Article pubs.acs.org/molecularpharmaceutics
Measuring Bipolar Charge and Mass Distributions of Powder Aerosols by a Novel Tool (BOLAR) Jennifer Wong,† Yu-Wei Lin,† Philip Chi Lip Kwok,‡ Ville Niemela,̈ § John Crapper,∥ and Hak-Kim Chan*,† †
Advanced Drug Delivery Group, Faculty of Pharmacy, The University of Sydney, Sydney, New South Wales 2006, Australia Department of Pharmacology and Pharmacy, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong SAR, China § Dekati Limited, Kangasala FI-36240, Finland ∥ Pharmaxis Limited, Frenchs Forest, New South Wales 2086, Australia ‡
ABSTRACT: The Bipolar Charge Analyzer (BOLAR) was evaluated for measuring bipolar electrostatic charge and mass distributions of powder aerosols generated from a dry powder inhaler. Mannitol powder (5, 10, and 20 mg) was dispersed using an Osmohaler inhaler into the BOLAR at air flow rates of 30 or 60 L/min. As the aerosol sample was drawn through the BOLAR, the air flow was divided into six equal fractions. Five of them entered individual detection tubes with a defined cutoff diameter in the range of 0.95 to 16.36 μm (depending on the flow rate) and the remaining (i.e., the sixth) fraction passed through a reference chamber. The aerosols that entered the detection tubes were separated according to the particle charge polarity (positive, negative, or neutral) and charge was measured by separate electrometers. The deposited powder of a single actuation from the inhaler was chemically assayed using high performance liquid chromatography. Additionally, the aerosol measurements were conducted on a modified Classic Electrical Low Pressure Impactor (ELPI) for comparison of the net specific charge per size fraction. Spray-dried mannitol carried significantly different positively and negatively charged particles in each of the five defined particle size fractions. The charge-to-mass ratio (q/m) of positively charged particles ranged from +1.11 to +32.57 pC/μg and negatively charged particles ranged from −1.39 to −9.25 pC/μg, resulting in a net q/m of −3.08 to +13.34 pC/μg. The net q/m values obtained on the modified ELPI ranged from −5.18 to +4.81 pC/μg, which were comparable to the BOLAR measurements. This is the first full report to utilize the BOLAR to measure bipolar charge and mass distributions of a powder aerosol. Positively and negatively charged particles were observed within each size fraction, and their corresponding q/m profiles were successfully characterized. Despite some potential drawbacks, the BOLAR has provided a new platform for investigating bipolar charge in powder aerosols for inhalation. KEYWORDS: Osmohaler, spray-dried mannitol powder, triboelectrification, Electrical Low Pressure Impactor (ELPI), Bipolar Charge Analyzer (BOLAR)
■
INTRODUCTION Pharmaceutical aerosols delivered by dry powder inhalers (DPIs) are likely to generate and accumulate bipolar charges when the solid particles undergo triboelectrification with each other and the surfaces of the inhaler.1−3 Charge transfer occurs between particles of various sizes and surfaces made from different materials due to a difference in work function. Work function is defined as the minimum amount of energy required to remove an electron from the surface of a solid to a distance infinitely far. The theory of work © 2015 American Chemical Society
function dictates the direction of electron transfer as moving from the material with lower work function to the one with a higher work function until the electron energies reach equilibrium.4,5 This also contributes to the magnitude and polarity of charge acquired. Received: Revised: Accepted: Published: 3433
June 5, 2015 July 20, 2015 July 29, 2015 July 29, 2015 DOI: 10.1021/acs.molpharmaceut.5b00443 Mol. Pharmaceutics 2015, 12, 3433−3440
Article
Molecular Pharmaceutics
It is the first commercially available impactor capable of separating and measuring positively and negatively charged particles according to particle size fractions by mass at air flow rates of 30 to 90 L/min. The concept, instrument design, and calibration of the BOLAR have been discussed in detail by Yli-Ojanperä et al.20 In brief, the instrument consists of two main components: a flow divider and five electrical detection tubes (Figure 1). The flow divider evenly separates air flow into six
Charged aerosols are of interest because electrostatic charge is one of five mechanisms that governs the deposition of inhaled particles in the lungs.6 In particular, “image charge” is the attractive force between a charged particle and the grounded lining of the respiratory tract. While theoretical and experimental studies have demonstrated the potential influence of electrostatic charge on lung deposition, the relationship between the magnitude and polarity of charged aerosols on the total and regional lung deposition in human subjects is still unclear.1 An important step toward establishing this relationship requires the accurate measurement of pharmaceutical aerosol charges. There is a large body of evidence in the literature, which confirms that a number of pharmaceuticals carry substantial levels of charge.1,7,8 More importantly, pharmaceutical powders are known to carry bipolar charges. Balachandran et al.9 investigated two pharmaceutical powders that contained either a carrier only or a carrier mixed with an active drug. The carrier only powder had particles that carried +2.5 × 10−9 C of positive charge and −2.8 × 10−9 C of negative charge, whereas the carrier mixed with an active drug powder had particles with +3.1 × 10−9 C positive charge and −2.6 × 10−9 C negative charge. Later, Beleca et al.10,11 measured the electrical properties of lactose. Lactose monohydrate particles with volume median diameter (VMD) of 30 μm exhibited positive charges of +3.2 × 10−15 C and negative charges of −4.0 × 10−15 C. Similarly, particles that had a VMD of 70 μm carried positive and negative charges of +2.8 × 10−15 C and −2.0 × 10−15 C, respectively.11 It is important to note that large carrier particles are not intended to be inhaled into the lungs and generally deposit in the throat. The bipolar charge of pharmaceutical aerosols generated by DPIs and metered-dose inhalers (MDIs) have also been measured.12,13 Individual particles with aerodynamic diameters between 3 to 5 μm had high levels of positive and negative charge in the range ±2.5 μC/g, even though the net charge-to-mass ratio was less than +0.2 μC/g.12 Despite the existence of bipolar charge, a majority of studies in the literature only characterized net charge. Static methods that measure the net charge and mass distribution of inhalable products have included the modified Electrical Low Pressure Impactor (ELPI),14−17 electrical Next Generation Impactor (eNGI),18 and Twin Stage Impinger.19 Each of the impactor stages operate as an individual Faraday pail, and the net charge is measured by separate electrometers. Although the net charge provides important information, it suffers from the limitation that bipolarity of particles within each size fraction cannot be discerned. Hence, methods able to determine bipolar charges would give more valuable details and are essential for understanding the role of electrostatic charge in pulmonary drug delivery. Dynamic methods assess bipolar charge by measuring the electrical mobility of individual particles. Instruments that operate on this principle have included the Bipolar Charge Measurement System (BCMS),9 modified Phase Doppler Anemometry (PDA),10,11 and the Electrical Single-Particle Aerodynamic Relaxation Time (E-SPART) analyzer.12,13 These instruments have been used to quantify bipolar charges of several DPIs. However, the experimental setups are generally complex, and particle charge is based on particle count, not mass. Charge distributions by mass are more direct and relevant for pharmaceutical aerosols because the dose (i.e., the mass of the active ingredient) determines the therapeutic outcome. Thus, a simpler instrument that can simultaneously measure bipolar charge and particle distribution by mass under realistic inhalational flow rates is needed. The Bipolar Charge Analyzer (BOLAR, Dekati Ltd., Kangasala, Finland) was designed to solve these problems.
Figure 1. Schematic diagram of the BOLAR. Diagram not drawn to scale.
branches, five of which lead to detection tubes and the sixth to a simple Faraday pail (reference chamber). Each detection tube is composed of an impactor with a specific cutoff diameter, an inner cylinder at positive potential (∼1000 V), a grounded outer cylinder, and a filter stage at the bottom. When a particle travels in the gap between the two cylinders, positively charged particles deposit on the outer cylinder, negatively charged particles deposit on the inner cylinder, and neutral particles pass through the gap and deposit on the filter stage. Both cylinders are connected to individual electrometers that measure the positive and negative charges separately. When the measurement is complete, the detection tubes can be dis-assembled for mass assay to determine the mass distribution and, in turn, the specific charges (charge-to-mass ratios). The primary objective of the present study was to evaluate the capability of the BOLAR to measure bipolar electrostatic charge and mass distributions of a DPI. The commercially available Osmohaler inhaler is used to disperse inhalable mannitol powder for the diagnosis of asthma and the treatment of cystic fibrosis. This DPI was selected for the study because it is simple and contains only pure drug. The capsule-based unit-dose DPI device system and relatively large unit dose (5−20 mg) enabled chemical analysis from a single actuation, thereby eliminating error due to interdispersion variations from multiple actuations. Additionally, two flow rates were examined to represent variable patient inspiratory flow rates. Comparisons of specific net charge data measured by both the BOLAR and modified ELPI were also made to confirm the reliability of the results obtained using the BOLAR.
■
MATERIALS AND METHODS Materials. Spray-dried mannitol powder and the commercial Osmohaler inhaler were supplied by Pharmaxis Ltd., Frenchs 3434
DOI: 10.1021/acs.molpharmaceut.5b00443 Mol. Pharmaceutics 2015, 12, 3433−3440
Article
Molecular Pharmaceutics
the DPI and charge measurement of the aerosol, and 20 s of baseline after the valves were closed again. All charges from a single dose were measured until the charge signals returned to baseline, which occurred around the first 10 s following the opening of the valves. Blank measurements were conducted in a similar way, except an Osmohaler inhaler loaded with a pierced empty capsule was held tightly at the throat. These blank readings were subtracted from the powder charge measurements to give the actual charge and account for disturbances in the electrical signals due to the presence of the inhaler at the throat. The modified ELPI measured and recorded charge using the Dekati ELPI VI 4.0 software with the following settings: current range at 400 000 fA, saving interval 1 s, data correction unselected, trap off, pressure drop adjusted to 100 mbar, and the electrometers were zeroed without flushing. Since the ELPI does not have internal valves to control the flow rate, the measurement sequence was different to the BOLAR. First, the vacuum pump was turned on. Second, when a stable zero baseline was obtained the loaded inhaler was inserted into the throat until the dose was completely emptied and the charge signals returned to baseline. Third, the inhaler was reinserted and removed from the throat three times to generate blank signals. Finally, the mean of these blank readings was subtracted from the charge measurements to give the actual charge and account for disturbances in the electrical signals due to the insertion of the inhaler into the throat. Mass Assay. Powder that deposited on the device, capsule, adaptor, USP throat, unit dose collector, and each part of the BOLAR or ELPI components was exhaustively washed with 2.5−25 mL of deionized water. The specific volumes for the BOLAR were 2.5 mL for the capsule; 5 mL each for the inhaler, adaptor, USP throat, flow divider inlet, impactor stages, and glass fiber filter; 10 mL for the filter stage; and 25 mL each for the flow divider, inner detection tube, and outer detection tube. For the ELPI, the volumes were 5 mL each for the capsule, inhaler, adaptor, USP throat, Y-piece, straight tube, impactor stages, and glass fiber filter; and 25 mL for the unit dose collector. The glass fiber filters were washed with deionized water and centrifuged at 13 400 rpm for 10 min (Minispin, Eppendorf, Westbury, USA), and the supernatant was collected for assay. Aliquots of the samples were chemically assayed using high performance liquid chromatography (HPLC). The Shimadzhu HPLC system used included an RID-10A detector, LC-20AT pump, and SIL-20A HT autosampler controlled by LCSolution software (all Shimadzhu Scientific Instruments, Kyoto, Japan). The mobile phase consisted of deionized water filtered through a 0.45 μm polyamide filter membrane (Sartorius Stedim Biotech GmbH, Göttingen, Germany). Samples were injected (20 μL) into a Resolve C18 5 μm 3.9 × 150 mm column (Waters, Milford, USA) at a flow rate of 1 mL/min, resulting in a retention time of ∼2 min. A calibration curve was constructed by serial dilution
Forest, Australia; deionized water (electrical resistivity >2 MΩ cm at 25 °C) was obtained from a Modulab Type II Deionization System (Continental Water Systems, Sydney, Australia); Slipicone silicone release spray was from DC products, Mount Waverly, Australia; and 45 mm glass fiber filters were from MicroAnalytix, Pty. Ltd., Sydney, Australia. Particle Size. The particle size distribution was measured by laser diffraction using a Mastersizer 2000 coupled to a Scirocco 2000 dry dispersion unit (both Malvern Instruments, Worcestershire, UK). Triplicate samples of mannitol powder, approximately 20 mg, were dispersed by compressed air at 4 bar of pressure. Measurement parameters were obscuration of 0.5−6%; particle refractive index of 1.52; particle absorption of 0.10; and the dispersant (air) refractive index of 1.00. The diameters corresponding to the cumulative volume distribution under 10, 50, and 90% were expressed as the d10, d50, and d90, respectively. The span, which quantified the broadness of the size distribution, was defined as (d90 − d10)/d50. Scanning Electron Microscopy. Particle morphology was observed using scanning electron microscopy (SEM). Samples were dispersed onto sticky carbon tape mounted on SEM stubs and coated with an approximately 15 nm thick layer of gold (K550X sputter coater, Quorom Emitech, Kent, UK). Images were obtained at 2 kV using a field emission SEM (Zeiss Ultra Plus, Carl Zeiss SMT Ag, Oberkochen, Germany). Charge Measurements. Inherent charges of aerosolized mannitol powder were characterized using the BOLAR and modified Classic ELPI (both Dekati Ltd., Kangasala, Finland). The ELPI normally operates at 30 L/min, but aerosolization can be achieved at 60 L/min through a Y-piece connected to a unit dose collector with an additional pump, as described previously.14,15,21 Consequently, air flow rates of 30 and 60 L/min were selected based on the achievable flow rates of the modified ELPI. To minimize particle bounce and re-entrainment, impactor stages in both instruments were sprayed with silicone, and the propellant was allowed to evaporate before the unit was reassembled. Additionally, the BOLAR detection tubes were not sprayed with silicone, and glass fiber filters were placed in the bottommost filter stage. The air flow rate was calibrated using a flow meter (TSI 3063, TSI Instruments Ltd., Shoreview, USA), which was fitted with rubber adapters for airtight sealing and attached to an Osmohaler inhaler loaded with a pierced empty capsule. Spray-dried mannitol powder (5, 10, or 20 mg) was filled into a size 3 hydroxypropyl methylcellulose capsule (Vcaps, Capsugel Australia Pty. Ltd., West Ryde, Australia), loaded into the inhaler, and pierced prior to experiments. A single actuation was discharged for each measurement and experiments were performed in triplicate without gloves to simulate patient use. Ambient environmental conditions were also monitored (38 ± 6% RH, 22 ± 1 °C). The BOLAR was operated according to the manufacturer’s instructions. First, a self-check test was performed before each measurement to ensure the BOLAR’s electrical control system was working properly. Subsequently, the loaded inhaler was attached to the USP throat, and the vacuum pump was turned on. When the automated measurement sequence was initiated, the internal valves opened air flow through the inhaler to disperse the powder into the BOLAR, and the charge was measured. One measurement sequence took approximately 60 s, and the inhaler was held tightly at the throat for the entire duration. This 60 s duration consisted of 10 s of baseline prior to opening the valves, 30 s of opening the valves that allowed air flow to disperse
Table 1. Aerodynamic Diameter of the Particles Collected in the BOLAR Bipolar Charge Detection Tubes at 60 L/min22
3435
tube
aerodynamic diameter (μm)
1 2 3 4 5
0.1 < dp < 0.95 0.1 < dp < 2.60 0.1 < dp < 4.17 0.1 < dp < 7.29 0.1 < dp < 11.57 DOI: 10.1021/acs.molpharmaceut.5b00443 Mol. Pharmaceutics 2015, 12, 3433−3440
Article
Molecular Pharmaceutics using freshly prepared standard solutions in deionized water at concentrations of 0.1 to 1000 μg/mL (R2 = 0.999). Data Analysis. The BOLAR consists of an unconventional parallel detection tube assembly (Table 1) where the charge and
measured mass of each detection tube was calculated by a subtraction between two sequential tubes (Table 2). Since only one-sixth of the total dose is sampled by each detection tube, the magnitude of charge and mass equals to the measured value multiplied by six. Moreover, the midpoint of the particle size range instead of the cutoff diameter of each detection tube was used for graphical presentation. Further information on the instrument design and calculations for BOLAR have been described elsewhere.20 The BOLAR detects charge as current where the total charge of the positive or negative particles was obtained by integrating the electrical current signals on the outer and inner detection tubes, respectively. The net charge in a particular detection tube was the sum of the positive and negative total charges. The charge-to-mass ratio (q/m) was the quotient of the total charge and drug mass on a particular detection tube, while the net q/m was the quotient of the net charge and total drug mass in a particular detection tube. Similarly, the ELPI also measures the electrical current on each impactor stage. The net charge was calculated by integration of the current-versus-time plot, and the q/m was the quotient of the net charge and drug mass on a particular impactor stage. At air flow rates of 60 L/min, only half of the total dose is sampled by the modified ELPI. Consequently, the magnitude of charge and mass were corrected by multiplying by two. Statistical Analysis. Statistical analysis was conducted using one-way analysis of variance (ANOVA), followed by the Tukey multiple comparisons post hoc test to determine differences between the means (GraphPad Prism 6, GraphPad Software Inc., La Jolla, USA). A p < 0.05 was considered statistically significant.
Table 2. Charge and Mass Calculations for BOLAR at 60 L/min22 Aerodynamic diameter range (μm)
midpoint diameter (μm)
charge, q (pC)
mass, m (μg)
0.1−0.95 0.95−2.60 2.60−4.17 4.17−7.29 7.29−11.57
0.525 1.78 3.39 5.73 9.43
6qtube 1 6(qtube 2 − qtube 1) 6(qtube 3 − qtube 2) 6(qtube 4 − qtube 3) 6(qtube 5 − qtube 4)
6mtube 1 6(mtube 2 − mtube 1) 6(mtube 3 − mtube 2) 6(mtube 4 − mtube 3) 6(mtube 5 − mtube 4)
Figure 2. Scanning electron micrograph of spray-dried mannitol powder (magnification at 5000×).
Figure 3. Mean electrostatic charge, mass deposition, and charge-to-mass profiles at 30 L/min using BOLAR. Data presented as mean ± one standard deviation (n = 3). Components with
