2.1 Aerosol generation

Four primary aerosols, fresh soot, aged (i.e. organically coated) soot, inorganic salt and mineral dust particles, were generated as depicted in Fig. 1. Fresh soot particles were generated with a miniCAST 6204 burner (Jing Ltd., Switzerland). The operation point was optimised to produce combustion particles with a geometric mean mobility diameter (GMD) of 90 nm and EC $/$ TC (elemental carbon to total carbon) mass fraction of >90 %. The combustion aerosol was split in two portions; one portion was led to the exhaust and the other through a metallic agglomeration tube (1.2 m long, 5 mm internal diameter), where the soot particles grew to about 120 nm. The mobility diameter was measured by a scanning mobility particle sizer (SMPS). The combustion aerosol was subsequently diluted by a factor of 10 with a VKL10 dilution unit (Palas, Germany). The outlet flow was delivered into an oxidation flow reactor known as the Micro Smog Chamber (MSC prototype developed by Keller and Burtscher, 2012, and used by Bruns et al., 2015; Corbin et al., 2015b, 2015a; Keller and Burtscher, 2012), where soot was mixed with a controlled amount of α-pinene vapours (≥97 % purity, Sigma Aldrich, Switzerland) under dry conditions (RH<5 %). The concentration of α-pinene at the inlet of the MSC was determined with a photoionisation detector (PID PhoCheck TIGER, Ion Science Ltd, UK) after filtering out the particles. The concentration could be varied by adjusting the flow of air through the α-pinene container (gas bubbler) and typically ranged between 60 and 70 ppm. RH was measured with a digital humidity sensor (FHAD 46 series/Almemo D6, Ahlborn, Germany). The aerosol flow through the MSC was set to 1.2 L min⁻¹ with the use of a miniature radial air blower (model H015X-525A9 with controller, Micronel AG, Switzerland). Higher aerosol flows through the MSC would lead to the residence time in the reactor being too short and should be avoided. α-Pinene underwent ozonolysis in the MSC, forming secondary organic aerosol (SOA), part of which condensed on the surface of the soot particles, simulating atmospheric ageing procedures (Ess et al., 2020).

Figure 1Schematic illustration of the experimental set-up. DUT stands for device under testing.

Download

The GMD of the soot mobility size distribution was shifted to 160 nm upon coating with SOA, and the EC $/$ TC mass fraction dropped to about 20 %. In parallel, fresh soot particles (120 nm mobility diameter) were sampled from the exhaust of the VKL10 dilution unit with the use of a second Micronel blower at flows between 1 and 2 L min⁻¹.

Mineral dust particles (ISO 12103-1 A2 fine test dust, Powder Technology Inc., USA) were generated with a rotating brush generator (RBG 1000, Palas, Germany) and were injected horizontally into an empty vessel, which acted as a swirl separator, filtering out the largest size fraction above PM₁₀. Alternatively, whenever calibration with respect to the PM_2.5 faction is desired, a PM_2.5 impactor can be installed right before injecting the dust particles into the homogeniser.

Inorganic salt particles were generated by nebulising aqueous mixtures of ammonium sulfate and ammonium nitrate at various ratios with the use of a TSI 3076 atomiser (TSI Inc., USA). The particles were passed through a 1.5 m long, spiral-shaped agglomeration tube to increase the GMD of the (number-based) mobility size distribution to about 100 nm (the mass-based aerodynamic size distribution shows a maximum at ≈200 nm). The aim was to simulate the presence of ammonium, nitrate and sulfate ions in the fine mode of atmospheric particle size distributions (Liu et al., 2000; Wall et al., 1988; Zhuang et al., 1999). Although generation of coarse-mode nitrate, formed at coastal areas by the reaction of gas-phase nitric acid with sea-salt or soil dust particles, or coarse-mode sulfate was not actively pursued, there is evidence (see Sect. 3) of coarse-sulfate formation. Presumably, this is either due to internal mixing of sulfate ions and mineral dust particles in the flow tube homogeniser or to deposition of salt particles in the aerosol pipes and consequent re-entrainment of agglomerates, which are larger than the particles initially produced by the generator.

The primary aerosols were introduced into a flow tube homogeniser (see Sect. 2.2) through separate injection ports. The flow of each primary aerosol entering the homogeniser could be regulated with separate mass flow controllers (Red-y MFC, Vögtlin, Switzerland) by splitting and directing part of the main primary aerosol flow to the exhaust. A filter (HEPA capsule, Pall Corporation, USA) was placed upstream of each MFC to remove the particles from the air flow. All four MFCs were connected to the same aerosol pump (VTE8, Thomas, Germany) as shown in Fig. 1.

The mobility diameter and number concentration of the soot and salt particles were determined with a scanning mobility particle sizer (SMPS 4.500, Grimm Aerosol Technik GmbH & Co. KG, Germany, L-DMA, Am-241 neutraliser, scan time 695 s). The mass concentration of each primary aerosol was measured with a tapered element oscillating microbalance (TEOM 1405, Thermo Scientific, USA), operated at a flow rate of 3 L min⁻¹ and a temperature of 30 ^∘C. The TEOM data were recorded via a custom-made LabVIEW routine every 6 s without averaging. The size distribution of the dust particles was measured with a Fidas Frog fine-dust monitor (Palas, Germany) and a high-resolution optical particle counter LAS-X II (Particle Measuring Systems, USA).

2.2 Aerosol homogenisation and sampling

The homogeniser is a 2.1 m long custom-made stainless steel tube with an inner diameter of 16.4 cm, placed vertically. The design is based on a previous study but has been significantly improved, and the facility has been shortened (Horender et al., 2019). The tube is equipped with five identical inlets, placed at the very top as shown in Figs. 1 and 2a. Dilution air (filtered, humidity and temperature controlled) is delivered to each one of the inlets at a flow rate of 24 L min⁻¹. The air is conditioned in two steps (Niedermeier et al., 2020) in such a way that the humidified air is particle-free: first, the dew point is adjusted by passing the air through a Nafion humidifier (Series FC125-240-10MP, PermaPure, USA) filled with water (ultra-analytic grade, Purelab ultra, ELGA, Switzerland) at a preselected water temperature, adjusted between 3 and 30 ^∘C with a cryostat–thermostat (LAUDA Ecoline Staredition RE 306, Lauda DR. R. Wobser GmbH & Co. KG, Germany). After being put through the Nafion humidifier, the air is fully saturated with water. Subsequently, the air is guided through a heated hose (Series T-7000, Thermocoax Isopad GmbH, Germany), where the temperature can be adjusted up to 100 ^∘C. The temperature and RH of the aerosol were monitored in the homogeniser at the height of the sampling probes with digital sensors (FHAD 46 series/Almemo D6, Ahlborn, Germany).

Figure 2(a) Computer-aided design (CAD, Inventor Professional 2019, Autodesk, USA) of the homogeniser. Panels (b) and (c) show enlarged views of the primary aerosol inlets and isokinetic sampling probes, respectively.

Download

The primary aerosols are injected in the middle of the tube through separate ports located 50 cm downstream as shown in Fig. 2b. The dilution air sweeps the particles down the tube, where they are further mixed by three turbulent jets of air. The three air-jet injection tubes (flow rate 20 L min⁻¹ each) are placed symmetrically around the homogeniser tube pointing 60^∘ downwards (Fig. 2b). The total flow rate of the homogenised aerosol is hence equal to 180 L min⁻¹ plus the flows of the four primary aerosols (in total less than 10 L min⁻¹). The temperature and relative humidity of the air jets are adjusted as described above for the dilution air. Finally, the homogeniser is surrounded by copper tubes with flowing water in order to maintain the stainless-steel tube at the same temperature as the aerosol. The temperature of water is adjusted by a flow-type cooler (AS-160 Green Line, Lindr, Czech Republic) or a thermostat (LAUDA EcoGold E4, Lauda DR. R. Wobser GmbH & Co. KG, Germany). The water flows in a closed loop, i.e. circulates back to the cryostat–thermostat as shown in Fig. 1. Currently, the homogeniser can only be cooled down to about 10 ^∘C, and this poses limitations to the environmental conditions which can be simulated in the laboratory; even though the aerosol entering the homogeniser can be preconditioned at a temperature down to about 5 ^∘C, the aerosol temperature at the outlet of the homogeniser will always be $\ge \mathrm{10}{}^{\circ }\mathrm{C}$.

The sampling zone is located 1.25 m downstream of the injection position and accommodates isokinetic sampling probes (funnels) placed at the bottom end of the homogeniser as illustrated in Fig. 2c. Isokinetic conditions are necessary when sampling with instruments operating at different flow rates to ensure representative sampling, e.g. by minimising sampling artefacts of larger particles. Several custom-made sampling probes with different cross sections have been therefore designed to match the flow rate of the various automated PM monitors, which typically ranges between 0.2 and 20 L min⁻¹. It is worth noting that the sampling system is highly adaptable; the lower end (outlet) of each sampling probe has custom-made threads so that it can be screwed in and out of the bottom metallic plate of the homogeniser. This ensures that the sampling probes can be readily exchanged before each experiment depending on the specifications of the PM monitors under testing. Finally, the excess aerosol flow exits the homogeniser through an exhaust outlet connected to a vacuum line as illustrated schematically in Fig. 1.

To characterise the aerosol homogeneity in the flow tube as a function of particle size, sodium chloride (NaCl) particles with a geometric mean mobility diameter of 50 nm and mineral dust particles with an aerodynamic diameter in the lower micrometre range (ISO A2 dust) were generated with a nebuliser and a rotating-brush generator, respectively, as described in Sect. 2.1. Two parallel sampling lines were inserted into the flow tube at the height where the sampling probes would be normally located; the position of the first sampling line was kept fixed at the centre of the flow tube (radial position 0), whereas the second one was placed consecutively at a distance $i=-\mathrm{70}$, −50, −30, −10, +10, +30, +50 and +70 mm with respect to the centre. The outlet of each sampling line was connected to a calibrated condensation particle counter (CPC; Models 3775 and 3776, respectively, TSI inc., USA). In total, concentration measurements at eight different positions along the diameter of the flow tube were performed. The particle number concentration measured at the centre was used as reference (C_ref=C₀), and the aerosol homogeneity was calculated as $C_i/C_{\text{ref}}$. The flow rate of each CPC was 0.3 L min⁻¹, and the inner diameter of the sampling line was 6 mm. This configuration ensured nearly isokinetic sampling.

The tests were performed with NaCl and mineral dust particles separately. In both cases the aerosol spatial homogeneity was found to be well within 3 % in number concentration as shown in Fig. 3a and b, respectively, indicating that the particle mixing characteristics do not depend on particle size in the tested range (i.e. from lower nanometre to lower micrometre range). A final test was performed by mixing NaCl and dust particles to investigate whether the particle mixing properties are affected when two primary aerosols are introduced into the homogeniser simultaneously. It was confirmed that the aerosol homogeneity remains well within ±3 % (measurements not shown), indicating that the simultaneous injection of primary aerosols into the homogeniser through separate ports (see Fig. 2b) does not compromise particle mixing in any way.

Figure 3Aerosol spatial homogeneity, η_hom=C_i/C_ref, at various radial positions along the diameter of the flow tube with (a) NaCl (sodium chloride) and (b) mineral dust particles as test aerosols. The measurements at positions $i=-\mathrm{10}$ and +10 mm were performed twice to assess measurement reproducibility. The error bars designate expanded uncertainties (95 % confidence level). These are type B uncertainties from the combined measurement uncertainties of the two CPCs and have no influence on the determination of homogeneity since they would shift all data points upwards or downwards by the same amount.

Download

Table 1Example of the uncertainty budget for a PM₁₀ mass concentration of 40 µg m⁻³ and a sampling time of 240 min.

^a The mass flow meter (Natec Sensors GmbH, Germany) was calibrated at METAS in a traceable manner. The expanded relative uncertainty on the calibration certificate amounts to 0.15 %. Here, a conservative estimation of 0.30 % was made to account for possible drifts since the time of calibration. ^b Assuming no loss of particulate mass during filter conditioning.

Download Print Version | Download XLSX

By calculating the standard deviation of all 28 measured data points, the spatial inhomogeneity of the aerosol in terms of number concentration was found to be 1.3 % for coverage factor k=1 (i.e. 68 % confidence level) or 2.6 % for k=2 (i.e. 95 % confidence level). This is used as an estimate for the uncertainty of the aerosol spatial homogeneity η_hom (see fourth row of Table 1). This is a crucial parameter which had not been evaluated so rigorously, if at all, in previous chamber studies (Hogrefe et al., 2004; Liu et al., 2017; Papapostolou et al., 2017; Schwab et al., 2004; Zhu et al., 2007).

2.3 Reference gravimetric method

The reference method used in this study for determining the PM₁₀ or PM_2.5 mass concentrations of particulate matter in the synthetic ambient aerosols is similar to the method described in the standard EN 12341:2014 (CEN/TC 264/WG-15, 2014); i.e. particulate matter was sampled on filters and weighed by means of a balance. The only major deviation from the requirements of the standard is the absence of any size-selective inlets upstream of the automatic PM samplers and the filter holder of the reference gravimetric method.

Briefly, model aerosols were drawn through 47 mm PTFE-coated glass fibre filters (Measurement Technology Laboratories, USA) placed in a metallic filter holder (C806 standard aerosol filter holder, Merck Millipore, Germany). The aerosol flow was controlled with a needle valve and measured with a calibrated mass flow meter (Natec Sensors GmbH, Germany) connected to an aerosol pump (VTE8, Thomas, Germany) in such a way that the volumetric flow corresponded to 2.3 m³ h⁻¹ at ambient conditions. Here, ambient conditions refer to the aerosol temperature and pressure in the homogeniser at the height of the sampling probes. In the EN 12341 standard, the requirement that the aerosol flow be set to 2.3 m³ h⁻¹ ($=\mathrm{38.33}\mathrm{L}{\min }^{-\mathrm{1}}$) at ambient conditions arises from the need to accurately define the size cut-off of the PM inlets, a property that depends on the inlet flow. Since the custom-made facility developed in this study aims at calibrating the PM monitors without their respective PM inlet, this flow requirement is here largely superfluous, apart from effects on sampling from the velocity of air through the filter. Nevertheless, during the experiments the aerosol flow was set to 2.3 m³ h⁻¹ at ambient conditions to facilitate comparison between the conventional field-based and the new laboratory-based procedures. The connecting tube between the isokinetic sampling probe (i.e. central sampling funnel in Fig. 2c) and the filter holder was made of inert, electrically conducting rubber material and was kept as short as possible (≈5 cm) without bends to minimise deposition losses of particulate matter by kinetic processes as well as losses due to thermal, chemical or electrostatic processes. Finally, the laboratory temperature and pressure were kept constant at (21±1) ^∘C and (950±20) hPa, respectively.

Before sampling, the filters were conditioned and weighed at the National Physical Laboratory (NPL) and shipped in individual plastic containers to the Federal Institute of Metrology METAS. After sampling, the filter samples were placed in Petri dishes, wrapped tightly in plastic covers and stored at 4 ^∘C for about a week. They were then shipped to NPL for conditioning and weighing. NPL use a Measurement Technology Laboratories robotic filter weighing system that comprises an environmental chamber ($\mathrm{20}{}^{\circ }\mathrm{C}\pm \mathrm{1}{}^{\circ }\mathrm{C}$ and 47.5 %±2.5 % relative humidity), an autohandler system and a Mettler Toledo XP2U balance. The filters are conditioned in the chamber for 48 h before weighing. The filters are weighed, and then the system pauses for 24 h before reweighing the filters to identify any time variation in filter mass. Numerous quality assurance and quality control checks are made before each set of weighings.

2.4 Uncertainty budget for the laboratory-based calibration of PM monitors

The reference mass concentration, C_m,ref, is given by the equation $C_{m,\text{ref}}={\mathit{\eta }}_{\text{hom}}\frac{m}{V}{P}_{\text{rel}}$, where η_hom is the aerosol homogeneity in the flow tube, m is the particulate mass collected on the filter and V is the sampled volume. V is given by the aerosol flow through the filter, Q, multiplied by the time duration of the measurement t. P_rel is defined as the relative particle penetration, $P_{\text{rel}}=P_{\text{DUT}}/P_{\text{ref}}$, where P_DUT and P_ref are the penetration through the sampling probe and connecting tube of the device under testing (DUT) and the reference method, respectively. The associated uncertainties are listed in Table 1.

Since sampling is carried out with isokinetic sampling probes and the tubes leading to the filter holder and the DUT are kept straight and as short as possible, particle losses are minimised. Penetration P_rel was set to 1; however, an uncertainty of 2 % was assigned to account for the higher impaction losses of supermicrometre particles in the sampling funnel of the reference method due to the higher sampling flow (von der Weiden et al., 2009). These losses are to some extent counteracted by the lower diffusion losses of submicrometre particles, which decrease with increasing sampling flow. Here, we followed a rather conservative approach and kept the uncertainty of P_rel at 2 %.