การพัฒนาปรับปรุงเปรียบเทียบและทดสอบมาตรฐานต้นแบบเครื่องวัดฝุ่นละอองในอากาศแบบใช้หลักการเชิงไฟฟ้าสถิตเพื่อไปสู่มาตรฐานเครื่องมือวัดสากล US EPA (พานิช อินต๊ะ,2562)

A manuscript to the Particuology to non-uniform electric field resulting from small surface scratches and imperfections. The needle cone angle was about 10o. The T-pipe electrode was made of stainless steel tube with inner diameter of about 6 mm. The PTFE insulator was an electrical isolator between needle electrode and T-pipe electrode. The PTFE insulator served to hold the needle electrode coaxial with the T-pipe electrode. The distance between the needle electrode and the T-pipe electrode apex was 6 mm. The length of the charging zone of the discharger was about 6 mm. The needle electrode could be screwed into the PTFE insulator to connect it to positive DC high voltage source (Model 602C-100P, Spellman High Voltage Electronics Corporation, Blvd Hauppauge, NY, USA), typically in the range between 2.8 and 3.2 kV, while the T-pipe electrode was grounded so as to produce the required corona discharge field. The discharge generates positive ions which move rapidly in the strong discharge field toward the wall of the T-pipe electrode. The positive polarity was set for the corona voltage in the developed discharger. This because of the production rate of ozone in the negative corona is one order of magnitude higher than in the positive corona (Chen & Davidson, 2003) and the negative corona revealed the fluctuations and dependence on surface conditions (English 1948). The discharge current, I , can be calculated by Townsend equation (Chang, Kelly & Crowley, 1995): I  aV (V Vc ) (1) where V is the applied voltage, Vc is the corona onset voltage and a is the dimensional constant depending on the inter-electrode distance, the needle electrode radius, the charge carrier mobility in the drift region and other geometrical factors. Therefore, the ion concentration, Ni , (ions/m3) as a function of the discharge current of the discharger

A manuscript to the Particuology can be estimated by (Intra, Yawootti & Rattanadecho, 2018): Ni  Ich (2) eZi Ech Ach where Ich is the discharge current of the discharger, e is the elementary charge (1.61  10-19 C), Zi is the electrical mobility of ions (1.4  10-4 m2/V s for the positive ion), Ech is the average electric field inside the discharger, and Ach is the inner surface area of the charging zone of the discharger where the discharge current is collected can be calculated by Ach  1.5 rL (3) where r is the radius of the T-pipe electrode and L is the length of the charging zone of the discharger. The inner surface area of the charging zone of the discharger was estimated as 8.5  10-5 m2. If the space charge effect is neglected, the electric field inside the corona discharger as a function of the distance between the corona-needle electrode tip and the T-pipe electrode is assumed to be given by Mason equation (Bamji, Bulinski & Prasad, 1993): E( )  (r  2   2 2V ln(1  4d / r) (4) / d)

A manuscript to the Particuology where V is the applied voltage, r is the radius of the needle electrode tip,  is the distance from the needle electrode tip along the z-axis and d is the distance between the needle electrode tip and T-pipe electrode. In the developed discharger, the average electric field in the charging zone was about between 8.89  105 and 1.02  106 V/m for applied voltage between 2.8 and 3.2 kV, respectively. At the needle electrode tip,   0 , therefore, the expression of the electric field in Equation (4) becomes (Mason, 1955): Emax  r 2V / r) (5) ln(1 4d Equation (5) is normally used in the literature for estimating the electric field at a needle electrode tip (Bamji, Bulinski & Prasad 1993; Mason, 1955). The electric field at the needle electrode tip in the charging zone of the developed discharger was about between 4.01  106 and 4.58  106 V/m for applied voltage between 2.8 and 3.2 kV, respectively. In this discharger, particle flow was directed across the corona discharge field and is charged by ion-to-particle collisions via diffusion charging and field charging mechanisms. The discharger performance is depended on the stable ion concentration ( Ni ), and the mean charging time of the particles to the ions ( t ) in the charging zone of the discharger, is given by t  r2L (6) Qp

A manuscript to the Particuology where Qp is the particle mass flow rate. In this work, the charging time of the discharger was about 17 ms when the particle mass flow rate was about 0.6 L/min with the Reynolds number of about 2,120. For this reason, a well-designed discharger should give a stable Nit product that can be accurately determined for any given operating conditions. As shown in Fig. 1, the ion trap zone removes the excess free positive ions mixing with the charged particles. This prevents contamination of the signal current to be measured by free ions potentially reaching the detector. In the ion trap zone, it consists of an ion trap electrode placed along the axis of a T-pipe electrode. The T-pipe electrode, a stainless steel tube, was about 6 mm in diameter and about 15 mm in length. The trap electrode, a stainless steel electrode, was 2.5 mm in diameter and 15 mm in length. For generating an electric field inside the ion trap zone for removal of the free positive ions, a positive DC voltage power source (Model 602C-100P, Spellman High Voltage Electronics Corporation, Blvd Hauppauge, NY, USA) was applied to the trap electrode, typically in the range between 200 and 300 V, while the T-pipe electrode is grounded. When the free positive ions with high electrical mobility are entered into the ion trap zone, the electrostatic force drives them toward the T-pipe electrode. They deposit on the wall of the T-pipe electrode, but most of the charged particles were not affected. The ion removal efficiency, trap , of the ion trap zone of the discharger can be estimated by Deutsch-Anderson equation as (Hinds, 1999; Intra, Limueadphai & Tippayawong, 2010) trap  Zi Etrap Atrap  (7)  1 exp   Qp 

A manuscript to the Particuology where Etrap is the average electric field inside the ion trap zone of the discharger and Atrap is the inner surface area of the ion trap zone of the discharger. Table 1 shows the comparison between our previous work by Intra & Tippayawong (2013) and this work in terms of the geometrical configuration of electrodes, the particle mass flow rate, the applied voltage range and the particle flow behavior. The differences between this discharger and our previous work by Intra & Tippayawong (2013) include; (i) concept of the this discharger was based on a simple design, compact, low cost and portable unit in order to reduce the production and operational costs (dimension of 44 mm  90 mm  25 mm and cost of less than 150 USD); (ii) T-pipe electrode configuration with short charging zone was used to reduce diffusion losses of the particles and eliminate the charged particle loss by electrostatic collection inside the discharger; and (iii) the ion trap was used to remove the excess free positive ions mixing with the charged particles in order to prevent the contamination of the measured signal current of free ions potentially reaching the detector. 3. Experimental Setup 3.1 Current-to-voltage characteristics Fig. 2 shows the experimental setup for evaluating the current-to-voltage characteristics of the developed discharger. It was consisted of a developed discharger, an adjustable DC high voltage power source and an electrometer. The positive voltage between 2.0 and 3.2 kV was applied to the needle electrode of the discharger with an adjustable commercial DC high voltage power source. In this experiment, no applied voltage on the trap electrode, therefore, the positive voltage of 0 V was maintained to

A manuscript to the Particuology the trap electrode by a second adjustable commercial DC voltage power source. The discharge currents in the discharge zone of the discharger were directly measured by the electrometer (Model 6517A, Keithley Instruments, Inc., Cleveland, OH, USA) via the T-pipe electrode which is electrical grounded. 3.2 Charging efficiency and particle losses determinations Fig. 3 shows the experimental setup for evaluating the particle charging efficiency and particle losses in the developed discharger. There are four indexes used for evaluating the particle charging performance of the developed discharger, including the intrinsic charging efficiency ( in ), the extrinsic charging efficiency ( ex ), the electrostatic loss ( Lel ) and the diffusion loss ( Ld ). The setup consisted of a developed discharger, an adjustable DC high voltage power source, an aerosol atomizer, a filtered air supply, an aerosol neutralizer, a concentration adjustment valves, a HEPA filter, a diffusion dryer, an electrostatic collector (ESC), an electrostatic classifier, an ultrafine condensation particle counter (UCPC). In this study, sodium chloride (NaCl) polydisperse particle was produced by nebulizing a NaCl solution (w/w 0.1 % in water) with an aerosol atomizer (Model 3076, TSI Inc., St. Paul, MN, USA) and a filtered air supply (Model 3074B, TSI Inc., St. Paul, MN, USA). The polydisperse particles coming out of the atomizer were still wet and were then dried to relative humidity less than about 30 % RH in the diffusion dryer (Model 3062, TSI Inc., St. Paul, MN, USA). It was well known that the generated polydisperse particles have some level of electric charge (Hinds, 1999). Therefore, some loss of particles due to electrostatic charges in the system was may be encountered, unless the particles are neutralized. Charged particles tend to deposit on tube walls and other surfaces. The soft X -ray aerosol

A manuscript to the Particuology neutralizer (Model 3088, TSI Inc., St. Paul, MN, USA) was used to neutralize the particle and to bring the particles to the Boltzman charge equilibrium. In this system, particle concentration can be changed by adjusting the concentration adjustment valves and a HEPA capsule filter (Model 1602051, TSI Inc., St. Paul, MN, USA). The mean diameter, concentration and geometric standard deviation of the generated particles were 59.8 nm, 4.54  105 particles/cm3 and 2.0, respectively, as shown in Fig. 4. NaCl polydisperse particles were then classified according to their electrical mobility using a soft X-ray aerosol neutralizer and an electrostatic classifier (Model 3082, TSI Inc., St. Paul, MN, USA) with a long-differential mobility analyzer (DMA), model 3081, TSI Inc., St. Paul, MN, USA, with clean air flow of 3.0 L/min allowing mobility diameter selection from 10 to 600 nm. The particles exiting the DMA at a given voltage were nearly singly charged monodisperse particles. The fraction of multiply charged large particles exiting the DMA was reduced to about 10% or less for all particles in the size range of 20 – 300 nm by having the particle size distribution of the test particle centered around a mode diameter of about 30 nm. The singly charged monodisperse particles were then passed through the second soft X-ray aerosol neutralizer and the 1st ESC to which a positive voltage of about 3.0 kV is applied by a DC high voltage power source to remove charged particles entirely. The uncharged monodisperse particles were obtained downstream from 1st ESC. Only uncharged particles were introduced into the discharger. After charged particle flow exited outside the discharger, it passed through the 2nd ESC. The charged aerosol flow was further entered into an UCPC (Model 3788, TSI Inc., St. Paul, MN, USA) to measure the particle concentration downstream of the developed discharger. In this study, both charging efficiencies and losses of the discharger were measured at the particle mass flow rate of about 0.6 L/min. The intrinsic charging

A manuscript to the Particuology efficiency can be calculated by (Ouf & Sillon, 2009): in 1 N1 / T (8) N2 where N1 is the number concentration of neutral particles downstream the 2nd ESC when the discharger and 2nd ESC voltages are on, N2 are the number concentration of total particles downstream the 2nd ESC when the discharger and 2nd ESC voltages are off and T is the transmission efficiency of neutral particles passing through the 2nd ESC. The extrinsic charging efficiency, which is the parameter of interest in a practical application, is defined as the fraction of particles exit outside the discharger carrying at least a unit of charge, and can be calculated by (Ouf & Sillon, 2009): ex  N3  N1 /T (9) N4 where N3 is the number concentration of charged particles downstream of the discharger when the discharger voltage is on and N4 is the number concentration of neutral particles upstream of the discharger. In this study, the electrostatic loss of particles, Lel , and diffusion loss, Ld , inside the discharger can be calculated by the following equations (Alonso, Martin & Alguacil, 2006): Lel  N2  N3 (10) N4

Ld 1 N2 A manuscript to the Particuology N4 (11) 4. Results and Discussion Fig. 5 shows the variations in discharge current and ion concentration with corona voltage of the developed discharger. The corona onset voltage of the developed discharger was about 2.4 kV. When the corona voltage increases from 2.4 kV to 3.2 kV, the discharge current will increase between 0.19 nA and 2.0 μA and the ion concentration will increase between 1.32  1011 ions/m3 and 1.03  1015 ions/m3, respectively. As a result, there is an ionization of the gas molecules and ignition of a corona discharge (Chang, Kelly & Crowley, 1995). As shown in Equation (2), the ion concentration inside the charging zone of the developed discharger can be controlled by the discharge current or the corona voltage. That is, an increasing corona voltage can lead to a higher discharge current; more ions were therefore generated from the surface of the needle electrode. It results in the possibility of an increasing intrinsic charging efficiency of the discharger. The electrical breakdown phenomena occurred for the applied voltages at larger than about 3.2 kV. Above this value, the discharge current was found to exhibit a fluctuation in an uncontrollable manner and no measurement could be made. It was shown that the developed discharger capable to operate stably at applied voltage ranging from 2.8 kV to 3.2 kV. In this work, the intrinsic charging efficiency was evaluated at different corona voltages. Fig. 6 shows the intrinsic charging efficiency of the discharger as a function of particle diameter in the range of about 20 nm to 300 nm at various corona voltages. The test particle mass flow rate was fixed at 0.6 L/min. Results show that the intrinsic

A manuscript to the Particuology charging efficiencies increase with increasing the corona voltages at a given particle size. This result is explained by an increase in the probability of collisions of particles with ions due to an increase in their concentration with the discharge current corresponding to the corona voltage (Efimov et al., 2018). At a given ion trap voltage, the intrinsic charging efficiencies increase from 74.09 to 94.40 %, 78.53 to 94.60 % and 79.90 to 95.08 % with particle sizes in the range of about 20 nm to 50 nm and decrease from 93.12 to 77.96 %, 93.36 to 76.91 % and 93.95 to 78.86 % with particle sizes in the range of about 150 nm to 300 nm for the applied voltage of 2.8, 3.0 and 3.2 kV, respectively. The intrinsic charging efficiency also reached a constant of ~93% for particles in the size range between 50 nm and 100 nm at a given corona voltage. It should be noted that the intrinsic charging efficiency of a unipolar discharger is primarily affected by the ion number concentration ( Ni ) and the particle residence time ( t ) in the charging zone. The ion concentration in the charging zone was controlled via the variation of discharge current only. At the particle mass flow rate of 0.6 L/min, the Nit values were about 1.79  1012 ions/m3 s and 1.76  1013 ions/m3 s at 2.8 kV and 3.2 kV applied voltages, respectively. However, the intrinsic charging efficiency can only present a part of the discharger performance. It is important for optimizing the discharger performance and to find out the optimal operating condition of the discharger. Fig. 7 shows the extrinsic charging efficiency of the discharger as a function of particle diameter in the range of about 20 nm to 300 nm at various corona voltages and ion trap voltage. The particle mass flow rate was fixed at 0.6 L/min. The extrinsic charging efficiency takes into account of the charged particle loss inside the discharger compared to the intrinsic charging efficiency. Results show that the extrinsic charging efficiency increases as the particle size increases. In addition, the higher extrinsic

A manuscript to the Particuology charging efficiency can be obtained with a sufficiently high corona voltage at an appropriate ion trap and particle mass flow rate. At the given corona voltage, the extrinsic charging efficiency decreases as the ion trap voltage increases. The best extrinsic charging efficiency of the discharger of about 20.80 % to 58.62% for particle diameter ranging from 20 nm to 300 nm is seen to occur at corona voltage and ion trap voltage of about 2.8 kV and 200 V, respectively. As shown in Fig. 6, when the corona voltages are increased to raise the intrinsic charging efficiencies, it’s important to let the charged particles pass through the discharger. The fraction of particle loss inside the discharger also needs to be reduced. Fig. 8 shows the electrostatic and diffusion losses of particles inside the discharger as a function of particle diameter in the size range of 20 nm to 300 nm at various corona voltages and ion trap voltage. The particle mass flow rate is fixed at about 0.6 L/min. For the electrostatic loss, the corona and ion trap voltages were varied in the range of about 2.8 kV to 3.2 kV, 0 – 300 V, respectively. As the given corona and ion trap voltages, larger particles were found to have lower electrostatic loss. The electrostatic loss increases as the corona and ion trap voltage increase. It was well known that the electrostatic loss depends on two competing factors; (i) it depends on how many particles can acquire a charge, which increases with particle size, and (ii) it also depends on the electrical mobility of the particles, which decreases with particle size (English, 1948). The highest electrostatic loss was observed to occur at particles with diameter of about 20 nm to be about 73.53, 83.66 and 53.98 % at the ion trap voltage of 300 V for the corona voltages of 2.8, 3.0 and 3.2 kV, respectively. For the diffusion loss, smaller particles were found to have higher diffusion loss due to Brownian diffusion effect on particle motion in the discharger than larger particles. The highest diffusion loss was seen to occur at particles with diameter of 10 nm to be about 18.9 %.

A manuscript to the Particuology 5. Conclusion In this paper, a corona discharger for unipolar charging of ultrafine particles has been developed and experimentally evaluated for the intrinsic and extrinsic particle charging efficiencies and the electrostatic and diffusion particle losses. The charging performance of the discharger was carried out under different operating conditions including corona and ion trap voltages for ultrafine particles in the size range of 20 nm to 300 nm. The applied voltage of the discharger ranged between 2.4 kV to 3.2 kV, corresponding to the discharge current from 0.19 nA to 2.0 μA and the ion concentration will increase from 1.32  1011 to 1.03  1015 ions/m3. Increasing the corona voltage can lead to a higher discharge current and ion concentration inside the discharger, leading to an increase in the intrinsic charging efficiency. In this discharger, the intrinsic charging efficiency was obtained between 74.09 % and 95.08 % for particles in the size range between 20 nm and 300 nm at a given corona and ion trap voltages. At the given corona voltage, the extrinsic charging efficiency decreases as the ion trap voltage increases. The best extrinsic charging efficiency of the discharger was found about 20.80 to 58.62% for particle diameter ranging from 20 nm to 300 nm at corona voltage and ion trap voltage of about 2.8 kV and 200 V, respectively. The highest electrostatic loss was observed to occur at particles with diameter of about 20 nm to be about 73.53, 83.66 and 53.98 % at the ion trap voltage of 300 V for the corona voltages of 2.8, 3.0 and 3.2 kV, respectively. Finally, the highest diffusion loss was seen to occur at particles with diameter of 10 nm to be about 18.9 %. In addition, the developed discharger has a simple design and robust operation that does not use additional ion-driving voltage, dilution or clean flows, and low cost system and easy to construct, which enables its use in a charging particles for electrical aerosol devices in the general applications of an airborne particulate monitoring.

A manuscript to the Particuology Acknowledgements The authors gratefully acknowledge the Electricity Generating Authority of Thailand (EGAT), Research contract no. GGR010100089000. The authors wish to thank Prof. Dr. Rainer Zawadzki of Governor State University for the valuable contribution during the preparation of the manuscript. References Alonso, M., Martin, M. I. & Alguacil, F. J. (2006). The measurement of charging efficiencies and losses of aerosol nanoparticles in a corona charger. Journal of Electrostatics, 64, 203 – 214. Bamji, S.S., Bulinski, A.T. & Prasad, K.M. (1993). Electric field calculations with the boundary element method. IEEE Transactions on Electrical Insulation, 28(3), 420 – 424. Chang, J., Kelly, A.J. & Crowley, J.M. (1995). Handbook of Electrostatic Processes. Marcel Dekker, Inc., New York. Chen, J. & Davidson, J. H. (2003). Ozone production in the negative DC corona: The dependence of discharge polarity. Plasma Chemistry and Plasma Processing, 23(3), 501 - 518. Efimov, A. A., Arsenov, P. V., Maeder, T. & Ivanov, V. V. (2018). Unipolar charging of aerosol particles in the size range 75-500 nm by needle-plate corona charger. Oriental Journal of Chemistry, 31(1), 214 – 221.

A manuscript to the Particuology English, W. N. (1948) Positive and negative point-to-plane corona in air. Physical Review, 74, 170 - 178. Flagan, R. C. (1998). History of electrical aerosol measurements. Aerosol Science and Technology, 28, 301 – 380. Hernandez-Sierra, A., Alguacil, F. J. & Alonso, M. (2003). Unipolar charging of nanometer aerosol particle in a corona ionizer. Journal of Aerosol Science, 34, 733 – 745. Hinds, W.C. (1999). Aerosol Technology, John Wiley & Sons, New York. Intra, P. & Tippayawong, N. (2006). Comparative study on electrical discharge and operation characteristics of needle and wire-cylinder corona chargers. Journal of Electrical Engineering & Technology, 1(4), 520 – 527. Intra, P. & Tippayawong, N. (2007). An overview of aerosol particle sensors for size distribution measurement. Maejo International Journal of Science and Technology, 1, 120 – 136. Intra, P. & Tippayawong, N. (2009). Progress in unipolar corona discharger designs for airborne particle charging: a literature review. Journal of Electrostatics, 67(4), 605 – 615. Intra, P. & Tippayawong, N. (2013). Development and evaluation of a high concentration, high penetration unipolar corona ionizer for electrostatic discharge and aerosol charging. Journal of Electrical Engineering & Technology, 8(5), 1175 – 1181.

A manuscript to the Particuology Intra, P., Limueadphai, P. & Tippayawong, N. (2010). Particulate emission reduction from biomass burning in small combustion systems with a multiple tubular electrostatic precipitator. Particulate Science and Technology, 28(6), 547 – 565. Intra, P., Yawootti, A. & Tippayawong, N. (2013). An electrostatic sensor for continuous monitoring of particulate air pollution. Korean Journal of Chemical Engineering, 30(12), 2205 – 2212. Intra, P., Yawootti, A. & Rattanadecho, P., (2018). Corona discharge characteristics and particle losses in a unipolar corona-needle charger obtained through numerical and experimental studies. Journal of Electrical Engineering & Technology, 12(5), 2021 - 2030. Lanki, T., Tikkanen, J., Janka, K., Taimisto, P. & Lehtimaki, M. (2011). An electrical sensor for long-term monitoring of ultrafine particles in workplaces. Journal of Physics: Conference Series, 304, 012013. Li, L., Chen, D.R. & Tsai, P.J. (2009). Use of an electrical aerosol detector (EAD) for nanoparticle size distribution measurement. Journal of Nanoparticle Research, 11(1), 111 – 120. Marquard, A., Meyer, J. & Kasper, G. (2006). Characterization of unipolar electrical aerosol chargers—Part II: Application of comparison criteria to various types of nanoaerosol charging devices. Journal of Aerosol Science, 37, 1069 – 1080. Mason, J. (1955). Breakdown of solid dielectrics in divergent fields. IEE Monograph 127: M. Medved, A., Dorman, F., Kaufman, S. L. & Pocher, A. (2000). A new corona-based charger for aerosol particles. Journal of Aerosol Science, 31, s616-s617.

A manuscript to the Particuology Murtomaa, M., Pekkala, P., Kalliohaka, T. & Paasi, J. (2005). A device for aerosol charge measurement and sampling. Journal of Electrostatics, 63(6 – 10), 571 – 575. Ouf, F.X. & Sillon, P. (2009). Charging efficiency of the Electrical Low Pressure Impactor's corona charger: Influence of the fractal morphology of nanoparticle aggregates and uncertainty analysis of experimental results. Aerosol Science and Technology, 43(7), 685 – 698. Park, D., An, M. & Hwang, J. (2007). Development and performance test of a unipolar diffusion charger for real-time measurements of submicron aerosol particles having a log-normal size distribution. Journal of Aerosol Science, 38(4), 420 – 430. Rostedt, A., Marjamaki, M., Yli-Ojanpera, J., Keskinen, J., Janka, K., Nienela, V., Ukkonen, A., et al. (2009). Non-Collecting Electrical Sensor for Particle Concentration Measurement. Aerosol and Air Quality Research, 9, 470 – 477. TSI Incorporated. (2002). Instruction Manual for Electrical Aerosol Detector Model 3070A, Minnesota, USA. Wei, J. (2007). Development of a method for measuring surface area concentration of ultrafine particles, D. Eng. Thesis, University of Duisburg-Essen, Germany. Whitby, K. T. (1961). Generator for producing high concentration of small ions. Review of Scientific Instruments, 32(12), 1351 – 1355.

A manuscript to the Particuology Table 1. Comparison between our previous work by Intra & Tippayawong (2013) and this work. Intra & Tippayawong (2013) This work Outer electrode configuration Cylindrical tube with tapered end T-pipe electrode Needle electrode diameter 3.0 mm 2 mm Needle cone angle 10o 10o Orifice diameter 3.5 mm - Electrodes distance 3.5 mm 6 mm Charging zone length 20 mm 6 mm Particle flow rate 1 – 5 L/min 0.6 – 5 L/min Corona voltage range 2.2 – 3.6 kV 2.8 – 3.2 kV Particle direction Circular Perpendicular Ion trap No Yes

A manuscript to the Particuology 2 mm Corona-needle Corona Voltage Ion Trap Voltage Electrode 10 mm PTFE Insulator 6 mm Electrode Tip 10o Ion Trap Electrode 6 mm Ion Trap Zone Aerosol Inlet Charging Zone Aerosol T-pipe Electrode Outlet Fig. 1. Schematic diagram of the developed corona discharger. DC High Voltage Power Source DC High Voltage Power Source (Spellman’s Bertan model 602C-100P) (Spellman’s Bertan model 602C-100P) Ion Trap Voltage Corona Voltage 0V 2.0 – 3.2 kV Test Discharger Electrometer (Keithley electrometer model 6517A) Fig. 2. Experimental setup for investigating the current-to-voltage characteristics of the developed discharger.

A manuscript to the Particuology Diffusion Dryer Soft X-ray HEPA filter TSI Model 3062 Aerosol Neutralizer Concentration TSI Model 3088 Adjustment Valves Compressed Air HEPA filter Filtered Air Supply TSI Model 3074B Constant Output Atomizer Soft X-ray Electrostatic Classifier TSI Model 3076 Aerosol Neutralizer TSI Model 3082 DC High Voltage Power Source TSI Model 3088 (Spellman’s Bertan model 602C-100P) DC High Voltage Power Source (Spellman’s Bertan model 602C-100P) Ion Trap Voltage DC High Voltage Power Source (Spellman’s Bertan model 602C-100P) Corona Voltage Test Discharger Test particles Soft X-ray (electrically neutral) 1st Electrostatic Aerosol Neutralizer 2nd Electrostatic Collector Collector TSI Model 3088 Condensation Particle Counter TSI Model 3788 DC High Voltage Power Source (Spellman’s Bertan model 602C-100P) Fig. 3. Schematic of the experimental setup for the measurement of particle loss, intrinsic and extrinsic charging efficiency of the developed discharger.

A manuscript to the Particuology particle concentration, dNp/dlogdp, cm-3 8.0x105 NaCl particles 6.0x105 4.0x105 Mean diameter : 59.8 nm 2.0x105 Mode diameter : 31.1 nm Geometric standard deviation : 2.0 0.0 Total number concentration : 4.54 x 105 cm-3 10 100 1000 particle diameter, nm Fig. 4. Nanoparticle concentration and size distribution of sodium chloride particles generated from an atomizer. discharge current, A 1015 ion number concentration, ions/m3 2.0 1.5 1014 1.0 1013 0.5 1012 0.0 2.6 2.8 3.0 1011 2.4 3.2 corona voltage, kV Fig. 5. Variations in discharge current and ion number concentration with corona voltage of the developed discharger.

A manuscript to the Particuology intrinsic charging efficiency, % 100 Corona voltage 80 2.8 kV 60 3.0 kV 40 3.2 kV 20 0 100 10 particle diameter, nm Fig. 6. Intrinsic charging efficiency of the discharger as a function of particle diameter. extrinsic charging efficiency, % 80 2.8 kV Corona voltage 60 40 Ion trap voltage 20 300 V 250 V 200 V 0V 0 10 100 particle diameter, nm (a) 2.8 kV

extrinsic charging efficiency, % A manuscript to the Particuology extrinsic charging efficiency, %80 3.0 kV Corona voltage Ion trap voltage 300 V 60 250 V 200 V 0V 40 20 0 10 100 particle diameter, nm (b) 3.0 kV 80 3.2 kV Corona voltage Ion trap voltage 300 V 60 250 V 200 V 0V 40 20 0 10 100 particle diameter, nm (c) 3.2 kV Fig. 7. Extrinsic charging efficiency of the discharger as a function of particle diameter.

A manuscript to the Particuology 100 2.8 kV Corona voltage Ion trap voltage electrostatic loss, 300 V 80 electrostatic loss, 250 V electrostatic loss, 200 V electrostatic loss, 0 V diffusion loss particle loss, % 60 40 20 0 10 100 particle diameter, nm (a) 2.8 kV 100 3.0 kV Corona voltage Ion trap voltage electrostatic loss, 300 V 80 electrostatic loss, 250 V electrostatic loss, 200 V electrostatic loss, 0 V diffusion loss particle loss, % 60 40 20 0 10 100 particle diameter, nm (b) 3.0 kV

A manuscript to the Particuology 100 3.2 kV Corona voltage Ion trap voltage electrostatic loss, 300 V 80 electrostatic loss, 250 V electrostatic loss, 200 V electrostatic loss, 0 V diffusion loss particle loss, % 60 40 20 0 10 100 particle diameter, nm (c) 3.2 kV Fig. 8. Particle losses in the discharger as a function of particle diameter.

Declaration of Interest Statement Declaration of interests ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

Journal of Electrical Engineering & Technology Development and performance evaluation of a fast response particulate detector for measurement of particle number concentration --Manuscript Draft-- Manuscript Number: Development and performance evaluation of a fast response particulate detector for Full Title: measurement of particle number concentration Article Type: Original Article Manuscript Classifications: 300: Electrophysics and Applications; 300.020: High Power, High Voltage and Section/Category: Discharge Funding Information: Electrophysics and Applications Electricity Generating Authority of Assoc.Prof.Dr Panich Intra Thailand (GGR010100089000) Abstract: The objective of this work is to develop and evaluate a fast response particulate detector for measurement of particle number concentration. The developed detector was experimentally evaluated the particle counting efficiency of sodium chloride (NaCl) and Di-Ethyl-Hexyl-Sebacat (DEHS) particles and compared with an ultrafine condensation particle counter (UCPC). It was shown that the particle number concentrations of the UCPC were in good agreement with the electrometer voltage of charged particles measured by developed particulate detector over a wide range, R2 of 0.998 and 0.999, and a slope of 0.003 and 0.002 for DEHS and NaCl particles, respectively. The time response for the particle number concentration changing from zero to slightly steady-state value was found to be in the order of about 750, 500 and 250 ms for the aerosol flow rate of 1.5, 3.0 and 6.0 L/min, respectively. It was found that the particle counting efficiencies of the developed particulate detector were about 99.60 to 100 and 99.55 to 100 for the DEHS and NaCl particles in the diameter range between 50 and 500 nm, respectively. Finally, this shows that this developed particulate detector proved particularly useful for measuring and detecting particulate air pollution, for number concentration of particles in the range between 0 and 104 particles/cm3 with a time response of less than 1 s. Corresponding Author: Panich Intra, Ph.D. RUEE RMUTL THAILAND Corresponding Author Secondary Information: Corresponding Author's Institution: RUEE RMUTL Corresponding Author's Secondary Institution: First Author: Panich Intra, Ph.D. First Author Secondary Information: Order of Authors: Panich Intra, Ph.D. Paisarn Wanusbodeepaisarn Thanesvorn Siri-achawawath Order of Authors Secondary Information: Author Comments: The Editor, Journal of Electrical Engineering & Technology 13 December 2019 Dear the Editor Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation

Enclosed are the typed original and the complete paper files of a manuscript by Panich Intra, Paisarn Wanusbodeepaisarn and Thanesvorn Siri-achawawath titled “Development and performance evaluation of a fast response particulate detector for measurement of particle number concentration” which is being submitted for consideration and possible publication in the Journal of Electrical Engineering & Technology. Any suggestion for improvement of the manuscript is welcomed. Yours sincerely, Assoc. Prof. Dr. Panich Intra Correspondence for manuscript, please contact: Assoc. Prof. Dr. Panich Intra Research Unit of Applied Electric Field in Engineering, College of Integrated Science and Technology, Rajamangala University of Technology Lanna, Chiang Mai 50220, Thailand Phone: +66-89755-1985 E-mail: [email protected] Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation

Title Page A manuscript to the Journal of Electrical Engineering & Technology, Springer Development and performance evaluation of a fast response particulate detector for measurement of particle number concentration Panich Intraa,* Paisarn Wanusbodeepaisarnb and Thanesvorn Siri-achawawathc aResearch Unit of Applied Electric Field in Engineering, College of Integrated Science and Technology, Rajamangala University of Technology Lanna, Chiang Mai 50220, Thailand bPico Innovation Co., Ltd., 336 M. 4, Yang Noeng, Saraphi, Chiang Mai 50140, Thailand cInnovative Instrument Co., Ltd., 7/139 Suntinakorn 11 Alley, Bang Kaeo, Bang Phli, Samut Prakan, Bangkok 10540, Thailand *Corresponding Author: E-mail: [email protected] Abstract The objective of this work is to develop and evaluate a fast response particulate detector for measurement of particle number concentration. The developed detector was experimentally evaluated the particle counting efficiency of sodium chloride (NaCl) and Di-Ethyl-Hexyl-Sebacat (DEHS) particles and compared with an ultrafine condensation particle counter (UCPC). It was shown that the particle number concentrations of the UCPC were in good agreement with the electrometer voltage of charged particles measured by developed particulate detector over a wide range, R2 of 0.998 and 0.999, and a slope of 0.003 and 0.002 for DEHS and NaCl particles, respectively. The time response for the particle number concentration changing from zero to slightly steady-state value was found to be in the order of about 750, 500 and 250 ms for the aerosol flow rate of 1.5, 3.0 and 6.0 L/min, respectively. It was found that the particle counting efficiencies of the developed particulate detector were about 99.60 to 100 and 99.55 to 100 for the DEHS and NaCl particles in the diameter range between 50 and 500 nm, respectively. Finally, this shows that this developed particulate detector proved particularly useful for measuring and detecting particulate air pollution, for number concentration of particles in the range between 0 and 104 particles/cm3 with a time response of less than 1 s. Keywords: Aerosol, Particulate, Faraday Cup, Electrometer, Fast Response

Blinded Manuscript Click here to access/download;Blinded Click here to view linked References Manuscript;MS_JEET_EM_Intra_et_al2019.pdf A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 1. Introduction 2 3 Aerosol sensors are important in particulate science and technology as well as in research and practical field 4 5 work for source characterization and particulate air pollution monitoring. Progress in developing measurement 6 7 methods and devices has been rapid and a large number of measurement methods and devices have become available 8 9 [1]. Electrical methods are widely used for measuring number concentration and size distribution of airborne aerosol 10 11 particles in the most cases. This method depends on the electrical properties of aerosol particles i.e. electrostatic 12 13 charge and force, and electrical mobility [2]. Advantage of the electrical methods is that it gives fast time response 14 15 monitoring of the particle number concentrations. 16 17 Typically, the electrical aerosol device consists of two key components: one for particle charging and the 18 19 other for measuring electric current on charged aerosols. An aerosol electrometer was used for this measurement. 20 21 The readout of a device depends on the net charge level on aerosol particles. Flagan [2] were reviewed the recent 22 23 developments in electrical aerosol instrumentations. The previous studies concerned themselves with measurements 24 25 of a nanoparticle number and surface area concentration [3], ambient ion and aerosol charge measurement [4], 26 27 aerosol charge measurement and dust sampling [5 – 6], aerosol integral parameter measurement [7], number 28 29 concentration and size distribution measurements [8], engine exhaust particle emission measurement [9], monitoring 30 31 of ultrafine particles in workplaces [10], and continuous ambient particulate air pollution measurement [11 – 12]. 32 33 Commercial devices that were designed for measuring net charge level on aerosol particles for example the TSI 34 35 Model 3070A Electrical Aerosol Detector [13]. These devices are most suitable for site-specific, high sensitivity and 36 37 resolution, but these devices were large in size, high cost, and complex system. In our previous work [4, 6, 11], an 38 39 aerosol electrometer was developed and theoretically and experimentally evaluated for ion and aerosol charge 40 41 42 measurements. The measured electric current range of this aerosol electrometer was between about 1 and 500 pA, 43 corresponding to number concentrations of ions and aerosol charges in the range between about 1011 and 1014 per m3. 44 45 46 However, in particulate air pollution studies typically the number concentration of aerosol particles should be less 47 than 1011 per m3. Therefore, an attempt was made to widen the measurement range of the number concentration, 48 49 50 resolution and sensitivity of aerosol as well as the time response of the of the aerosol electrometer of the previous 51 52 work. 53 54 Therefore, a fast response particulate detector was developed and constructed, suitable for measurement of 55 56 number concentration of airborne aerosol particles in this study. This paper introduces and discusses the detailed 57 58 description of the operating principle of the particulate detector as well as the performance evaluation of the detector. 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 A prototype of the fast response particulate detector was experimentally evaluated the particle counting efficiency of 2 3 NaCl and DEHS particles and compared with an UCPC at different operating conditions. 4 5 6 7 2. Materials and Methods 8 9 2.1 Description of the fast response particulate detector 10 11 Fig. 1 shows the schematic diagram of the fast response particulate detector. This apparatus is typically a 12 13 metal casing with a metal filter for collecting the charged particle. It mainly consisted of a metal casing, a filter 14 15 holder, a metal filter, an insulator, and electrometer circuit. In order to measure ultra-low current of charged particles 16 17 less than 1 pA, a metal casing was made of brass and connected to ground for shielding the metal filter. The metal 18 19 casing helped for completely shielding noise due to the electromagnetic and radio-frequency interferences from an 20 21 external source. The filter holder was made of a brass tube in a diameter of 52 mm and electrically isolated from the 22 23 metal housing by a Teflon® insulator which eliminates noise due to leakage currents. The metal filter in a diameter of 24 25 47 mm was housed in the filter holder. In this design, the connecting rod was connected between the filter holder and 26 27 the electrometer circuit input to transfer particle charges collected on the filter to an electrometer circuit that was 28 29 inside the electrometer circuit casing. The electric current measured by the electrometer circuit is proportional to the 30 31 collection rate of the charged particle on the filter. If the number of charges per particle, np , and the volumetric 32 33 34 aerosol flow rate, Qa , are known, the particle number concentration, N p , can be estimated from the measured 35 36 current of particle charges, I p , by a following equation as 37 38 39 40 41 Np  Ip (1) 42 n p eQa 43 44 45 46 47 where e is the elementary unit of charge. 48 49 A schematic diagram of an electrometer circuit is shown in Fig. 2. This circuit is a current-to-voltage 50 51 converter, where the voltage drop caused by an electric current flowing through a feedback resistor is measured. A 52 53 500 G resistor with a percent error of about 10 % (Model RX-1M5009KE, OHMITE, USA) was used for the 54 55 feedback resistor in this circuit. In this circuit, a commercially-available monolithic operational amplifier (LMC662, 56 57 Texas Instruments Incorporated, Texas, USA) was used. This was designed for ultra-low current measurement and 58 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 featured a ultra-low input bias current of about 2 fA and a low offset voltage drift of about 1.3 μV/oC. The output 2 3 voltage of this circuit, Vo , can be calculated by 4 5 6 7 8 Vo  AeVi (2) 9 10 11 12 13 14 where 15 16 17 Ae  Rf / Ri , (3) 18 19 20 21 22 23 (4) 24 Vi  Ii Ri , 25 26 27 28 29 30 Ae is the amplifier gain, Rf is the feedback resistor, Ri is the input resistor, Vi is the input voltage and Ii is the 31 32 input current. Substituting Eqs. (3) and (4) into Eq. (2), the output voltage is given by 33 34 35 36 37 38 Vo  Rf Ii (5) 39 40 41 42 43 44 This circuit has an output voltage ranging between 0 and 10 V corresponding to an input current ranging between 0 45 46 and 20 pA, typically 5 mV per 10 fA. In this work, the performance of the electrometer circuit was compared with a 47 48 high-impedance current source. Good agreement was found from the comparison, i.e. the average differences were 49 50 about 7.4 % and the measured current ratio was in the range between 1.07 and 1.09 []. Other components in this 51 52 circuit were primarily for input protection (10 MΩ resistor) and RC low-pass filtering (33 kΩ resistor and 1µF 53 54 capacitor), used to reduce high-frequency noise and to prevent oscillations of the amplifier output. The cut-off 55 56 frequency and the time constant of this circuit were about 4.82 Hz and 0.033 s, respectively. According to the Eq. 57 58 (5), the measured current of charged particles can be calculated by 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 I p  Vo / Rf (6) 2 3 4 5 6 7 Substituting Eq. (6) into Eq. (1), the number concentration of charged particles as a function of the output voltage is 8 9 given by 10 11 12 13 Np  Vo (7) 14 Rf npeQa 15 16 17 18 19 2.2 Experimental setup 20 21 Fig. 3 shows the experimental setup for evaluating the performance of the particulate detector. The setup 22 23 mainly consisted of a filtered air supply, an aerosol atomizer, a diffusion dryer, an aerosol neutralizer, an electrostatic 24 25 classifier, an ultrafine condensation particle counter (UCPC) and a developed particulate detector. In this work, a 26 27 sodium chloride (NaCl) and Di-Ethyl-Hexyl-Sebacat (DEHS) aerosols were used for evaluation of the particle 28 29 counting efficiency of solid and liquid particles. The NaCl and DEHS aerosols generated by atomizing a NaCl 30 31 solution (w/w 0.1 % in water) and a 0.6% DEHS solution in isopropyl alcohol with an aerosol atomizer (Model 3076, 32 33 TSI Inc., St. Paul, MN, USA) and a filtered air supply (Model 3074B, TSI Inc., St. Paul, MN, USA). The 34 35 polydisperse aerosol of NaCl or DEHS coming out of the aerosol atomizer were then delivered to a diffusion dryer 36 37 (Model 3062, TSI Inc., St. Paul, MN, USA) for water removing. It was well known that the generated polydisperse 38 39 aerosol of NaCl or DEHS have some level of electric charge []. Therefore, some loss of particles due to electrostatic 40 41 charges in the system was may be encountered, unless the particles are neutralized. Charged particles of NaCl or 42 43 DEHS tend to deposit on tube walls and other surfaces. The soft X-ray aerosol neutralizer (Model 3088, TSI Inc., St. 44 45 Paul, MN, USA) was used to neutralize the particle and to bring the particles to the Boltzman charge equilibrium. In 46 47 this system, the particle number concentration could be regulated by the concentration adjustment valves and a 48 49 HEPA filter. polydisperse particles of NaCl or DEHS were then classified according to their electrical mobility using 50 51 52 the electrostatic classifier (Model 3082, TSI Inc., St. Paul, MN, USA) with a long-differential mobility analyzer, long 53 54 DMA (model 3081, TSI Inc., St. Paul, MN, USA), allowing for a mobility diameter selection from 50 to 500 nm. 55 56 The monodisperse particles of NaCl or DEHS exiting the DMA were nearly singly-charged, which was mixed with 57 58 makeup flow and then sampled by an UCPC (Model 3788, TSI Inc., St. Paul, MN, USA) and the developed 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 particulate detector with a flow splitter (Model 3708, TSI Inc., St. Paul, MN, USA). The aerosol flow rate of the 2 3 UCPC and the developed particulate detector were both set to be 1.5 L/min. The flow paths were symmetrical to 4 5 keep the same particle diffusion losses. It means that the aerosol flow rates and conductive silicone tube lengths from 6 7 the flow splitter to the UCPC and the developed particulate detector inlet are the same. The output voltage from the 8 9 developed particulate detector was measured and recorded by a highly-accurate digital voltmeter and oscilloscope. 10 11 The output voltage reading was then translated into the measurement of particle number concentration via Eq. 7. 12 13 14 15 4. Results and Discussion 16 17 Fig. 4 shows the warm-up time and zero offset of the developed particulate detector. It was shown that the 18 19 electrometer voltage of the developed particulate detector was decreased from 25 mV to 0 within 90 min. It was 20 21 mean that the warm-up time of the developed detector could be about 90 min. For zero offset of the developed 22 23 particulate detector, a standard deviation was about 0.870. It was well known that the stability of the particulate 24 25 detector was dependant on the environment including temperature and humidity. Vibrations and fluctuations in the 26 27 power source can lead to discrepancies in the measurements. Fig. 5 shows the comparison between the particle 28 29 number concentration measured by the UCPC and the electrometer voltage measured by the developed particulate 30 31 detector for 100 nm DEHS and NaCl particles. Note that the UCPC has 100% counting efficiency for 100 nm NaCl 32 33 particles. Particle number concentrations were in the range of 0 to 1.23 × 104 particles/cm3 and 0 to 9.90 × 103 34 35 particles/cm3, corresponding to the electrometer voltages were in the range of 0 to 38 mV and 0 to 26.5 mV for 36 37 DEHS and NaCl particles, respectively. It was shown that the DEHS and NaCl particle number concentrations of the 38 39 UCPC are in good agreement with the electrometer voltage of DEHS and NaCl particles measured by developed 40 41 42 particulate detector over a wide range. 43 44 Fig. 6 shows the linear correlations between the particle number concentrations measured by the UCPC and 45 46 the electrometer voltage measured by the developed particulate detector for 100 nm DEHS and NaCl particles. For 47 48 the DEHS particle, the electrometer voltage was in range from 0 to 31 mV, corresponding to the particle number 49 concentration was in range from 0 to 1.0 × 104 particles/cm3. The electrometer voltage was in range from 0 to 28 50 51 52 mV, corresponding to the particle number concentration was in range from 0 to 1.0 × 104 particles/cm3 for the NaCl 53 54 particle. The comparison between the measured particle number concentration by the UCPC and the measured 55 56 electrometer voltage by the developed particulate detector data resulted in R2 of 0.998 and 0.999, and a slope of 57 58 0.003 and 0.002 for DEHS and NaCl particles, respectively. At the same particle number concentration, the 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 measured electrometer voltage of the DEHS particle has higher than the measured electrometer voltage of the NaCl 2 3 particle. This is because the difference in the particle density and the dielectric constant of material between DEHS 4 5 and NaCl particles. Note that the particle density and the dielectric constant of the DEHS and NaCl particles were 6 7 0.912 g/cm3 and 2.170 g/cm3, and 5.1 and 3.21, respectively. It was evident that the intrinsic charged fraction and 8 9 mean charge per particle differences among different dielectric constant of particle materials. 10 11 Fig. 7 shows the ratio of the particle number concentration to the electrometer voltage as a function of 12 13 particle diameter in the range from 50 nm to 500 nm for DEHS and NaCl particles. The ratio of the particle number 14 15 concentration to the electrometer voltage increased with increasing particle size for the DEHS particles. The ratio of 16 17 the particle number concentration to the electrometer voltage of the DEHS particles was about 328.12, 341.47, 18 19 362.38, 404.44, 436.84 and 457.77 for particle diameter of 50, 100, 200, 300, 400 and 500 nm, respectively. For the 20 21 NaCl particles, the ratio of the particle number concentration to the electrometer voltage decreased with increasing 22 23 particle size in the range between 50 nm and 200 nm and reached a constant for particles in the size range between 24 25 300 nm and 500 nm. The ratio of the particle number concentration to the electrometer voltage of the NaCl particles 26 27 was about 464.83, 382.60, 387.32, 398.09, 400 and 408 for particle diameter of 50, 100, 200, 300, 400 and 500 nm, 28 29 respectively. 30 31 The time response, defined as sample transport delay time between the aerosol sample inlet and the metal 32 33 filter of the developed particulate detector, was evaluated by means of the step change in the number concentration 34 35 of the incoming aerosol sample. Fig. 8 shows the time response of the developed particulate detector to step changes 36 37 in number concentration of the aerosol sample at different operating aerosol flow rate of 1.5, 3.0 and 6.0 L/min, 38 39 respectively. As shown in Fig. 8, the time response for the particle number concentration changing from zero to 40 41 42 slightly steady-state value was in the order of about 750, 500 and 250 ms for the aerosol flow rate of 1.5, 3.0 and 6.0 43 44 L/min, respectively. It was shown that the time response of the particle number concentration changing from steady- 45 46 state value to zero was about 1500, 1250 and 1000 ms for the aerosol flow rate of 1.5, 3.0 and 6.0 L/min, 47 48 respectively. Results show that the time response of the developed particulate detector increases with increasing the 49 50 aerosol flow rate, ways to improve this instrument time response may be done by increasing the aerosol flow rate in 51 52 of the developed particulate detector. 53 54 Fig. 9 shows the response time curves to a step change in number concentration of the UCPC and the 55 56 developed particulate detector for 100 nm DEHS and NaCl particles at the aerosol flow rate of about 1.5 L/min. It 57 58 was shown that the response time curves to a step change in number concentration of the developed particulate 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 detector are in good agreement with the response times of the UCPC for both DEHS and NaCl particles. Fig. 10 2 3 shows the results of evaluating the particle counting efficiency for DEHS and NaCl particles in the diameter range 4 5 from 50 nm to 500 nm at the aerosol flow rate of about 1.5 L/min. It was shown that the particle counting efficiencies 6 7 of the developed particulate detector were about 99.60 to 100 and 99.55 to 100 for the DEHS and NaCl particles in 8 9 the diameter range between 50 and 500 nm, respectively. 10 11 12 13 5. Conclusion 14 15 In this paper, a particulate detector for measuring particle number concentration with fast response was 16 17 developed and evaluated. This detector was typically a metal casing with a metal filter for collecting the charged 18 19 particle. The detector was evaluated experimentally the particle counting efficiency of NaCl and DEHS particles and 20 21 the time response and compared with an UCPC for particle diameter of 50, 100, 200, 300, 400 and 500 nm, 22 23 respectively. It was shown that the DEHS and NaCl particle number concentrations of the UCPC are in good 24 25 agreement with the electrometer voltage of DEHS and NaCl particles measured by developed particulate detector 26 27 over a wide range, R2 of 0.998 and 0.999, and a slope of 0.003 and 0.002 for DEHS and NaCl particles, respectively. 28 29 The time response for the particle number concentration changing from zero to slightly steady-state value was found 30 31 to be in the order of about 750, 500 and 250 ms for the aerosol flow rate of 1.5, 3.0 and 6.0 L/min, respectively. It 32 33 was found that the particle counting efficiencies of the developed particulate detector were about 99.60 to 100 and 34 35 99.55 to 100 for the DEHS and NaCl particles in the diameter range between 50 and 500 nm, respectively. Finally, 36 37 this shows this developed detector demonstrated particularly useful for measuring number concentration of particles 38 39 in the range between 100 and 104 particles/cm3 with a time response of less than 1 s. 40 41 42 43 44 Acknowledgement 45 46 The authors gratefully acknowledge the Electricity Generating Authority of Thailand (EGAT), Research 47 48 contract no. GGR010100089000. The authors wish to thank Prof. Dr. Rainer Zawadzki of Central New Mexico 49 50 Community College for the valuable contribution during the preparation of the manuscript. 51 52 53 54 References 55 56 [1] Hinds, W.C. (1999). Aerosol Technology. John Wiley & Sons, New York. 57 58 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 [2] Flagan, R. C. (1998). History of electrical aerosol measurements. Aerosol Science and Technology, 28, 301 – 2 3 380. doi: 10.1080/02786829808965530. 4 5 [3] Wei, J. (2007). Development of a Method for Measuring Surface Area Concentration of Ultrafine Particles, D. 6 7 Eng. Thesis, University of Duisburg-Essen, Germany. 8 9 [4] Intra, P. & Tippayawong, N. (2011). Performance evaluation of an electrometer system for ion and aerosol charge 10 11 measurements. Korean Journal of Chemical Engineering, 28(2), 527 – 530. 12 13 [5] Murtomaa, M., Pekkala, P., Kalliohaka, T. & Paasi, J. (2005). A device for aerosol charge measurement and 14 15 sampling. Journal of Electrostatics, 63(6 – 10), 571 – 575. 16 17 [6] Intra, P. & Tippayawong, N. (2015). Development and evaluation of a Faraday cup electrometer for measuring 18 19 and sampling atmospheric ions and charged aerosols. Particulate Science and Technology, 33(3), 257 – 263. 20 21 [7] J-Fatokun, F.O., Morawska, L., Jamriska, M. & Jayaratne, E.R. (2008). Application of aerosol electrometer for 22 23 ambient particle charge measurements. Atmospheric Environment, 42, 8827 – 8830. 24 25 [8] Li, L., Chen, D.R. & Tsai, P.J. (2009). Use of an electrical aerosol detector (EAD) for nanoparticle size 26 27 distribution measurement. Journal of Nanoparticle Research, 11(1), 111 – 120. 28 29 [9] Rostedt, A., Marjamaki, M., Yli-Ojanpera, J., Keskinen, J., Janka, K., Nienela, V., Ukkonen, A., et al. (2009). 30 31 Non-collecting electrical sensor for particle concentration measurement. Aerosol and Air Quality Research, 9, 32 33 470 – 477. 34 35 [10] Lanki, T., Tikkanen, J., Janka, K., Taimisto, P. & Lehtimaki, M. (2011). An electrical sensor for long-term 36 37 monitoring of ultrafine particles in workplaces. Journal of Physics: Conference Series, 304, 012013. 38 39 [11] Intra, P., Yawootti, A. & Tippayawong, N. (2013). An electrostatic sensor for continuous monitoring of 40 41 particulate air pollution. Korean Journal of Chemical Engineering, 30(12), 2205 – 2212. 42 43 44 [12] Yawootti, A., Intra, P., Tippayawong, N. & Sampattagul, S. (2015). Field evaluation of an electrostatic PM10 45 46 mass monitor used for continuous ambient particulate air pollution measurements. Journal of Electrostatics, 78 47 (1), 46 – 54. 48 49 50 [13] TSI Incorporated. (2002). Instruction Manual for Electrical Aerosol Detector Model 3070A, Minnesota, USA. 51 52 [14] National Semiconductor Corporation. (2003). LMC662 CMOS Dual Operational Amplifier Data Sheet. 53 54 [15] Chang, J., Kelly, A.J., & Crowley, J.M. (1995). Handbook of Electrostatic Processes. Marcel Dekker, Inc., New 55 56 York. 57 58 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 Aerosol Inlet 2 3 Metal Filter 4 5 Metal Housing Filter Holder 6 7 8 9 10 11 12 13 Aerosol Outlet Teflon® 14 Insulator 15 16 17 18 Power Input and Signal Output Port 19 Connecting Electrode 20 To Electrometer Circuit 21 22 23 Electrometer Circuit Housing 24 25 26 27 Fig. 1. Schematic diagram of the fast response particulate detector. 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 2 10 M 500 G 3 Input 4 +5V 5 6 0.1uF 7 8 26 1 33 k 9 78 10 LMC662 Output 11 3 4 1 uF 12 5 0.1uF 13 14 15 16 -5V 17 18 19 20 Fig. 2. Schematic diagram of the electrometer circuit. 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 HEPA Filter 2 3 Diffusion Dryer Soft X-ray Valve 4 TSI Model 3062 Aerosol Neutralizer Concentration HEPA Filter Valve TSI Model 3088 Adjustment Valves 5 Excess Air 6 7 Polydisperse 8 Aerosol Outlet Makeup Flow 9 Compressed 2.4 L/min 10 Air Flowmeter 11 Filtered Air Supply 12 TSI Model 3074B 13 NaCl/ HEPA Filter 14 DEHS 15 Constant Output Atomizer Valve 16 TSI Model 3076 0.6 L/min 17 18 19 Electrostatic Classifier 20 TSI Model 3082 21 22 1.5 L/min 1.5 L/min 23 24 25 26 27 Vacuum Pump Mass flow controller 28 29 Test Electrometer Condensation Particle Counter TSI Model 3788 30 31 32 Fig. 3. Experimental setup for evaluating the performance of the particulate detector. 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 times, sec 2 0 50 100 150 200 250 300 25 1.0 3 4 Warm-up time 5 Zero offset 6 0.5 7 20 8 electrometer voltage, mV 9 electrometer voltage, mV 10 15 11 12 0.0 13 14 10 15 16 -0.5 17 5 18 19 20 21 0 -1.0 0 10 20 30 40 50 60 70 80 90 22 23 times, min 24 25 26 Fig. 4. Warm-up time and zero offset of the developed particulate detector. 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 2 3 1.4x104 4 CPC 40 5 35 particle number concentration, particles/cm3 1.2x104 This Work 6 7 8 1.0x104 30 electrometer voltage, mV 9 10 25 11 8.0x103 12 20 13 6.0x103 14 15 10 15 4.0x103 16 17 18 2.0x103 5 19 20 0.0 0 21 0 200 400 600 800 1000 22 times, s 23 24 25 26 (a) DEHS 27 28 29 30 30 25 31 CPC 32 particle number concentration, particles/cm3 1.0x104 33 This Work 34 35 8.0x103 20 electrometer voltage, mV 36 37 38 6.0x103 15 39 40 41 4.0x103 10 42 43 44 2.0x103 5 45 46 47 0.0 0 48 0 200 400 600 800 1000 49 times, s 50 51 52 53 (b) NaCl 54 55 56 57 Fig. 5. Comparison between the particle number concentration measured by the UCPC and the electrometer voltage 58 59 measured by the developed particulate detector for 100 nm DEHS and NaCl particles. 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 2 3 4 1.0x104 5 DEHS 6 particle number concentration, particles/cm3 7 NaCl Linear Fit of NaCl 8 8.0x103 Linear Fit of DEHS 9 10 11 12 6.0x103 13 14 Equation y = a + b*x 15 Weight No Weighting 16 4.0x103 Residual Sum of 17 Squares 0.78459 18 Pearson's r Adj. R-Square 0.99964 NaCl 0.99922 Value Standard Error Intercept 0.0854 0.1187 Slope 0.00277 2.14091E-5 19 Equation y = a + b*x 20 2.0x103 Weight No Weighting 21 Residual Sum of Squares 1.65392 22 Pearson's r Adj. R-Square 0.99934 23 DEHS 0.99858 Value Standard Error Intercept 0.80322 0.15158 Slope 0.00307 3.09918E-5 24 0.0 25 0 5 10 15 20 25 30 35 26 electrometer voltage, mV 27 28 29 30 Fig. 6. Linear correlations between the number concentration of DEHS and NaCl particles and the electrometer 31 32 voltage. 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 2 3 4 500 particle number concentration/electrometer voltage 5 6 7 8 400 9 10 11 300 12 13 14 15 200 16 17 18 19 100 DEHS 20 21 22 NaCl 23 0 1000 24 100 25 26 particle diameter, nm 27 28 29 Fig. 7. Ratio of the particle number concentration and the electrometer voltage as a function of particle diameter for 30 31 32 DEHS and NaCl particles. 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 (a) 1.5 L/min 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 (b) 3.0 L/min 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 (c) 6.0 L/min 58 59 Fig. 8. Response time of the developed particulate detector at different operating aerosol flow rate. 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 2 3 8.0x103particle number concentration, particles/cm3 CPC 4 This work 5 6 7 8 6.0x103 9 10 11 12 4.0x103 13 14 15 16 17 2.0x103 18 19 20 21 0.0 22 0 100 200 300 400 23 times, s 24 25 26 27 (a) DEHS 28 29 30 31 8.0x103 32 CPC 33 particle number concentration, particles/cm3 This work 34 35 36 6.0x103 37 38 39 40 4.0x103 41 42 43 44 45 2.0x103 46 47 48 49 0.0 50 0 100 200 300 400 51 times, s 52 53 54 55 (b) NaCl 56 57 58 59 Fig. 9. Response time of the UCPC and the developed particulate detector. 60 61 62 63 64 65

A manuscript to the Journal of Electrical Engineering & Technology, Springer 1 2 3 100.0 4 5 6 7 8 99.5 9 counting efficiency, % 10 11 12 99.0 13 14 15 16 17 98.5 18 19 DEHS 20 NaCl 21 98.0 100 22 1000 23 particle diameter, nm 24 25 26 27 Fig. 10. Results of evaluating the particle counting efficiency for DEHS and NaCl particles. 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

ภาคผนวก ง มาตรฐานผลติ ภัณฑ์อตุ สาหกรรม เครอ่ื งตรวจวัดฝ่นุ ละอองลอยแบบหลกั การไฟฟ้าสถติ

หน้า ๓๕ ๒๙ พฤษภาคม ๒๕๖๓ เล่ม ๑๓๗ ตอนพเิ ศษ ๑๒๖ ง ราชกจิ จานุเบกษา ประกาศกระทรวงอตุ สาหกรรม ฉบับท่ี 5708 (พ.ศ. 2563) ออกตามความในพระราชบญั ญตั ิมาตรฐานผลติ ภัณฑ์อตุ สาหกรรม พ.ศ. 2511 เร่อื ง กาหนดมาตรฐานผลติ ภัณฑ์อตุ สาหกรรม เครือ่ งตรวจวัดฝนุ่ ละอองลอยหลักการไฟฟา้ สถติ อาศัยอานาจตามความในมาตรา 15 แห่งพระราชบัญญัติมาตรฐานผลิตภัณฑ์อุตสาหกรรม พ.ศ. 2511 ซึ่งแก้ไขเพิ่มเติมโดยพระราชบัญญัติมาตรฐานผลิตภัณฑ์อุตสาหกรรม (ฉบับที่ 7) พ.ศ. 2558 รัฐมนตรีว่าการกระทรวงอุตสาหกรรมออกประกาศกาหนดมาตรฐานผลิตภัณฑ์อุตสาหกรรม เครื่องตรวจวัดฝุ่นละอองลอยหลักการไฟฟ้าสถิต มาตรฐานเลขที่ มอก. 3030 - 2563 ไว้ ดังมรี ายละเอยี ดต่อท้ายประกาศนี้ ทั้งนี้ ใหม้ ีผลตั้งแตว่ ันท่ปี ระกาศในราชกิจจานเุ บกษาเป็นตน้ ไป ประกาศ ณ วนั ที่ 4 มนี าคม พ.ศ. ๒๕๖3 สรุ ิยะ จงึ รงุ่ เรืองกจิ รฐั มนตรวี ่าการกระทรวงอตุ สาหกรรม

ข้อมูลมาตรฐานผลิตภณั ฑ์อุตสาหกรรม แนบทา้ ยประกาศกระทรวงอตุ สาหกรรม ฉบบั ที่ 5708 (พ.ศ.2563) ช่อื มาตรฐาน : เครอ่ื งตรวจวดั ฝนุ่ ละอองลอยหลักการไฟฟ้าสถติ ELECTROSTATIC AEROSOL PARTICULATE MATTER (PM) MONITOR มาตรฐานเลขท่ี : มอก. 3030−2563 ผู้จดั ทํา : สํานักงานมาตรฐานผลติ ภณั ฑอ์ ตุ สาหกรรม กรรมการวชิ าการ : คณะกรรมการวิชาการรายสาขา คณะที่ 70 ผลติ ภณั ฑน์ วตั กรรม ขอบขา่ ย : มาตรฐานผลติ ภณั ฑ์อตุ สาหกรรมนี้ - ครอบคลุมเฉพาะเครอ่ื งตรวจวัดฝนุ่ ละอองลอยในอากาศด้วยหลักการไฟฟา้ สถิต ใชต้ รวจวัดฝุ่นละอองลอย PM10 PM2.5 PM1.0 และชว่ งการวัด (measuring range) มวลฝนุ่ ละอองลอยตอ้ งอยใู่ นชว่ ง 0.1 μg/m3 ถึง 5 000 μg/m3 เน้อื หาประกอบด้วย : ขอบข่าย บทนิยาม แบบ คณุ ลกั ษณะทต่ี อ้ งการ การบรรจุ เครอื่ งหมายและ ฉลาก การชกั ตวั อย่างและเกณฑต์ ัดสิน การทดสอบ ภาคผนวก ก. ภาคผนวก ข. และภาคผนวก ค. จาํ นวนหนา้ : 13 หนา้ ISBN : 978-616-475-526-0 ICS : 13.040 สถานทจ่ี ดั เกบ็ : หอ้ งสมดุ สาํ นกั งานมาตรฐานผลติ ภณั ฑอ์ ตุ สาหกรรม ถนนพระรามที่ 6 กรงุ เทพมหานคร 10400 สถานทจ่ี ําหนา่ ย : กองสง่ เสรมิ และพฒั นาดา้ นการมาตรฐาน สํานักงานมาตรฐานผลติ ภณั ฑ์อุตสาหกรรม ถนนพระรามท่ี 6 กรงุ เทพมหานคร 10400 โทรศพั ท์ 0 2202 3426

มาตรฐานผลติ ภณั ฑ์อุตสาหกรรม เคร่อื งตรวจวดั ฝนุ่ ละอองลอยหลักการไฟฟา้ สถติ 1. ขอบขา่ ย 1.1 มาตรฐานผลิตภณั ฑอ์ ุตสาหกรรมน้ีครอบคลุมเฉพาะเคร่ืองตรวจวัดฝุ่นละอองลอยในอากาศด้วยหลักการไฟฟ้า สถิต ใช้ตรวจวัดฝุ่นละอองลอย PM10, PM2.5, PM1.0 และช่วงการวัด (measuring range) มวลฝุ่นละออง ลอยต้องอยใู่ นชว่ ง 0.1 µg/m3 ถงึ 5000µg/m3 2. บทนิยาม ความหมายของคาท่ใี ชใ้ นมาตรฐานผลติ ภณั ฑ์อุตสาหกรรมน้ี มีดงั ต่อไปนี้ 2.1 เครื่องตรวจวัดฝุ่นละอองลอย [aerosol particulate matter (PM) monitor]ซ่ึงต่อไปในมาตรฐานนี้จะ เรียกว่า “เคร่ืองวัดฝุ่น” หมายถึงเคร่ืองตรวจวัดปริมาณของอนุภาคขนาดเล็กในอากาศหรือละอองลอย (aerosol)โดยอาศัยหลักการไฟฟ้าสถิต (electrostatic)ของอนุภาคผ่านแผ่นกรองอากาศ (filter)แล้วแปรค่า สัญญาณไฟฟ้าเป็นปรมิ าณหรอื มวลอนภุ าค (PM concentration) 3. แบบ 3.1 เครือ่ งวดั ฝ่นุ แบง่ ตามการวัดฝ่นุ เป็น2แบบ คือ 3.1.1 แบบช่องวัดเดยี่ ว (single channel type) คอื แบบทีผ่ ลิตสาหรบั วดั ฝ่นุ ได้ขนาดเดยี ว 3.1.2 แบบหลายช่องวดั (multi channel type) คือ ท่ผี ลิตสาหรบั วัดฝุ่นได้พรอ้ มกนั หลายขนาด 4.คุณลักษณะทต่ี อ้ งการ 4.1 ลักษณะทัว่ ไป เคร่ืองวัดฝุ่นต้องประกอบอย่างเรียบร้อยขอบและมุมของเคร่ืองต้องปราศจากส่วนแหลมคมท่ีอาจเป็นอันตราย ต่อผู้ใช้งาน สายไฟฟ้าต้องมีการป้องกันความเสียหายและต้องไม่มีตาหนิอื่นใดท่ีอาจส่งผลเสียหายต่อการใช้ งาน การทดสอบใหท้ าโดยการตรวจพนิ จิ 4.2 อุปกรณ์ควบคมุ ประกอบดว้ ย 4.2.1 เครื่องแสดงผล ต้องเป็นดงั น้ี (1) ตอ้ งมจี อแสดงผล ทอี่ ่านงา่ ยและเหน็ ชดั เจน (2) ชว่ งการวดั (measuring range) มวลฝุ่นต้องอยู่ในช่วง 0.1 µg/m3ถงึ 5000µg/m3 (3) มาตรเชิงตวั เลข ต้องมคี วามละเอียดถึง 0.1 µg/m3 (4) มาตรวัดอตั ราการไหล ตอ้ งมีความละเอยี ดถงึ 0.1 liter/min (5) แสดงระยะเวลาการรายงานผล เช่น ค่าเฉลย่ี ทุก ๆ 1 min 1 h 8 h หรือ 24 h -1-