Characterization of condensed phase nitric acid particles formed in the gas phase

NIRONMAvailableonlineatwww.sciencedirect.comRNAL OFScience Direct181m29Journal of Environmental Sciences 2011, 23(3)412418Characterization of condensed phase nitric acid particles formedin the gas phaseLong Jia", Yongfu Xu'1. State Key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry, Institute of Atmospheric Physics,Chinese Academy of Sciences, Beijing 100029, China. E-mail: Jialong@ mail iap. accn2. Graduate University of Chinese Academy of Sciences, Beijing 100049, ChinaReceived 25 April 2010: revised 21 June 2010: accepted 24 June 2010AbstractThe formation of nitric acid hydrates has been observed in a chamber during the dark reaction of NO2 with O in the presence ofair. The size of condensed phase nitric acid was measured to be 40-100 nm and 20-65 nm at relative humidity(rH)<5% and rh67% under our experimental conditions, respectively. The nitric acid particles were collected on the glass fiber membrane and theirnd NO3"xH2O(x> 4)at low RH, whereas at high RH HNO3. H2O, HNO. 2H2O, HNO3 3H2O and NO3"xH2O(xe 4)all exist in theondensed phase At high RH HNO, xH2O (x< 3)collected on the glass fiber membrane is greatly increased, while NO3. xH2O(x>4)decreased, compared with low RH. To the best of our knowledge, this is the first time to report that condensed phase nitric acid canbe generated in the gas phase at room temperatureKey words: nitric acid; particles; relative humidity; chamberDOI:10.1016/S1001-0742(10604147Citation: Jia L, Xu Y F, 2011. Characterization of condensed phase nitric acid particles formed in the gas phase. Journal ofEnvironmental Sciences, 23 (3): 412-418IntroductionIt is usually believed that gaseous HNO3 can reactto form secondary aerosols on the surfaces of or withinThe hydrates of nitric acid are of great interest because existing aerosol particles, but it can hardly nucleate tof their roles in the formation of polar stratospheric clouds form new particles under normal circumstances, becauseand acid rain, therefore there are many studies that re. the vapor pressure of HNO at 293 K is approximatelyveal the properties of nitric acid hydrates in cold vapor 10 to 10 times higher than that of H2SO4( Spurny, 2000)(Tolbert and Middlebrook, 1990: Ritzhaupt and Devlin, Hence, the results have not been reported about nitric acid1991: Koehler et al., 1992; Tolbert et al., 1992; Hanson particles bom in the gas phase at room temperature andand Mauersberger, 1988; Worsnop et al., 1993: Barton et pressureal., 1993). Some studies showed that hydration of nitric The previous studies of the reaction of HNO, with wateracid can also take place under normal atmospheric con- were mainly focused on low temperature or on the surditions. Lee et al. 1980)studied the gas-phase clustering faces. If HNO can particulate in the ambient air condition,equilibria of NO2" and NO3- with high-pressure mass there will be two obvious effects:(1)the influences of thespectrometry, and pointed out that gas-phase nitric acid hydrates of nitric acid on the ratio of VOC/NOx and on thecan reversibly bind water molecules to form HNO, nH2O, ozone pollution in the photochemical reactions, and (2)thewhere the value n is dependent on the atmospheric water direct contribution to the nucleation of clouds and hazes.concentration. Eatough et al. (1985) studied the collection The purpose of the present work was to characterize theof gas phase nitric acid by diffusion denuders and further nitric acid particles formed during the reaction of NO2 withproved that the species collected on the nylon denuders O at room temperature and ambient air pressure. To theis actually HNO3 nH2O. Ramazan et al. ( 2006)recently best of中国煤化工 me to characterizestudied the heterogeneous reaction of NO2 with H2O on size digas phase at roomthe glass surface and suggested that HNO can react with tempCNMHGH20 to form nitric acid hydrates on the glass surface at atemperature of 296Korresponding author. E-mail: xyf@mailiap.accnCharacterization of condensed phase nitric acid particles formed in the gas phase1 ExperimentalTable 1 Initial conditions for experimentsThe experiments were performed in a 100 L reactorNo(ppb)made of 0. 1 mm FEP Teflon film at a temperature range #1of 289-295 K with a total pressure of 1 atm. The details of #2the apparatus have been described by Xu et al. (2006). Gas #36200533phase HNO was generated by the reaction of NO, with #4%7<≤≤<9Oy in the absence of irradiation. Before each experiment, #611270the reactor was washed using N2(99.999%), until O3and NOx concentrations were under the detecting limiRH: relative humidityAfter the residual gas was pumped out, the given volumesof 101 ppm NO2, synthetic air and ca. 500 ppm O3 2 Results and discussionwere sequentially introduced into the reactor according to Table 1 presents the experimental conditions for thethe chosen reactant concentrations. Then, the reactor wasvigorously shaken to make the reactants mixed thoroughly, NO, wall losses were obtained to be 40x 10-6 and 2.41after which timing the reaction began. At the same time, x 10-6 sec-I in the chamber, respectively ia et al., 2009)the ozone concentration was monitored on line. Syntheticair was used as background gas. The fine particles in syn. Experiments #1 and#2 were used to characterize the sizethetic air were removed by an Teflon holder that contains adistribution of nitric acid hydrate clusters under rH=67%and RH 5%, respectively; #3 and #4 were used to studyPTFE(polytetrafluoroethylene) fiber membrane (pore size the total nitrate products and gas phase nitrate; and #5 and0.2 um, Sartorius, Germany). During each experimentalcourse the temperature varied about t 1 K. After each #6 were used to determine the influences of rH on theexperiment the reactor was flushed using synthetic air forcondensed phase nitrate products10 hr with a 40 w blacklight lamp on2.1 Determination of gas phase nitric acidThe gas-phase species such as N2O5, HNO3, NOx andO, were determined by a long path Fourier transform Many studies(Spicer et al, 1982; Akimoto et al., 1980;infrared spectrometer (LP-FT-IR, Nicolet iS10, Ther-Grosjean, 1985) pointed out that the"NO2"reading of amo Fisher Scientific Inc, USA), which includes a 33 m chemiluminescent NOx analyzer includes gaseous HNO:permanently aligned long path gas cell(PIKE Tech-Experiments #l and #2, after 20 min of reaction, the Modelnologies, USA). NOx (NO, NO2)was simultaneously 42c detected that the"NO2"concentration in Experimentmonitored by a chemiluminescence analyzer(Model 42CThermo Fisher Scientific Inc, USA). O was monitored #2(RH 5%)was over 580 ppb higher than that inwith a UV photometric O, analyzer(Model 49C, Ther- Experiment #1(RH= 67%)(Fig. 1), which suggests themo Fisher Scientific Inc, USA). The formation and important role of RH in the NO2-O3 systemThe gas-phase NO2-0 system is well characterizedevolution of particles in the range of 13.5-645 nm were and is included in all kinetic models, such as in MCMmonitored using a scanning mobility particle sizer (tsModel 3034: DMA Model 3081 and CPC Model 3775, (the Master Chemical Mechanism)(Saunders et al., 2003).TSI Inc,USA). The chemical composition of particles Thus, the gas-phase concentrations of reactants and prod-collected on the glass fiber membrane(pore size 0. 45ucts like NO2 can be obtained by computer modelingStaplex, USA)was analyzed by FT-IRThe theoretically calculated NO2 concentrations are ingood agreement with"NO2 "readings under the high RHcondition(Fig. la). However, when RH is low(< 5%6),2500X NOx NO中国煤化工CNMHGFig. 1 Observed(symbols) and modeled (lines)concentration-time profiles of O and NOx obtained in the NO2-O3 experiments(measured NOzinclude NOz. HNO and N2O5). (a)RH= 67%(Exp #l);(b)RH 5%(Exp #2)41Joumal of Environmental Sciences 2011, 23(3)412-418/ Long Jia et aL.theNO2 "reading is much higher than the calculated Noz that gaseous HNO,(or N2Os)was only 80 ppb, and that(Fig. 1b). The possible reason is that gaseous nitric acid over 93% of nitric acid was probably converted into thewas converted into the condensed phase under the high particlesRH condition and was filtered out by the Teflon membranein the Model 42C inlet. Thus, the interference from gas-2.2 Size distribution of nitric acid hydrate particlesphase HNO3 can be removed under the high RH conditionTo further ensure whether there was condensed phaseSeveral repeated experiments all confirmed the conclusion. nitric acid in the chamber, the size and number concenAccording to the model calculation of the NOx-O, tration of nitric acid particleswas measured bsystem, after 60 min (in Experiments #l and #2), NO2 DMA TSI Model 3081 with CPC TSI Model 3775. Resultswould be almost consumed(Fig. 1). Thus, we take 60 for both low and high RH conditions are shown in Fig. 2.min as a reference time point. In Experiment #2( Fig. Ib), Both lower panels show the variation of mean numberafter 60 min of the reaction over 558 ppb"" was concentrations with time. During the first 10 min, particlesstill measured, which implied that there was over 558 ppb were seldom detected under both low and high RH condi-HNO(or N2O5)in the gas phase and another 688 ppb tions, because formed HNO3 was very limited during thisnitric acid was formed in the condensed phase or on the period. After 10 min, the number concentration began tochamber wall because of its combination with water. It accumulate and the particle size started to grow quicklycan be considered that the condensed nitric acid particles After 30 min both particle size and number concentrationwere mainly formed because the wall loss of HNO was reached a relatively steady state. It can be seen that whengnorable. According to Bloss et al. (2005)who considered RHis <5%, the particle size is in the range of 30 to 100 nmhat the rate constant of the HNO, wall loss ranged from and the mean number concentration is 2060 cm-3.When4.1x10-5 sec-I to 1.4x10-4 secin the NO2-O3 system, RH is 67%, the particle size is focused in the range of 20we estimate that the wall loss of HNO was only 62 ppb to 60 nm and the mean number concentration is 820 cm-3.at 60 min of reaction time In Experiment #l (Fig. la), the The mean number concentration of nitric acid particles atmeasured"NO2"was only 80 ppb at 60 min, indicating high RH (67%)are decreased 60%o, compared to that at lowRH≤5%RH=67%00200300400500600700800900100010102030405060708090Mean lineMean line△203040Time(min)Fig. 2 Size and number concentration of nitric acid particles vs reaction timeRH≤5%RH=67%中国煤化工001201CNMHG 120 140Diameter(nm)Diameter(nm)Fig 3 Size distribution of nitric acid particles at 60 min of reaction time at RH 5% and RH = 67%.Characterization of condensed phase nitric acid particles formed in the gRH(<5%), as shown in Fig. 2.min. The reduced"NO2"should come from gas phaseFigure 3 shows the size distribution of nitric acid parti- HNO3. This illustrates that gas phase HNO, accounts forcles, which is extracted from Fig. 2 at 60 min time point. 76% of total nitrogen-containing compounds in the gasThis clearly indicates that RH changes the size distribution phase. After the glass fiber filter, there are still 386 ppbof particles. Under low RH of <5%6, the size of nitric acid" NO2"remaining in the gas phase, which may come fromarticles ranges from 40 to 100 nm, whereas under high gas phase N2ORH of 67%, the measured particle size is mainly in a range The FT-IR spectrum of nitrate collected on the glassof 20 to 65 nm. The largest number concentration appears fiber membrane in the 1250-3650 cm region is shownto be at 58 nm at low RH, whereas it appears to be at 24 in Fig. 4b, in which the peak at near 3126 cm"is mainlynm at high RH. Thus, we consider that high RH leads to nitric acid complexes with one or two water moleculesthe removal of nitric acid particles larger than 65 nm from The spectrum(Fig. 4b)can be deconvoluted into sixthe gas phase, and that high humidity is favorable for the contributing peaks at 3380-3414, 3078, 2858, 2479, andsmall particles to be bon and to grow1422-1347", which is shown in Fig.4b1,b2.Accord2.3 Identification of total nitrate productsing to Ramazan et al. (2006)and Mccurdy et al. (2002).it is obvious that the peak at 3380-3414 cm is H2OTo determine the component of detected"NO2"under The peaks at 3080 and 2858 cm" can be assigned tolow RH conditions, a glass fiber filter was used to sample HNO] H2O and HNO].2H2O, respectively. The shouldertotal nitrate. The glass fiber filter that is relatively alkaline at 2479 cm"in Fig. 4b1 is assigned to HNO33H20.Thecan approximate efficient total inorganic nitrate samplers absorption bands at 1417 and 1356 cm- are assignedand retain both particulate nitrate and HNO3 with high to v3 asymmetric stretch of NO3 complexed to waterefficiency under atmospheric conditions(Newman, 1993). (NO3"xH2O)(x: 4-5)( Finlayson-Pitts et al., 2003). ForFigure 4a shows the detected"NO2 "before and after the this experiment, the spacing between the peaks of 1417 andglass fiber filter for Experiment #3. The difference between 1356 is 61 cm". In other experiments, the spacing can be"NO2"readings before and after the glass fiber filter is 65 and 75 cm" at rH 5% and rH= 90%, respectively1204 ppb at 60 min. The detected"NO2"is reduced by Experiment #4 was used to separate condensed phase76% after the glass fiber filter in the time period of 30-60 HNO, and gas phase HNO3. The combination of Teflon6000NO, before glass fiber filter一▲-NO2 after glass fiber filter0.152000300020001500Wavenumber(cm-)0200.150.12858中国煤化工CNMH35013001250Wavenumber(cm-)Wavenumber(cm-)tion-time profiles of NO and O3 for Exp. #3 before and after the glass fiber filter(a), and FT-IR spectrum of nitric acid hydratescollected on the glass fiber membrane from Exp. #3(b)Journal of Environmental Sciences 2011. 233)412418/Long Jia et alVoL 23and glass fiber filters in series was used to remove particles condensed phase at RH $%.and determine gaseous HNO3 (Mehlmann and Warneck,1995). Gaseous HNO, collected on the glass fiber mem- 2.4 Relative humidity impacts on nitric acid hydratesbrane was determined by FT-IR. when the RH is <5%, theTo quantitatively study the infuences of rh to nitricpeaks(not shown)at both 3078 and 2858 cm"assigned acid hydrates, the NO2-03 system with high concentrationto HNO. H2O and HNO3 2H20 are almost as strong as was used. The nitric acid hydrates were collected with thethose in Fig. 4b. The band at 2479 cm-Iassigned to glass fiber membrane at different RH conditions. TheirHNO33H20 disappears. The peak heights at both 1356 spectrums and deconvolution are shown in Fig. 5. Thecm" and 1417 cm"assigned to NO3 are about 14% of quantity of nitric acid complexes is estimated from thethose in Fig. 4b. Although these weak peaks at both 1356 Beer-Lambert law, as shown in Table 2.and 1417 cm"for Exp. #4 indicate that condensed HNO3 Of these peaks, the peak at 3080 cm assigned toNO3.xH2O) either may not be completely removed by HNO3. H2O is the largest contributor under both lowthe Teflon filter or is probably formed after the Teflon filter. and high RH conditions. The second contributor comesIt can be considered that HNO3 3H20 and NO3-xH2o from the peak of NO 3".xH2O. Compared with the lotdo not exist in the gas phase. Therefore, we can conclude RH condition(RH 5%), the nitric acid hydrates ofthat HNO3 H2O and HNO3 2H20 mainly exist in the gas HNO3 xH2O(x 3)are all increased, but NO3"xH20phase, and that HNO3 3H2O and NO3xH20 are in the is slightly decreased (1422-1347 cm")under the highRH<5%RH=90%0.152500Wavenumber(cm)Wavenumber(cm-RH≤5%025}dRH=90%0.150.100050051300Wavenumber(cm-)Wavenumber(cm)Fig. 5 Deconvolution of nitric acid hydrate absorbance bands(Exps. #5 and #6)collected on the glass fiber membrane. The irregular line representsthe spectrum, the smooth lines indicate the individual peaks that make up the spectrum, and the solid bold line through the data is a summation of theTable 2 Estimated concentrations of nitric acid complexes collected on the glass fiber membraneRH≤5%RH=90%HExperimental integrated Quantity(cm-1)km/mol)aband intensity(cm")中国煤化工”k0mr0.11335.1CNMHGHNO3 3H,02950.041NO3H2O(x4-5)1347-142259243.10073The theoretical intensities reported by Ramazan et al. (2006).Characterization of condensed phase nitric acid particles formed in the gas phaseH condition(RH=90%). This is in agreement with the 3 Conclusionschange of the number concentrations of nitric acid particlesunder different RH conditions, including the increase of The characterization of nitric acid hydrate particlesparticles with the size of 20-40 nm and great decrease of formed from the dark reaction of NO2 with O3 has beenarticles larger than 65 nm.investigated in the chamber under normal temperature andThe spacing between NO3"xH2O absorption peaks is pressure. At low RH of $%, HNO3. H2O and HNO3 2H2O75 cm-at RH=90%, whereas the spacing is 65 cm-at are mainly in the gas phase, while HNO, 3H20 andRH <5%, as shown in Fig. 5c, d. The measured spacing is NO3"xH2O(x> 4)are in the condensed phase. The sizedifferent under different conditions, particularly at different of condensed phase nitric acid hydrates were measured tomatrix. Ritzhaupt and Devlin(1977)obtained the spacing be 40-100 nm. At high RH, HNO3-H2O, HNO3 2H2O,of 150 and 65 cm"in an argon matrix containing 6% water HNO3 3H20 and NO3 H2O(x> 4)were all detectedand the matrix containing 100% water, respectively. Ra- to be in the condensed phase, and the particle size wasmazan et al.( 2006)observed a splitting of 93 cm-at rh mainly in 20-65 nm with the increase of RH, the quantity50%. They pointed out that the spacing decreases as the of HNO3 H2O, HNO3 2H20 and HNO3 3H20 collectedamount of water associated with the nitrate ion increases. on the glass fiber membrane is greatly enhanced, whereasTheoretical calculation indicates the splitting of 101 cm-I NO3-xH2O(x>4)is slightly decreased. Compared withfor nitrate associated with four water molecules and 62 low Rh of 5%, HNO3 3H20 is increased by overcm-for the complex with five water molecules(Ritzhaupt 100% at high RH(90%). High RH is favorable for theand Devlin, 1977). Therefore, our results indicates that small particles to be bom and to grow, and can removethe H20 number x in NO3"xH20 is reduced under the the larger particles containing more H2O. To the besthigh RH condition. This conclusion is in agreement with of our knowledge, our experiments are the first time tothe size distribution of nitric acid particles shown in Fig. demonstrate that the formation of condensed nitric acid can3, in which the larger particles containing more H2O are take place in the gas phase at room temperatureremoved under the high RH condition.Figure 6 shows the changes of composition in nitrateAcknowledgmentsparticles for different RH conditions. When RH increas-This work was supported by the Knowledge Innova-es from 5% to 90%, the largest increase comes from tion Program of the Chinese Academy of Sciences(NoHNO3 xH2O(x< 3), in which HNO3 3H20 is increased KZCXZ-Yw-Q02-03)by over 100%. We consider that as rH increases, thenumber x in HNO3 xH2O becomes great. When x>4, ReferencesHNO3 xH2O forms droplets and dissociates to form H3O+and NO3"xH2O. Compared to the change of particle size Akimoto H, Bandow H, Sakamaki F, Inoue G, Hoshino M,distribution in Fig 3, when RH increases from 5% to 67%,Okuda M, 1980. Photooxidation of the propylene-NOxthe increased particles are mainly particles with the size ofair system studied by long-path Fourier transform infrared20-40 nm. These particles should be HNO3 xH2o (r< 3)spectrometry. Ervironmental Science 6 Technology, 14The reduced particles are mainly particles larger than 65172-179,nm, which should be NO3 .xH2O. Ramazan et al. (2006)Barton N, Rowland B, Devlin J P, 1993. Infrared spectra of largeproposed a hypothesis that Reactions(1)H4)might happenacid hydrate clusters: formation conditions of submicronon the glass surfaces. Our experimental results indicate thatparticles of HNO3 2H2O and HNO3 3H2O. Journal of Phys-these reactions can also take place in the gas phase.ical Chemistry, 97: 5848-5851Bloss C, Wagner V, Bonzanini A, Jenkin M E, wirtz K, Martin-HNO3+H2O→HNO3H2O(1)Reviejo m, Pilling M J, 2005. Evaluation of detailedaromatic mechanisms(MCMv3 and MCMv3. 1) againstHNO3. H2O+ H2O HNO3 2H2Oenvironmental chamber data. Atmospheric Chemistry andHNO3 2H20+H0- HNO3 3H2O(3) Eatough D White V E. Hansen L D, Eatough N L, Ellis EHNO3 3H20+ yH20 H3O+NO3xH2OC, 1985. Hydration of nitric acid and its collection in theatmosphere by diffusion denuders. Analytical Chemistry,RH=5%Finlayson-Pitts B J, wingen L M, Sumner A L, Syomin D,Ramazan K A, 2003. The heterogeneous hydrolysis of0.08NOz in laboratory systems and in outdoor and indoor at-mospheres: An integrated mechanism. Physical ChemistryysICS5:223-242.004Grosjean D, 1985. Wall loss of gaseous pollutants in outdoor中国煤化工 ice a lechnologyHansHNO, H,O HNO, 2H-O HNO, 3H 0 NO, xH,OCNMHGssures of HNO3/H20(x:4-5)Istry,92:6167-6170.Fig 6 Comparisons of estimated nitric acid complexes at different RH Jia L, Xu Y F, Ge M F, Du L, Zhuang G S, 2009. Smog chamber418Jourmal of Environmental Sciences 2011: 23(3)412418/ Long Jia et alstudies of ozone formation potentials for isopentane. Chi- Ritzhaupt G. Devlin J P, 1977. lonic vs. molecular nature ofnese Science Bulliten. 54: 4624-4632monomeric ammonium and hydronium nitrate. InfraredKoehler B G, Middlebrook A M, Tolbert M A. 1992. Charac-spectra of hydronium nitrate(H3ONO, and ammoniumterization of model polar stratospheric cloud films usingnitrate(NH NO )solvated in argon matrices. Journal ofFourier transform infrared spectroscopy and temperaturePhysical Chemistry, 81: 521-525programmed desorption. Journal of Geophysical Research, Saunders SM, Jenkin M E, Derwent R G, 2003. Protocol for the97:80658074.development of the master chemical mechanism, MCMv3Lee N, Keesee R G, Castleman A wJr. 1980. The(PartA): tropospheric degradation of non-aromatic volatileclusters in the gas phase IV. Complexes of Horganic compounds. Atmospheric Chemistry and Physics,clustering on NOx Joumal of Chemical Physi3:161-180.1094Spicer C W, Howes Jr E, Bishop T A, Arnold L H, StevensRK,McCurdy P R, Hess W P, Xantheas S S, 2002. Nitric acid-1982. Nitric acid measnt methods: An intercompareater complexes: theoretical calculations and comparisonson Atmosphere Environment, 16: 1487-1500to experiment. Journal of Physical Chemistry A, 106(33): Spurny K R, 2000. Aerosol Chemical Processes in the Environ-ment. Lewis Publishers, New York. 398.Mehlmann A, Warneck P, 1995. Atmospheric gaseous HNO,, Tolbert M A, Koehler B G, Middlebrook A M, 1992.articulate nitrate, and aerosol size distributions of majoronic species at a rural site in western Germany. Atmospher-troscopic studies of model polar stratospheric cloudEnvironmen,29(l7):2359-2373Spectrochimica Acta, 48A: 1303-131Newman L, 1993. Measurement Challenges in AtmosphericTolbert M A, Middlebrook A M, 1990. Fourier transform infraredChemistry. American Chemical Society, Washington DCstudies of model polar stratospheric cloud surfaces: Growth6and evaporation of ice and nitric acid/ice Jourmal of Geo-Ramazan K A wingen L M, Miller Y, Chaban G M, Gerberphysical Research, 95: 22423-22431R B. Xantheas S S et al, 2006. New experimental and Worsnop D R, Fox L E, Zahniser M S, Wofsy SC,1993.Vaportheoretical approach to the heterogeneous hydrolysis ofpressures of solid hydrates of nitric acid: Implications forNO2: key role of molecular nitric acid and its complexespolar stratospheric clouds. Science, 259: 71-74Journal of Physical Chemistry A, 110: 6886-6897Xu Y F, Jia L, Ge M F, Du L, Wang G C, Wang D X, 2006Ritzhaupt G, Devlin J P, 1991. Infrared spectra of nitric andA kinetic study of the reaction of ozone with ethylene inhydrochloric acid hydrate thin films. Journal of Physicala smog chamber under atmospheric conditions. ChineseChemistry, 95: 90-95Science bulletin, 51: 2839-2843中国煤化工CNMHG