Field study of a soft X-ray aerosol neutralizer combined with electrostatic classifiers for nanoparticle size distribution measurements

Highlights

• A comparison between a soft X-ray (SXR) neutralizer and radioactive neutralizers was made.

• Particle penetration inside SXR neutralizer was parameterized with equivalent pipe length method.

• SXR neutralizer could be used with particle sizer spectrometers to replace radioactive one.

Abstract

Most conventional aerosol neutralizers are based on radioactive sources, which are controlled by strict regulations restricting their handling, transport, and storage. The TSI 3087 soft X-ray (SXR) neutralizer circumvents these legal restrictions. The aim of the present work is to compare the performance of a standalone SXR aerosol neutralizer with that of conventional radioactive aerosol neutralizers based on ⁸⁵Kr (TSI 3077) and ²⁴¹Am (Grimm 5522) by performing field tests in a real environmental scenario. The results obtained when the SXR neutralizer was connected to a mobility particle sizer spectrometer (MPS), different from the device suggested by the manufacturer, were comparable with those obtained with the use of radioactive aerosol neutralizers. In changing the neutralizer, the particle number concentrations, measured with the MPS connected to the SXR neutralizer, almost remained within the 10% uncertainty bounds for the particle size interval 10–300 nm, when diffusion losses inside the SXR tube were considered. Based on our comparisons, the SXR neutralizer can be regarded as a standalone instrument that could solve the problems associated with legal restrictions on radioactive neutralizers and fulfil the need for a portable instrument for different field test purposes.

Introduction

The mobility particle sizer spectrometer (MPS) is a widely used device for investigating the in situ evolution of nanoparticle size distributions (Justino, Rocha-Santos, & Duarte, 2011). Types of MPS include differential mobility particle sizers and scanning mobility particle spectrometers. These two systems are similar, except for the electric-field ramp applied in the measurements. Each system features a combination of a differential mobility analyzer (DMA) for sizing the distribution and a condensation particle counter (CPC) for counting the particle concentration.

MPS raw data, based on electrical mobility segregation, can be converted into a particle size classification only if the aerosol charge distribution at the inlet is known (Fuchs, 1963, Hoppel and Frick, 1986, Stolzenburg and McMurry, 2008, Wang and Flagan, 1990). This requirement is fulfilled with a diffusion charger, also known as a neutralizer, which brings the aerosol particles into a steady-state charge distribution, irrespective of their initial charge state (Adachi, Kousaka, & Okuyama, 1985; Cooper & Reist, 1973; Liu & Pui, 1974; Reischl, Mäkelä, Karch, & Necid, 1996).

Most conventional aerosol neutralizers are based on radioactive sources, such as ⁸⁵Kr, ²¹⁰Po, and ²⁴¹Am. However, the use of radioactive-based neutralizers is controlled by regulations restricting their handling, transport, and storage, even when the radioactive sources are sealed. To circumvent expensive safety measures and the legal restrictions imposed on the use of radioactive neutralizers, researchers have attempted to apply different physical principles to attain equilibrium charge distributions with non-radioactive ion sources.

Non-radioactive methods proposed to date include: corona discharge (Stommel & Riebel, 2004), photoelectric/UV-light sources (Shimada, Han, Okuyama, & Otani, 2002), and surface discharge microplasma (Kwon, Sakurai, Seto, & Kim, 2006). Among these methods, particle charging with soft X-ray (SXR) photoionizers has been widely applied. Shimada et al. (2002) observed that the particle charging probabilities were equal when using a SXR of <9.5 keV and an ²⁴¹Am foil source of 3.7 MBq at two different resident times (0.5 and 3 s). Lee, Kim, Shimada, and Okuyama (2005) compared the SXR neutralizer (<9.5 keV) with ²¹⁴Am bipolar neutralizers of approximately 3 and 6 MBq at 1.5 and 0.7 L/min, and observed almost equivalent concentrations of positive and negative ions generated by the two methods, particularly at low flow rates.

Shimada et al. (2002) and Lee et al. (2005) reported that the mobility distribution and total charging ratios for SXR and ²⁴¹Am neutralizers were similar, particularly for particles larger than 40 nm, in relation to the considered flow rate (1.5 L/min). Yoon, Bong, and Kim (2015) analyzed the neutral fraction at different flow rates, and reported effective performance of the SXR neutralizer for flow rates between 0.3 and 2 L/min, owing to the neutral fraction being close to that predicted by bipolar diffusion charging theory (Fuchs, 1963, Wiedensohler, 1988). At flow rates higher than 3 L/min, Yoon et al. (2015) reported that the neutral fraction of the SXR neutralizer did not follow the predictions of bipolar charging theory. At a flow rate of 0.3 L/min, the SXR neutral fraction was nearly the same as that of the ²⁴¹Am neutralizer; however, at a flow rate of 1 L/min the negatively charged fraction was higher than the positively charged fraction, although this difference was consistently less than 6%. They concluded that the particle charge distribution and the neutral fraction obtained using the SXR neutralizer agreed well with the prediction of the bipolar diffusion charging theory. Unfortunately, the authors did not provide details on the SXR and ²⁴¹Am neutralizers considered in their study.

Since these investigations, bipolar ion production systems based on SXR (<9.5 keV) sources have become commercially available. Such devices can be easily turned on and off, thereby increasing their safety and overall operating lifetime, and enabling more flexible regulations for their use. The first commercial unit was developed after the work of Shimada et al. (2002) and Yun, Lee, Iskandar, Okuyama, and Tajima (2009). A different unit was subsequently developed and sold by TSI (model 3087, TSI Inc., USA). The housing design for TSI 3087 is quite different from that reported by Shimada et al. (2002) and Yun et al. (2009). Furthermore, the TSI 3087 has a 0.4-mm aluminum foil installed to attenuate the SXR intensity and prevent potential particle generation inside the neutralizer owing to SXR photoionization.

The ion properties of the TSI 3087 neutralizer and its performance have been analyzed by Kallinger, Steiner, and Szymanski (2012) and Modesto-Lopez, Kettleson, and Biswas (2011), while the aerosol charge fraction downstream has been reported by Jiang et al. (2014). Their results showed excellent agreement of the measured spectra for the considered neutralizers. The same conclusions were reported by Kallinger and Szymanski (2015) who investigated the performance of four nanoparticle chargers, including an ²⁴¹Am-based neutralizer (60 MBq) and the TSI 3087.

The TSI 3087 datasheet also provides details of a comparison study with a ⁸⁵Kr-neutralizer (model 3077A, TSI), reporting that the particle size distributions (PSD) obtained with the two neutralizers had geometric means and geometric standard deviations within 5% and the total concentrations were within 10%–20% over the entire test matrix. These specifications apply only to the TSI instruments, because the X-ray neutralizer 3087 was specifically designed to interface with the TSI Electrostatic Classifier 3082. Nicosia, Belosi, and Vazquez (2014) first attempted to evaluate the combination of the SXR neutralizer with a Grimm MPS under controlled laboratory conditions. They showed a 0.998 correlation in the particle number concentration and differences in the distribution parameters (median and geometric standard deviation) were below 4% for polydisperse NaCl particles measured with a SMPS + C (series 5.400 Vienna type, long DMA, Grimm Aerosol Technik, Germany) equipped with either a SXR neutralizer (TSI model 3087) or an ²⁴¹Am radioactive source of 3.7 MBq (Grimm Aerosol Technik model 5.522). The TSI SXR neutralizer has also been used in combination with an MPS of a system different from the TSI in the “MPS-train project”, as part of the “CLOUD-T experiments” (Stolzenburg & Winkler, 2015). This experiment combined two TSI 3088 SXR neutralizers (an alternative version of the TSI 3087) with six Grimm nanoMPS to evaluate the growth of nanoparticles in a narrow size range below 20 nm. Nevertheless, the use of a SXR neutralizer as a standalone system coupled with MPS devices made by other manufacturers has yet to be validated.

The aim of the present work is to extend the performance of the SXR (TSI 3087) as a standalone aerosol neutralizer in combination with a Grimm MPS for field tests.