previously carried out experiments withoutgrease and the same results were obtained, suggesting thegrease had little effect on the freezing results [Dymarska etal., 2006].[10] The upper portion of the cell body and the inlet andoutlet were made from stainless steel. A sapphire window (1mm thick) was positioned at the top of the cell body,allowing optical access to the bottom surface of the cell.The reflected-light microscope was coupled to a high-resolution monochrome digital video camera (Sony, XCD-X700) which captured images of the particles deposited onthe hydrophobic slide during the course of the experiments.The images were analyzed with the Northern Eclipsesoftware package to determine particle size and total surfacearea available for ice nucleation.[11] The bottom surface of the flow cell was a hydropho-bic slide made from a glass cover slide treated withdichlorodimethylsilane (DCDMS). This hydrophobic layerwas added to reduce the probability of ice nucleationdirectly on the surface. Prior to the treatment with DCDMSthe glass slide was thoroughly cleaned in ‘‘piranha’’solution(3:1 mixture by volume of sulfuric acid and hydrogenperoxide), rinsed in high-purity water (distilled water furtherpurified with a Millipore system) and methanol (HPLCgrade). Any remaining contaminant particles removed witha dry ice cleaning system (Sno Gun-IITM, Va-Tran Sys-tems). The treatment with DCDMS involved placing theslides in an airtight chamber together with 2–3 droplets ofDCDMS solution (Fluka, 5% DCMS in heptane). The slideswere not in direct contact with the DCDMS, rather theDCDMS coated the glass slides via vapor deposition.[12] All samples were prepared and the flow cell con-structed within a filtered air laminar flow hood. This greatlyreduced the possibility of sample contamination by ambientatmospheric and laboratory particles. All mineral dustparticles with the exception of calcite were deposited on ahydrophobic glass slide using the following technique: thedry dust particulates were placed in a glass vessel immersedin an ultrasonic bath. A flow of ultra-high-purity N2waspassed through the glass vessel, and vibrations from theultrasonic bath caused the dust particles to be suspended inthe flow of N2. This flow was then directed at the hydro-phobic glass slide, and the dust particles were deposited onthe slide by impaction. Calcite particles were not readilysuspended by the vibrations from the ultrasonic bath, sothese were deposited on the hydrophobic slide simply bysprinkling them directly on the slide using a small spatula.In all cases, dust particles deposited on the slide were lessthan 50 mm in diameter. The optical resolution limit of themicroscope was C241 mm. A typical sample held between 100to 1000 individual particles, the majority of which werebetween 1 and 20 mm in diameter. The average sizes of theparticles used in our experiments were 7.7 mm for kaolinite,9.0 mm for muscovite, 8.1 mm for montmorillonite, 10.0 mmfor quartz and 14.2 mm for calcite based on the opticalmicroscope images.[13] During the ice nucleation experiments, a flow ofhumidified He gas was introduced to one side of the cell andexited on the other where its frost point was measured witha frost point hygrometer (General Eastern). From the frostpoint measurements, the water vapor pressure (pH2O) wascalculated using the parameterization of Murphy and Koop[2005]. A flow of humidified gas was generated by passinga flow of ultra-high-purity He gas (99.999 %) over areservoir of ultra-pure water (distilled water further purifiedusing a Millipore system). The desired pH2Owas adjustedby altering the temperature of the water reservoir anddiluting the humidified flow with a second flow of dryHe. A continuous flow of between 1900 to 2100 cm3minC01(at 273.15 K and 1 atm) was maintained throughout thecourse of the ice nucleation experiments. For purification,the He gas used in these experiments was first passedthrough a trap containing molecular sieve (1/16’’ pellets,Type T4A) at 77 K and then through a 0.02 mm filter(Anodisc 25).[14] In our experimental apparatus, a Pt-100 resistancetemperature detector (RTD) was located just beneath theslide containing the mineral dust particles. The RTD wascalibrated against the dew point or ice frost point withinthe cell, similar to methods used by other researchers[Middlebrook et al., 1993; Parsons et al., 2004b; Dymarskaet al., 2006]. Calibration involved observing the change insize of ice crystals on the slide as the temperature wasramped up and down. The temperature at which the sizeof the ice crystals remained constant was determined fromthese measurements; at that point the ice crystals were inequilibrium with the water vapor inside the cell, whosefrost point was precisely known from the hygrometermeasurements. Hence it was possible to determine theoffset temperature between the temperature reported bythe RTD and the temperature of the ice crystals formedon the hydrophobic glass slide in the cell, and use thisoffset to correct experimentally measured temperaturesobtained with the RTD.[15] The RHiwithin the cell was then calculated usingequation (1):RHi¼ pH2O=piceðTcellÞ * 100 ð1ÞWhere pice(Tcell) is the saturation vapor pressure of ice at thetemperature of the cell, calculated using the parameteriza-tion of Murphy and Koop [2005].[16] In each ice nucleation experiment, the RHiwasramped from below 100% to water saturation by decreasingD22203 EASTWOOD ET AL.: ICE NUCLEATION ON MINERAL DUST PARTICLES3of9D22203the temperature of the cell at approximately 0.1 K minC01(corresponding to a change in RHiof approximately 1%minC01) while maintaining a constant pH2Oinside the cell.Typical experimental RHitrajectories are illustrated inFigure 2 for three different initial temperatures of 246.7 K,241.7 K, and 236.7 K. Images of the dust particles wererecorded digitally every 20 seconds or C240.033 K, whilesimultaneously recording pH2Oand the cell temperature.From these images, the RHiand temperature at which icecrystals first formed was determined (i.e., the onset con-ditions of ice nucleation). Shown in Figure 3 are images ofthe kaolinite particles recorded in a typical freezing exper-iment before and after ice nucleation, respectively. Theformation of ice crystals is clearly discernable.3. Results and Discussion3.1. Measurements of Onset Conditions of IceNucleation[17] As mentioned above, the onset conditions weredetermined for muscovite, kaolinite, montmorillonite,quartz and calcite. The total surface area of mineral dustexposed in any particular experiment ranged from 5 x 10C05to5x10C03cm2, based on the geometric surface area of theparticles. Shown in Figure 4 are the onset conditions for allfive minerals at each of the three temperatures studied. Theerror bars represent 95% confidence intervals based on atleast six measurements per data point. For kaolinite, mus-covite and montmorillonite ice nucleation occurred belowliquid water saturation and there was no indication of liquidwater condensing prior to ice nucleation as expected. This isalso the case for quartz and calcite at the two lowesttemperatures. For these experiments the data correspondto deposition freezing, which occurs when vapor absorbsonto a solid surface and is transformed into ice [Valie,1985]. For quartz and calcite at the warmest temperaturestudied the data overlap with liquid water saturation. Forapproximately half of the quartz and calcite experiments atthe warmest temperature we observed only ice nucleationwith no indication of liquid water condensation prior to icenucleation. For the other half of the quartz and calciteexperiments at the warmest temperature we first observedthe condensation of liquid water. In this case it appears thatcondensation freezing may be important. Condensationfreezing refers to the sequence of events whereby liquidwater first condenses followed by freezing of the liquid[Valie, 1985].[18] On the basis of Figure 4, kaolinite and muscovite areeffective ice nuclei at all temperatures studied, with onsetRHivalues below 112%. Quartz and calcite were poor icenuclei at all temperatures studied, requiring relative humid-ities close to water saturation before ice nucleation occurred.Montmorillonite was an effective ice nucleus at the twolowest temperatures studied (236.0 and 240.8 K), but arelatively poor ice nucleus at the highest temperaturestudied (244.6 K). Overall, the data show significant differ-ences in the ice nucleating abilities of the five mineralsstudied over this temperature range.[19] The reason why some minerals are better ice nucleithan others is still relatively poorly understand. However,previous research suggests that it is likely a combination ofthe strengths of the chemical bonds at the mineral surface,the crystallographic match between the substrate and iceFigure 2. Typical experimental trajectories of RHi, wheretemperature was reduced at a rate of 0.1 K minC01, while thewater partial pressure was constant. The trajectories werecalculated using the saturation vapor pressures of water andice from the parameterizations of Murphy and Koop [2005].Figure 3. Optical microscope images of kaolinite particles(a) before and (b) after ice nucleation. Figure 4. Onset conditions for all minerals studied.D22203 EASTWOOD ET AL.: ICE NUCLEATION ON MINERAL DUST PARTICLES4of9D22203embryo, and the presence of active sites on the mineralsurface, which can promote ice nucleation [Pruppacher andKlett, 1997]. For example, it has been speculated that therelatively good ice nucleation ability of kaolinite may bedue to the pseudo-hexagonal arrangement of the hydroxyl(C0OH) groups at the kaolinite surface [Pruppacher andKlett, 1997]. Our laboratory results provide further data thatcan be used to test these various theories.[20] In our experiments, the mineral particles are sus-pended on a hydrophobic glass substrate. Before we discussour results further, we first address the possible effect of thehydrophobic glass substrate on the ice nucleation results.[21] First, in blank experiments (when no mineral dustwas used) ice nucleation did not occur below liquid watersaturation, (see Figure 4a in the study of Dymarska et al.[2006]). Second, from direct observations of the opticalimages we confirmed that ice nucleation always occurred ona mineral particle, rather than a bare spot on the hydropho-bic glass substrate. Third, to further ensure that the hydro-phobic glass support was not influencing our results, we didsome tests with other types of supports. Tests were carriedout with bare glass slides that were not coated with thehydrophobic monolayer. In this case the glass slides werejust cleaned as described above, resulting in a hydrophilicsubstrate. Also, tests were carried out using thin Teflonsheets as the bottom surface of the flow cell. In the testexperiments we measured the onset conditions for icenucleation on kaolinite particles using the different supports(bare glass, hydrophobic glass, and Teflon). Within theuncertainty of our measurements, the results were indepen-dent of the type of support used. This gives further confi-dence that the hydrophobic glass support is not influencingour results.[22] Below we compare our onset conditions with resultspublished in the literature. The comparison below focusesmainly on results from six different studies: those ofRoberts and Hallett [1968], Bailey and Hallett [2002],Dymarska et al. [2006], Kanji and Abbatt [2006], Salamet al. [2006] and Zimmerman et al. [2007]. Some of theexperimental conditions from these measurements are listedin Table 3 for reference.3.2. Comparison of Measured Onset Conditions forKaolinite With Literature Data[23] In Figure 5, we compare our kaolinite results withprevious measurements. At our temperature range, therehave been three previous studies: those of Dymarska et al.[2006], Bailey and Hallett [2002], and Salam et al. [2006].The data from Dymarska et al. [2006] are in agreement withour measurements, which is not surprising since the sameinstrument and experimental protocol were employed. Theonset conditions for Salam et al. [2006] are also in agree-ment with our current studies at 233 to 246 K. Theseauthors observed that 0.5% of the kaolinite particles acti-vated as ice nuclei even at the lowest supersaturations withrespect to ice (close to 100%). The fraction of particlesactivated remained almost constant in these studies untilabove 120–130 % RHiat which point the fraction activatedincreased sharply. The results from Bailey and Hallett[2002] differ from our results by approximately 10–15%RHi. These differences may be due to variation in mineralsurface area available in the different experiments, variationin particle diameters, or variability due to different exper-imental techniques or different observation times in theexperiments. Also, differences in the source of the kaoliniteparticles may result in some variability in the interaction ofthe particles with water. Hoffer [1961] and Schuttlefield etal. [2007] reported that water uptake by kaolinite andmontmorillonite varied significantly with the location ofthe mineral source. These points should be investigated inmore detail in future experiments.Table 3. Summary of Experimental Conditions for Previously Published ResultsStudySizeRange (mm)Numberof ParticlesSurface AreaRange (cm2)Current study 1–50 102–1033 C2 10C06–1C2 10C02Roberts and Hallett [1968] 0.5–3 101–1048 C2 10C08–3C2 10C03Bailey and Hallett [2002] 5–10 Not determined Not determinedKanji and Abbatt [2006] 0.5–5 4.6 C2 102–3.2C2 1049 C2 10C05–3C2 10C02Dymarska et al. [2006] 1–20 2 C2 102–8C2 1026 C2 10C06–1C2 10C02Salam et al. [2006] < 0.5–5 Not determinedaNot determinedaZimmerman et al. [2007] 1–10 Not determinedbNot determinedbaThese authors used a continuous flow diffusion chamber. Typical total aerosol number concentrations were <15 particlescmC03.bZimmerman et al. [2007] spread particles on a silicon plate (5 C2 5 mm), and the particle density was 100–150 particles/mm2. However, the exact fraction of particles activated could not be determined precisely because only a small part of the totalsilicon plate was imaged.Figure 5. Summary of ice nucleation results for kaolinite.D22203 EASTWOOD ET AL.: ICE NUCLEATION ON MINERAL DUST PARTICLES5of9D22203[24] If we consider all the data shown in Figure 5, thefollowing conclusions seem appropriate. At temperaturesabove 255 K, ice nucleation does not occur until liquidwater saturation is reached. At temperatures below 250 K,all data suggest that kaolinite is an effective ice nucleus (i.e.,RHivalues less than water saturation are required for icenucleation). Quantitatively, however, there are relativelylarge differences between the different experiments at tem-peratures below 250 K as mentioned above.3.3. Comparison of Measured Onset Conditions forMontmorillonite With Literature Data[25] In Figure 6 we compare our ice nucleation data formontmorillonite with previous data from Kanji and Abbatt[2006], Salam et al. [2006] and Zimmerman et al. [2007].Our results are consistent with the more recent study byZimmerman et al. [2007] if we extrapolate our results towarmer temperatures, but appear to be inconsistent with theresults from Kanji and Abbatt [2006]. At the warmertemperatures our results also appear to be inconsistent withthe results from Salam et al. [2006]. These authors foundthat 1.2% of montmorillonite particles were active ice nucleievenatverysmallsupersaturationswithrespecttoice(RHiC25100%) over the temperature range of 258 K and 233 K. Thefraction of particles activated remained almost constant untilabove approximately 107–115 % RHi, at which point thefraction activated, increased sharply.[26] Not shown in Figure 6 are the results from Robertsand Hallett [1968]. Roberts and Hallett [1968] observedice nucleation at 248 K and at liquid water saturationwhen using approximately 104particles. This result isconsistent with our observations.[27] Overall, there appear to be large differences betweensome of the montmorillonite studies. Some possible reasonsfor these differences are discussed above. Kanji and Abbatt[2006] also presented other reasons why their results maydiffer from other measurements. First, they suggested a timedependence may be causing the observed differences, withlonger exposure times leading to lower RHivalues requiredfor activation. Second, they suggested that some of thedifferences may be due to different preparation techniques.Kanji and Abbatt [2006] prepared their particles by nebu-lizing an aqueous suspension of the mineral dust. Incontrast, all other studies mentioned above for montmoril-lonite used dry dispersion. Future studies that investigate theeffect of these different parameters are needed for a com-plete understanding of ice nucleation on mineral particles.3.4. Comparison of Measured Onset Conditions forMuscovite, Quartz, and Calcite With Literature Data[28] To our knowledge, these are the first measurementsof the onset conditions (both RHiand temperature) for icenucleation on muscovite. Mason and Maybank [1958]reported that muscovite was inactive as an ice nucleus attemperatures above 255 K, which does not contradict ourresults.[29] For quartz, there are no previous reports of the onsetconditions for ice nucleation. Mason and Maybank [1958]reported that this mineral was inactive as an ice nucleusabove 248 K, which is consistent with our data.[30] For calcite, Roberts and Hallett [1968] reported onetemperature (255 K) and RHi(120%) at which one particlein 104nucleated ice. These conditions correspond to liquidwater saturation. At 244 K, the warmest temperature weinvestigated, we also observed that relative humidities closeto liquid water saturation were needed for ice nucleation.We conclude that the limited results from Roberts andHallett [1968] do not contradict our measurements.3.5. Nucleation Rates, Jhet[31] Above we reported onset conditions (RHiand tem-perature), which may depend on several experimentalparameters, such as observation time and surface areaavailable for nucleation. A more useful parameter fordescribing ice nucleation is the heterogeneous nucleationrate, Jhet, which allows for a more direct comparisonbetween laboratory studies and for extrapolation to theatmosphere. Jhetis defined as the number of nucleationevents per unit surface area of solid material per unit time.Note that Jhetis referred to as both a rate [Pruppacher andKlett, 1997; Martin, 2000; Hung et al., 2003; Parsons et al.,2004a, 2004b; Archuleta et al., 2005; Pant et al., 2006] anda rate coefficient [Dymarska et al., 2006; Marcolli et al.,2007] in the literature. The heterogeneous nucleation rate isrelated to the onset data through equation (2):Jhet¼ w=Ast ð2Þwhere w is the number of ice crystals nucleated, Asis thetotal mineral dust surface area available for heterogeneousnucleation, and t is the observation time. At the onset of icenucleation, w was equal to one.[32] Table 4 lists the nucleation rates determined in ourexperiments. The uncertainty in Jhetwas determined byconsidering the uncertainties in Asand t. We used 10 s forthe observation time with an upper limit of 20 s (the timebetween image captures) and a lower limit of 1 s. Note,however, that nucleation may have happened at a shortertime than 1 second. If this is the case the calculatednucleation rates will be lower limits to the true nucleationrates. For the surface area available for nucleation we usedthe geometric surface area of the particles determineddirectly from the optical microscope images using digitalFigure 6. Summary of ice nucleation results for mon-tmorillonite.D22203 EASTWOOD ET AL.: ICE NUCLEATION ON MINERAL DUST PARTICLES6of9D22203image software (Northern Eclipse). For an upper limit to thesurface area available for nucleation we multiplied thegeometric surface area of the particles by a factor of 50,based on scanning electron microscope (SEM) measure-ments where we more accurately determined the specificsurface area of a limited number of particles. The datashown in Table 4 suggests that our experiments are typicallysensitive to values of Jhetranging from 1 to 22000 cmC02sC01.3.6. Classical Nucleation Theory Parameters From Jhet[33] The applicability of standard classical nucleationtheory to heterogeneous nucleation on minerals remains tobe determined. In fact, some measurements show that forprecise predictions, active site theory is required. See forexample the studies of Hung et al. [2003], Archuleta et al.[2005], and Marcolli et al. [2007]. Nevertheless, classicalnucleation theory has been used in the past to describeheterogeneous nucleation in atmospheric cloud models[Ka¨rcher, 1996, 1998; Jensen and Toon, 1997; Jensen etal., 1998; Ka¨rcher et al., 1998; Demott, 2002; Morrison etal., 2005]. Also classical nucleation theory is a relativelyconvenient and simple way to parameterize laboratory data.Hence classical nucleation theory is a reasonable startingpoint for analyzing our experimental data. Below weanalyze the nucleation rates using classical nucleationtheory. From this analysis, we determined the contact anglebetween an ice nucleus and the mineral surface.[34] For this analysis we only consider the case ofdeposition freezing. As a result, we do not consider anynucleation data where the onset conditions overlap theliquid water saturation line, since in this case nucleationmay have occurred by either deposition or condensationfreezing, as discussed above.[35] According to standard classical nucleation theory, therate of heterogeneous nucleation (Jhet) by deposition freez-ing is defined as: [Pruppacher and Klett, 1997]Jhet¼ A C1 expC0DFg;hetkTð3Þwhere A is the pre-exponential factor in units of cmC02sC01,DFg,hetis the free energy of formation of the critical embryoin joules (J), k is the Boltzmann constant in J KC01, and T isthe temperature in K. Assuming that an ice embryo on acurved solid substrate can be described as a spherical capmodel, the free energy of formation of the critical embryo isgiven by [Pruppacher and Klett, 1997]:DFg;het¼16pM2ws3i=v3½RTrInSiC138C1 fðm;xÞð4Þwhere Mwis the molecular weight of water in g molC01, si/vis the surface tension at the ice-vapor interface in mJ mC02,Ris the universal gas constant in J molC01KC01, r is the densityof ice in g cmC03,Siis the supersaturation ratio with respectto an ice surface, f(m,x) is the geometric factor, m is thecompatibility parameter for ice on a solid substrate, and x isthe ratio of the radius of the substrate to the radius ofspherical ice germ. The compatibility parameter, m, is equalto cosq, where q is the contact angle between an ice nucleusand the mineral surface.[36] Assuming the radius of the substrate to be muchlarger than the radius of the ice germ (a good approximationunder our experimental conditions), f(m,x) is defined asfollows:fðm;xÞ¼m3C0 3m þ 24ð5ÞTo calculate q, we first calculated the free energy offormation of the critical nucleus using equation (3), ourexperimentally determined Jhetvalues, and assuming a pre-exponential term (A) equal to 1025cmC02secC01[Fletcher,1958, 1959; Pruppacher and Klett,1997].Then,wecalculated the contact angle (q) using equations (4) and(5), assuming the density of ice (r)is0.92gcmC03[CRC,2001–2002], Mwis 18.015 g molC01, and si/vequals 106 ±5mJmC02[Pruppacher and Klett, 1997]. For si/v,weareusing the surface tension appropriate for hexagonal ice.Recent work has shown that cubic ice is the first phase tonucleate when homogeneous nucleation dominates [Murrayet al., 2005; Murray and Bertram, 2006], but furtherresearch is needed to determine if this also the case forheterogeneous nucleation.Table 4. J Values and Contact Angles for all Five Minerals StudiedMineralOnsetTemperature (K) RHiceJhet(cmC02sC01)Jhet, upper(cmC02sC01)Jhet, lower(cmC02sC01) qlowerqqupperKaolinite 246.1 105 ± 5 420 4200 4.2 5.2 9.7 13.9Kaolinite 241.1 104 ± 5 200 2000 2.0 2.4 7.8 12.4Kaolinite 236.4 104 ± 2 280 2800 2.8 3.4 9.2 13.5Muscovite 246.2 107 ± 5 940 9400 9.4 7.1 10.7 14.7Muscovite 241.3 106 ± 6 2200 22,000 22 5.5 9.5 14.3Muscovite 237.0 102 ± 2 490 4900 4.9 0.7 6.2 11.2Montmorillonite 244.6 124 ± 5 460 4600 4.6 19.6 22.3 25.3Montmorillonite 240.8 110 ± 4 870 8700 8.7 10.9 14.3 17.6Montmorillonite 236.0 106 ± 4 930 9300 9.3 7.7 12.0 15.6Quartz 243.9 133 ± 1 360 3600 3.6 - - -Quartz 238.7 137 ± 3 340 3400 3.4 24.5 27.1 30.1Quartz 234.0 136 ± 4 480 4800 4.8 23.7 26.3 29.3Calcite 244.1 130 ± 3 620 6200 6.2 - - -Calcite 238.9 135 ± 3 250 2500 2.5 23.8 26.4 29.4Calcite 234.2 132 ± 6 130 1300 1.3 22.2 24.9 27.9D22203 EASTWOOD ET AL.: ICE NUCLEATION ON MINERAL DUST PARTICLES7of9D22203[37] In Table 4, the contact angles calculated using theprocedure discussed above are listed. The contact anglevalues are also illustrated in Figure 7. The data show that forefficient ice nuclei such as muscovite and kaolinite, thecontact angles are small (below 18C176). For poor ice nucleisuch as quartz and calcite, the contact angles are larger(above 20C176). These values may be useful for future model-ing studies of ice nucleation in the atmosphere and forcomparing results between different laboratories.4. Summary and Conclusions[38] An optical microscope coupled to a flow cell was useto characterize the ice nucleation ability of muscovite,kaolinite, montmorillonite, quartz and calcite over thetemperature range of 233 to 247 K. Onset conditions forice nucleation, nucleation rates and contact angles weredetermined.[39] Onset measurements indicate that muscovite andkaolinite are very good ice nuclei with onset RHivaluesof less than approximately 110%, well below water satura-tion. This can be explained by a better crystallographicmatch between the pseudo-hexagonal arrangement of thehydroxyl (C0OH) groups at the kaolinite surface and thehexagonal ice structure which allows a better hetero-epitax-ial growth of the ice structure though a H-bond framework[Pruppacher and Klett, 1997]. Onset measurements indicatethat quartz and calcite are the poorest ice nuclei. For thetemperature range studied, the RHivalues needed to induceice nucleation on quartz and calcite are approximately 20 to40% higher than those needed for kaolinite and muscovite.In contrast, montmorillonite was an effective ice nucleus atthe two lowest temperatures studied (236.0 and 240.8 K),but a relatively poor ice nucleus at the highest temperaturestudied (244.6 K), based on the onset data. Overall, the datashow significant differences in the ice nucleating abilities ofthe five minerals studied over this temperature range.[40] The measured onset conditions for the mineral dustswere compared with previously published data. In severalcases, there was a lack of quantitative agreement amongpublished work. This can be explained by several factorsincluding the mineral source, the particle sizes, the surfacearea available for nucleation, observation and equilibriumtimes. Future studies that investigate the effect of thesedifferent parameters are needed for a complete understand-ing of ice nucleation on mineral particles.[41] The heterogeneous nucleation rates (Jhet) and contactangles (q) were determined according to classical nucleationtheory for all five minerals studied. The contact anglesmeasured for kaolinite and muscovite ranged from 6 to 12C176;whereas for quartz and calcite the contact angles rangedfrom 25 to 27C176. The reported Jhetand q values may allow fora more direct comparison between laboratory studies andcan be used when modeling ice cloud formation in theatmosphere. Future studies should investigate the depen-dence of onset conditions and nucleation rates on particlesize and the surface area available for nucleation to estab-lish more accurate predictions of ice nucleation in theatmosphere.[42] Acknowledgments. The authors would like to thank B. J.Murray, D. A. Knopf, and D. L. Macalady for their many helpfuldiscussions. This research was supported by the Canadian Foundation forClimate and Atmospheric Science (CFCAS), the Natural Sciences andEngineering Research Council of Canada (NSERC), and the CanadaResearch Chair Program.ReferencesAnthony, J. W., R. A. Bideaux, K. W. Bladh, and M. C. Nichols (1995),Handbook of Mineralogy, Mineral Data, Tucson, Ariz.Archuleta, C. M., P. J. Demott, and S. M. Kreidenweis (2005), Ice nuclea-tion by surrogates for atmospheric mineral dust and mineral dust/sulfateparticles at cirrus temperatures, Atmos. Chem. Phys., 5, 2617–2634.Bailey, M., and J. Hallett (2002), Nucleation effects on the habit of vapourgrown ice crystals from C018 to C042 C176C, Q. J. R. Meteorol. Soc., 128,1461–1483.Cantrell, W., and A. J. Heymsfield (2005), Production of ice in troposphericclouds—A review, Bull. Am. Meteorol. Soc., 86(6), 795–807.Chen, Y. L., S. M. Kreidenweis, L. M. McInnes, D. C. Rogers, and P. J.DeMott (1998), Single particle analyses of ice nucleating aerosols in theupper troposphere and lower stratosphere, Geophys. Res. Lett., 25(9),1391–1394.Cziczo, D. J., D. M. Murphy, P. K. Hudson, and D. S. Thomson (2004),Single particle measurements of the chemical composition of cirrus iceresidue during CRYSTAL-FACE, J. Geophys. Res., 109(D4), D04201,doi:10.1029/2003JD004032.DeMott, P. J. (2002), Laboratory studies of cirrus cloud processes, in Cir-rus, edited by D. K. Lynch, pp. 102–136, Oxford Univ. Press, New York.DeMott, P. J., K. Sassen, M. R. Poellot, D. Baumgardner, D. C. Rogers,S. D. Brooks, A. J. Prenni, and S. M. Kreidenweis (2003a), African dustaerosols as atmospheric ice nuclei, Geophys. Res. Lett., 30(14), 1732,doi:10.1029/2003GL017410.Demott, P. J., D. J. Cziczo, A. J. Prenni, D. M. Murphy, S. M. Kreidenweis,D. S. Thomson, R. Borys, and D. C. Rogers (2003b), Measurements ofthe concentration and composition of nuclei for cirrus formation, Proc.Natl. Acad. Sci. U.S.A., 100(25), 14,655–14,660.Dymarska, M., B. J. Murray, L. M. Sun, M. L. Eastwood, D. A. Knopf, andA. K. Bertram (2006), Deposition ice nucleation on soot at temperaturesrelevant for the lower troposphere, J. Geophys. Res., 111(D4), D04204,doi:10.1029/2005JD006627.Field,P.R.,O.Mo¨hler,P.Connolly,M.Kra¨mer,R.Cotton,A.J.Heymsfield,H. Saathoff, and M. Schnaiter (2006), Some ice nucleation characteristicsof Asian and Saharan desert dust, Atmos. Chem. Phys., 6, 2991–3006.Fletcher, N. H. (1958), Size effect in heterogeneous nucleation, J. Chem.Phys., 29(3), 572–576.Fletcher, N. H. (1959), On ice-crystal production by aerosol particles,J. Meteorol., 16, 173–180.Fletcher, N. H. (1962), Physics of Rainclouds,CambridgeUniv.Press,London, U.K.Girard, E., and J. A. Curry (2001), Simulation of low-level troposphericclouds during FIRE experiment in May 1998, J. Geophys. Res.,106(D14), 15,139–15,154.Girard, E., J.-P. Blanchet, and Y. Dubois (2005), Effects of sulphuric acidaerosols on wintertime low-level atmospheric ice crystals, humidity, andtemperature at Alert, Nunavut, Atmos. Res., 73, 131–148.Figure 7. Contact angle as a function of temperature foreach mineral studied.D22203 EASTWOOD ET AL.: ICE NUCLEATION ON MINERAL DUST PARTICLES8of9D22203Glaccum, R. A., and J. M. Prospero (1980), Saharan aerosols over thetropical North Atlantic—Mineralogy, Mar. Geol., 37, 295–321.Haag, W., and B. Ka¨rcher (2004), The impact of aerosols and gravity waveson cirrus clouds at midlatitudes, J. Geophys. Res., 109, D12202,doi:10.1029/2004JD004579.Hanisch, F., and J. N. Crowley (2001), The heterogeneous reactivity ofgaseous nitric acid on authentic mineral dust samples, and on individualmineral and clay mineral components, Phys. Chem. Chem. Phys., 3,2474–2482.Heintzenberg, J., K. Okada, and J. Stro¨m (1996), On the composition ofnon-volatile material in upper tropospheric aerosols and cirrus crystals,Atmos. Res., 41, 81–88.Hoffer, T. E. (1961), A laboratory investigation of droplet freezing,J. Meteorol., 18, 766–778.Hung, H. M., A. Malinowski, and S. T. Martin (2003), Kinetics of hetero-geneous ice nucleation on the surfaces of mineral dust cores inserted intoaqueous ammonium sulfate particles, J. Phys. Chem. A, 107(9), 1296–1306.Jensen, E. J., and O. B. Toon (1997), The potential impact of soot particlesfrom aircraft exhaust on cirrus clouds, Geophys. Res. Lett., 24(3), 249–252.Jensen, E. J., O. B. Toon, R. F. Pueschel, J. Goodman, G. W. Sachse, B. E.Anderson, K. R. Chan, D. Baumgardner, and R. C. Miake-Lye (1998),Ice crystal nucleation and growth in contrails forming at low ambienttemperatures, Geophys. Res. Lett., 25(9), 1371–1374.Kanji, Z. A., and J. P. D. Abbatt (2006), Laboratory studies of ice formationvia deposition mode nucleation onto mineral dust and n-hexane soot sam-ples, J. Geophys. Res., 111(D16), D16204, doi:10.1029/2005JD006766.Ka¨rcher, B. (1996), Aircraft-generated aerosols and visible contrails, Geo-phys. Res. Lett., 23(15), 1933–1936.Ka¨rcher, B. (1998), Physicochemistry of aircraft-generated liquid aerosols,soot, and ice particles—1: Model description, J. Geophys. Res.,103(D14), 17,111–17,128.Ka¨rcher, B. (2004), Cirrus clouds in the tropical tropopause layer: Role ofheterogeneous ice nuclei, Geophys. Res. Lett., 31, L12101, doi:10.1029/2004GL019774.Ka¨rcher, B., R. Busen, A. Petzold, F. P. Schroder, U. Schumann, and E. J.Jensen (1998), Physicochemistry of aircraft-generated liquid aerosols,soot, and ice particles—2: Comparison with observations and sensitivitystudies, J. Geophys. Res., 103(D14), 17,129–17,147.Knopf, D. A., and T. Koop (2006), Heterogeneous nucleation of ice onsurrogates of mineral dust, J. Geophys. Res., 111, D12201, doi:10.1029/2005JD006894.Lohmann, U., and K. Diehl (2006), Sensitivity studies of the importance ofdust ice nuclei for the indirect aerosol effect on stratiform mixed-phaseclouds, J. Atmos. Sci., 63, 968–982.Lohmann, U., and J. Feichter (2005), Global indirect aerosol effects: Areview, Atmos. Chem. Phys., 5, 715–737.Mangold, A., R. Wagner, H. Saathoff, U. Schurath, C. Giesemann, V. Ebert,M. Kramer, and O. Mo¨hler (2005), Experimental investigation of icenucleation by different types of aerosols in the aerosol chamber AIDA:Implications to microphysics of cirrus clouds, Meteorol. Z., 14(4), 485–497.Marcolli, C., S. Gedamke, T. Peter, and B. Zobrist (2007), Efficiency ofimmersion mode ice nucleation on surrogates of mineral dust, Atmos.Chem. Phys., 7, 5081–5091.Martin, S. T. (2000), Phase transitions of aqueous atmospheric particles,Chem. Rev., 100(9), 3403–3453.Mason, B. J., and J. Maybank (1958), Ice nucleating properties of somenatural mineral dusts, Q. J. R. Meteorol. Soc., 84, 235–241.Meyers, M. P., P. J. Demott, and W. R. Cotton (1992), New primary icenucleation parameterization in an explicit cloud model, J. Appl. Meteorol.,31, 708–721.Middlebrook,A.M.,L.T.Iraci,L.S.McNeill,B.G.Koehler,M.A.Wilson, O. W. Saastad, M. A. Tolbert, and D. R. Hanson (1993), Fouriertransform-infrared studies of thin H2SO4/H2O films—Formation, water-uptake, and solid-liquid phase-changes, J. Geophys. Res., 98(D11),20,473–20,481.Morrison, H., J. A. Curry, and V. I. Khvorostyanov (2005), A new double-moment microphysics parameterization for application in cloud and cli-mate models. part 1: Description, J. Atmos. Sci., 62, 1665–1677.Mo¨hler, O., et al. (2006), Efficiency of the deposition mode ice nucleationon mineral dust particles, Atmos. Chem. Phys., 6, 3007–3021.Murphy, D. M., and T. Koop (2005), Review of the vapour pressures of iceand supercooled water for atmospheric applications, Q. J. R. Meteorol.Soc., 131(608), 1539–1565.Murray, B. J., D. A. Knopf, and A. K. Bertram (2005), The formation ofcubic ice under conditions relevant to Earth’s atmosphere, Nature, 434,202–205.Murray, B. J., and A. K. Bertram (2006), Formation and stability of cubicice in water droplets, Phys. Chem. Chem. Phys., 8, 186–192.Pant, A., M. T. Parsons, and A. K. Bertram (2006), Crystallization ofaqueous ammonium sulfate particles internally mixed with soot and kao-linite: Crystallization relative humidities and nucleation rates, J. Phys.Chem. A, 110(28), 8701–8709.Parsons, M. T., D. A. Knopf, and A. K. Bertram (2004a), Deliquescenceand crystallization of ammonium sulfate particles internally mixed withwater-soluble organic compounds, J. Phys. Chem. A, 108(52), 11,600–11,608.Parsons, M. T., J. Mak, S. R. Lipetz, and A. K. Bertram (2004b), Deliques-cence of malonic, succinic, glutaric, and adipic acid particles, J. Geophys.Res., 109(D6), D06212, doi:10.1029/2003JD004075.Prospero, J. M. (1999), Long-range transport of mineral dust in the globalatmosphere: Impact of African dust on the environment of the south-eastern United States, Proc. Natl. Acad. Sci. U. S. A., 96, 3396–3403.Pruppacher, H. R., and J. D. Klett (1997), Microphysics of Clouds andPrecipitation, Kluwer Academic Publishers, Dordrecht, Netherlands.Richardson, M. S., et al. (2007), Measurements of heterogeneous ice nucleiin the western United States in springtime and their relation to aerosolcharacteristics, J. Geophys. Res., 112(D2), D02209, doi:10.1029/2006JD007500.Roberts, P., and J. Hallett (1968), A laboratory study of the ice nucleatingproperties of some mineral particulates, Q. J. R. Meteorol. Soc., 94,25–34.Rogers, D. C., P. J. Demott, and S. M. Kreidenweis (2001), Airbornemeasurements of tropospheric ice-nucleating aerosol particles in theArctic spring, J. Geophys. Res., 106(D14), 15,053–15,063.Salam,A.,U.Lohmann,B.Crenna,G.Lesins,P.Klages,D.Rogers,R.Irani,A. MacGillivray, and M. Coffin (2006), Ice nucleation studies of mineraldust particles with a new continuous flow diffusion chamber, Aerosol Sci.Technol., 40, 134–143.Sassen, K. (2002), Indirect climate forcing over the western US from Asiandust storms, Geophys. Res. Lett., 29(10), 1465, doi:10.1029/2001GL014051.Sassen, K. (2005), Dusty ice clouds over Alaska, Nature, 434, 456.Sassen, K., P. J. DeMott, J. M. Prospero, and M. R. Poellot (2003), Saharandust storms and indirect aerosol effects on clouds: CRYSTAL-FACEresults, Geophys. Res. Lett., 30(12), 1633, doi:10.1029/2003GL017371.Schuttlefield, J. D., D. Cox, and V. H. Grassian (2007), An investigation ofwater uptake on clays minerals using ATR-FTIR spectroscopy coupledwith quartz crystal microbalance measurements, J. Geophys. Res., 112,D21303, doi:10.1029/2007JD008973.Toon, O. B. (2003), African dust in Florida clouds, Nature, 424, 623–624.Twohy, C. H., and M. R. Poellot (2005), Chemical characteristics of iceresidual nuclei in anvil cirrus clouds: Evidence for homogeneous andheterogeneous ice formation, Atmos. Chem. Phys., 5, 2289–2297.Vali, G. (1985), Nucleation Terminology, Bull. Am. Meteorol. Soc., 66(11),1426–1427.Vali, G. (1996), Ice nucleation – a review, in Nucleation in AtmosphericAerosols, edited by M. Kulmala and P. Wagner, pp. 271–279, PergamonPress, Oxford, UK.Zimmerman, F., M. Ebert, A. Worringen, L. Schu¨tz, and S. Weinbruch(2007), Environmental scanning electron microscopy (ESEM) as a newtechnique to determine the ice nucleation capability of individual atmo-spheric aerosol particles, Atmos. Environ., 41, 8219–8227.Zuberi, B., A. K. Bertram, C. A. Cassa, L. T. Molina, and M. J. Molina(2002), Heterogeneous nucleation of ice in (NH4)2SO4–H2O particleswith mineral dust immersions, Geophys. Res. Lett., 29(10), 1504,doi:10.1029/2001GL014289.C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0A. K. Bertram, S. Cremel, M. L. Eastwood, and C. Gehrke, Departmentof Chemistry, University of British Columbia, 2036 Main Mall, Vancouver,BC, Canada V6T 1Z1. ( Girard, Department of Earth and Atmospheric Sciences, University ofQuebec at Montreal, P.O. Box 8888, Downtown, Montreal, QC H3C 3P8,Canada.D22203 EASTWOOD ET AL.: ICE NUCLEATION ON MINERAL DUST PARTICLES9of9D22203


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics

Country Views Downloads
China 35 18
Republic of Moldova 2 0
Pakistan 2 0
Saudi Arabia 2 0
Germany 2 2
United States 1 16
City Views Downloads
Unknown 14 126
Guangzhou 14 0
Shenzhen 6 18
Kunming 2 0
Islamabad 2 0
Chişinău 2 0
Hangzhou 2 0
Tianjin 1 0
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items