A Thermodesorber for Onlne studes of Combuston Aerosols - Influence of partcle dameter, resdence tme and mass concentraton Dahl, Andreas; Pagels, Joakm Publshed: 2003-01-01 Lnk to publcaton Ctaton for publshed verson (APA): Dahl, A., & Pagels, J. (2003). A Thermodesorber for Onlne studes of Combuston Aerosols - Influence of partcle dameter, resdence tme and mass concentraton. Abstract from 1st Internatonal Symposum on Incomplete Combuston, Unversty of Kuopo, Fnland. General rghts Copyrght and moral rghts for the publcatons made accessble n the publc portal are retaned by the authors and/or other copyrght owners and t s a condton of accessng publcatons that users recognse and abde by the legal requrements assocated wth these rghts. Users may download and prnt one copy of any publcaton from the publc portal for the purpose of prvate study or research. You may not further dstrbute the materal or use t for any proft-makng actvty or commercal gan You may freely dstrbute the URL dentfyng the publcaton n the publc portal? Take down polcy If you beleve that ths document breaches copyrght please contact us provdng detals, and we wll remove access to the work mmedately and nvestgate your clam. Download date: 24. Jun. 2016 L UNDUNI VERS I TY PO Box117 22100L und +46462220000
A THERMODESORBER FOR ONLINE STUDIES OF COMBUSTION AEROSOL PARTICLES - INFLUENCE OF PARTICLE DIAMETER, RESIDENCE TIME AND MASS CONCENTRATION ANDREAS DAHL and JOAKIM PAGELS Dvson of Aerosol Technology (EAT), Lund Insttute of Technology Sölvegatan 26, SE-221 00 Lund, Sweden Keywords: Thermodesorber, Volatlty TDMA INTRODUCTION Onlne physcal and chemcal characterstcs of ultrafne partcles are much needed n studes related e.g. to human health and new partcle formaton n the atmosphere. The condtonng of aerosol partcles at sutable temperatures n a thermodesorber (TD), thereby removng volatle substances s a promsng tool for such studes. TD:s are usually operated n connecton wth a partcle sze spectrometer (e.g. a SMPS or an ELPI). When operated at hgh concentratons a desorber part flled wth e.g. actve carbon s sometmes connected after the heater (e.g. Burtscher et al., 2001 and Wehner et al., 2002). By selectng monodsperse partcles wth a DMA before the sample pass the TD, hgh-resoluton spectroscopy of volatle components can be performed n a Tandem Dfferental Moblty Analyzer (TDMA-setup). For sphercal partcles, volatle volume fractons and vapour pressures can then be estmated. By combnng measurements of the partcle mass and moblty of the condtoned and uncondtoned partcles, quanttatve volatle volume fractons can be determned also for fractal-lke agglomerates (Park, 2003). Sakura et al. (2003) used a TD operated n a TDMA-setup to get nformaton about the vapour pressures of more volatle desel nano-partcles. Half the partcle volume was evaporated at around 343 K for 30 nm partcles and at around 368 K for 70 nm partcles. It s desrable to compare data collected n TD:s wth dfferent resdence tme, for dfferent partcle dameters and mass concentratons. Therefore experments to study the nfluence of these parameters are needed. In ths paper we descrbe the desgn and charactersaton experments of the recently bult LUND-TD for temperatures up to 873 K. We have used ths setup to study nfluence of partcle mass concentraton, partcle dameter and resdence tme n the TD on the evaporaton profle for relevant laboratory generated aerosols. THEORY In the free molecular regme the change n partcle dameter due to evaporaton can be wrtten as: dd p c M K p, ps, = dt 4 ρ R T Tp Eq.1 Where c denotes the mean molecular velocty of speces, M s the molar mass, K s the Kelvn factor, ρ s the partcle materal densty, p s, s the saturaton vapour pressure of speces just above the surface and p, s the partal pressure of speces, far away from the partcle. T p and T are the temperature at the partcle surface and far away from the partcle respectvely. Note that n ths model, the dameter change (nm/s) s ndependent of partcle dameter (neglectng the Kelvn effect). At low mass concentratons p, can be set to zero and the evaporaton rate becomes ndependent of mass-concentraton. In the contnuum regme dd p /dt s proportonal to D p -1. For partcle dameters smaller than 80 nm, the error of eq. 1, compared to the general formula (Flagan and Senfeld, 1988) s less than 30% for T=293 K and less than 10% for T=773 K.
METHODS The desgn of the LUND-TD s optmsed for flow rates between 0.3 and 1.0 l/mn, flows whch are preferred for the Scannng Moblty Partcle Szers (SMPS, TSI, Inc., USA). The heater conssts of an 18 mm n nner dameter, 550 mm long nsulated stanless steel tube. Thermocouples are brazed on the outsde of the heater, one at the mddle and one 50 mm from the ext. The regulator s connected to the mddle thermocouple, whle the second can be used to check the ext temperature. A 500 mm long adsorber tube wth actvated carbon s used as adsorbent. The heater s capable of operatng up to 1073 K. In order not to ncnerate the carbon, a 500 mm long tube wth an nner dameter of 8 mm replaces the adsorber when the TD s used at hgher temperatures (see fgure 1). Fgure 1. Schematc overvew of the LUND-TD and the locaton of the thermocouples. The resdence tme for the aerosol at the set TD temperature, T s, was calculated as tme were the aerosol was at a temperature hgher than (T s -25 K). The temperature profles are shown n fgure 2 and the calculated resdence tmes are dsplayed n table 1. Table 1. Estmated resdence tmes ( t) for the aerosol at dfferent T s and flow rates T s Flow rate [l/mn] [ K] 0.3 1.0 373 15.2 s 4.4 s 573 6.8 s 1.9 s 873 1.0 s Fgure 2. Temperature profles at a flow rate of 1.0 l/mn at 373, 573 and 873 K. The penetraton was measured for a monodsperse aerosols wth partcle dameters 20, 30, 50 and 100 nm at the flow rates 0.3, 0.6 and 1.0 l/mn and at TD temperatures from room temperature up to 773 K. A nebulzer model 3076 (TSI. Inc., USA) was used to generate a K 2 SO 4 -aerosol (chosen because of ts low vapour pressure). An electrostatc classfer (EC) model 3071 (TSI. Inc., USA) was used to select the partcle szes. A Scannng Moblty Partcle Szer (SMPS)-system, consstng of an EC model 3071 and a Condensaton Partcle Counter (CPC) model 3010 (TSI. Inc., USA), was used to count the partcles leavng the LUND-TD.
The evaporaton performance regardng ablty to evaporate and remove a volatle substance from the aerosol was nvestgated wth dfferent aerosols at dfferent mass concentraton. The tests of the thermodesorber were made wth a monodsperse KCl-aerosol and a condensaton aerosol consstng of d-octyl sebacate (DOS) partcles wth NaCl nucle. The setup for the KCl evaporaton was the same as for the penetraton measurements. For the dfferent TD temperatures the partcle sze dstrbuton was measured. Due to the rapd partcle dameter decrease n a narrow temperature range above 673 K, the TD temperature was ncreased wth 15 to 25 K steps n that regon. The DOS/NaCl-aerosol was prepared wth a condensaton aerosol generator (Model: SLG 270 Topas, GmbH, Germany). The generator was setup to produce an aerosol wth a mean number dameter of 500 nm and a σ g of <1.2. RESULTS AND DISCUSSION At room temperature and a flow rate of 0.3 l/mn, the penetraton of partcles wth a dameter of 100 nm s 90% and at hgher flow rate of 1.0 l/mn 96%. At 573 K the penetratons descents to 69% and 86% respectvely. The man deposton processes n the TD are thermophoress and dffuson and most of the deposton takes place n the secton of the TD where they work n the same drecton,.e. where the aerosol temperature decreases. The partcle penetraton through the TD s less dependent of partcle sze (n the range 20-100 nm) and more related to the flow rate and TD-temperature. The results from the evaporaton of the KCl-aerosols are dsplayed n fgure 3 as dameter change ( D p ) n nm at the set temperature. The results are from sngle measurements. The dameter change s ndependent of the ntal partcle dameter, as s expected from theory. From fgure 3 t s also obvous that there s no nfluence of the mass concentraton on the evaporaton profle at the low mass concentratons studed. In fgure 3 dfferences between the evaporaton profles at flow rates of 0.3 and 1.0 l/mn are also dsplayed. Strong resdence tme dependence s ndcated, then the partcle dameter decrease seen at 0.3 l/mn takes place approxmately 30 K hgher wth a flow rate of 1.0 l/mn. Comparng measurements of the two flow rates at the same temperature gves a rato n D p of around 3-4, whch accordng to eq. 1 (neglectng the Kelvn effect) gves a rato of the resdence tmes ( t) of 3-4. Ths s very close to the expermentally found dfference n resdence tme measured at 573K. A hgh flow rate wll yeld a hgher partcle penetraton but because of the temperature-resdence tme relatonshp a hgher temperature s needed to evaporate a gven compound. Fgure 3. Partcle dameter change for monodsperse KCl-aerosols wth ntal mean dameters of 40 and 80 nm at flow rates of 0.3 and 1.0 l/mn. The measurements on 80 nm partcles at the flow rate of 0.3 l/mn as well as the measurements on 40 nm partcles at 1.0 l/mn where performed at two dfferent ntal partcle mass concentratons The measurements were related to dameters at 473 K.
Fgure 4 dsplays partcle sze dstrbutons of DOS/NaCl-aerosols wth ntal partcle mass concentratons of 70 and 400 µg/m³. At deal evaporaton only the NaCl-nucleus should reman. The aerosol wth hgh mass concentraton yelds slghtly larger partcles extng the thermodesorber. Ths s ether due to nsuffcent evaporaton n the heater, lmted by the vapour pressure of DOS, or readsorbton of DOS on the partcles when the aerosol cools and enters the adsorber. The effect s not notceable at even hgher temperatures. Fgure 4. Partcle sze dstrbuton from evaporaton of DOS/NaCl-aerosol wth the LUND-TD at a flow rate of 0.3 l/mn. The samples A, B and C where measured on an aerosol wth an ntal mass concentraton of 400 µg/m³ and the samples a, b and c had the ntal mass concentraton 70 µg/m³. The A samples where led bypass the TD and the B where measured at TD a temperatures of 353 and C at 358 K. CONCLUSIONS A thermodesorber sutable for onlne measurements of combuston aerosols has been desgned and tested usng relevant laboratory generated aerosol s. The temperature range has been up to 873 K, but may be extended to 1073 K n future applcaton. The partcle penetraton through the thermodesorber and the temperature profle are comparable to exstng equpment. Evaporaton experments have been performed from 3 ng/m 3 n a TDMA setup wth monodsperse partcles, to 400 µg/m 3 n a sngle DMA-setup The nfluence of resdence-tme, partcle dameter and mass-concentraton on the evaporaton profle of the TD was found to be comparable to the theory n the free molecular regme. Takng the resdence tme n the TD nto account allows for a better comparson between measurements taken wth dfferent TD desgns and flow rates. More measurements are needed at hgh mass concentratons to nvestgate the performance of the adsorber and potental recondensaton of volatlsed materal on partcles. REFERENCES Burtscher H., Baltensperger U., Buckoweck N., Cohn P., Hügln C., Mohr M., Matter U., Nyek S., Schmatloch V.,Stret N., Wengartner E. (2001). Separaton of volatle and non-volatle aerosol fractons by thermodesorpton: Instrumental development and applcatons, Journal of Aerosol Scence, 32, 427 442. Sakura H., Tobas HJ., Parka K., Zarlng D., Docherty KS., Kttelson DB., McMurry PH., Zemann PJ., (2003) On-lne measurements of desel nanopartcle composton and volatlty, Atmospherc Envronment 37, 1199 1210 Wehner B., Phlppn S., Wedensohler A., (2002), Desgn and calbraton of a thermodenuder wth an mproved heatng unt to measure the sze-dependent volatle fracton of aerosol partcles, Journal of Aerosol Scence, 33, 1087-1093 Park K. (2003) In-Stu Measurements of 50-500 nm partcles: Mass-Moblty Relatonshp. PhD-Thess, Unversty of Mnnesota, USA Flagan RC., and Senfeld JH., (1988), Fundamentals of ar polluton engneerng, Prentce Hall, Englewood Clff, NJ, USA