Differences in estimates of size distribution of beryllium powder materials using phase contrast microscopy, scanning electron microscopy, and liquid suspension counter techniques
© Stefaniak et al; licensee BioMed Central Ltd. 2007
Received: 27 October 2006
Accepted: 28 February 2007
Published: 28 February 2007
Accurate characterization of the physicochemical properties of aerosols generated for inhalation toxicology studies is essential for obtaining meaningful results. Great emphasis must also be placed on characterizing particle properties of materials as administered in inhalation studies. Thus, research is needed to identify a suite of techniques capable of characterizing the multiple particle properties (i.e., size, mass, surface area, number) of a material that may influence toxicity. The purpose of this study was to characterize the morphology and investigate the size distribution of a model toxicant, beryllium. Beryllium metal, oxides, and alloy particles were aerodynamically size-separated using an aerosol cyclone, imaged dry using scanning electron microscopy (SEM), then characterized using phase contrast microscopy (PCM), a liquid suspension particle counter (LPC), and computer-controlled SEM (CCSEM). Beryllium metal powder was compact with smaller sub-micrometer size particles attached to the surface of larger particles, whereas the beryllium oxides and alloy particles were clusters of primary particles. As expected, the geometric mean (GM) diameter of metal powder determined using PCM decreased with aerodynamic size, but when suspended in liquid for LPC or CCSEM analysis, the GM diameter decreased by a factor of two (p < 0.001). This observation suggested that the smaller submicrometer size particles attached to the surface of larger particles and/or particle agglomerates detach in liquid, thereby shifting the particle size distribution downward. The GM diameters of the oxide materials were similar regardless of sizing technique, but observed differences were generally significant (p < 0.001). For oxides, aerodynamic cluster size will dictate deposition in the lung, but primary particle size may influence biological activity. The GM diameter of alloy particles determined using PCM became smaller with decreasing aerodynamic size fraction; however, when suspended in liquid for CCSEM and LPC analyses, GM particle size decreased by a factor of two (p < 0.001) suggesting that alloy particles detach in liquid. Detachment of particles in liquid could have significance for the expected versus actual size (and number) distribution of aerosol delivered to an exposure subject. Thus, a suite of complimentary analytical techniques may be necessary for estimating size distribution. Consideration should be given to thoroughly understanding the influence of any liquid vehicle which may alter the expected aerosol size distribution.
Relating the physicochemical properties of an aerosol to its toxicity following inhalation is of specific interest for beryllium and of general interest for a wide range of particles, including nanomaterials. Various hypotheses have been put forth regarding particle mass, surface area, and number as metrics of toxicity of inhaled particles. Detailed control and physicochemical characterization of aerosols generated for inhalation toxicology studies is essential for obtaining meaningful study results.
A variety of techniques exist for exposing the lung to particles in pulmonary toxicology studies, including inhalation, intratracheal instillation, and pharyngeal aspiration. A high research priority is the characterization of the exposure material as encountered by workers and as administered in a pulmonary toxicology study. Characterization of bulk material as produced or supplied may not accurately reflect the properties of the particles in the workplace atmosphere or of particles delivered to the lung by inhalation, instillation, or pharyngeal aspiration. For example, current methods for characterizing dry material using microscopy can reveal general morphology and particle size, but cannot determine whether material consists of solid particles or clusters, or whether material will deagglomerate in liquids such as lung surfactant. Similarly, current methods for characterizing material in suspension can reveal "equivalent" particle diameter, but cannot determine particle morphology. Thus, a suite of techniques is needed to understand the multiple particle properties that may influence toxicity.
The purpose of this study was to apply a suite of three different standard particle sizing techniques to investigate the size distributions of three physicochemical forms of beryllium and to assess the meaning and implication of any differences in the measured size distributions.
We studied finished product metal powder (product type I-400, Brush Wellman Inc., Elmore, OH); finished product beryllium oxide (BeO) powder (product type UOX-125, Brush Wellman Inc.); process-sampled particles from the BeO powder production line; and process-sampled aerosol from an arc furnace in the copper-beryllium alloy production line. These materials were selected for study because they represent three industrially important chemical forms of beryllium and because inhalation of these materials is associated with development of chronic beryllium disease. The process-sampled BeO material is aerosol particles generated during screening of finished product BeO powder.
Bulk metal and BeO powders were aerosolized and aerodynamically size-separated using a 5-stage aerosol cyclone operated at 24 L min-1 and 20°C; the 50% aerodynamic cutoff diameters (D50) were >6, 2.5, 1.7, 0.9, and 0.4 μm for stages 1 to 5 of the aerosol cyclone, respectively. Process-sampled particles were collected from ventilation ductwork using a 5-stage aerosol cyclone operated at 28 L min-1 and 23°C; D50 were >5.7, 2.3, 1.5, 0.7, 0.4 μm for stages 1 to 5 of the aerosol cyclone, respectively. For the purposes of this study, we characterized material collected in stages 2, 3, and 4 of the aerosol cyclone. The density of the metal powders determined using pycnometry by Hoover et al.  were 1.89 g/cm3 (stage 2), 1.89 g/cm3 (stage 3), and 1.91 g/cm3 (stage 4); the densities of the oxide powder and process-sampled BeO materials were not measured directly, but based on material purity , were taken to be the theoretical value for BeO of 3.0 g/cm3. The masses of processed sampled alloy particles collected were insufficient for direct measurement of density by pycnometry.
Detailed morphology of the dry powders and particles was determined using scanning electron microscopy (SEM, Model 6300FXV, JEOL USA, Peabody, MA) operated at 5 kV. The SEM samples were prepared on 300 mesh gold electron microscopy grids coated with a lacey carbon substrate (Ted Pella Inc., Redding, CA).
The geometric mean (GM) particle size and geometric standard deviation (GSD) was determined using phase contrast microscopy (PCM), a liquid suspension particle counter (LPC), and computer-controlled scanning electron microscopy (CCSEM).
PCM (Carl Zeiss Inc., Germany) was used to determine particle Feret diameter (distance between tangents drawn from the extreme left and right edges of a particle measured perpendicular to a reference line) using a Filar ocular micrometer; the theoretical limit of resolution was 0.4 μm. Prior to analysis, the ocular micrometer was calibrated using a stage micrometer (10 μm scale division). For each material, 1000 particles were sized from slides of particles dispersed in mounting medium on a slide.
A calibrated LPC (Multisizer II, Coulter Electronics, Inc., Hialeah, FL) operated with 30 μm diameter aperture tube (0.6 – 18 μm actual cutoff range) was used to determine particle spherical equivalent diameter (spherical size equivalent to the volume of liquid displaced by a particle that produces the same voltage pulse when drawn between two electrodes). Prior to analysis, the LPC was calibrated using monodisperse 5-μm polystyrene latex spherical reference particles (Beckman-Coulter) and the calibration verified using monodisperse 2-μm polystyrene latex spherical reference particles (Beckman-Coulter). For analysis, a dilute suspension of each material in physiologic saline was subjected to ultrasonic agitation for 30 sec to mix, then an aliquot of each suspension was added to electrolyte solution (Coulter Electronics, Inc.); tens of thousands of particles were sized for each material. Size data were background corrected for particle counts in fluids. Additionally, an aliquot of suspension was deposited onto a 300 mesh gold electron microscopy grid (Ted Pella Inc.) and particles imaged using SEM (JEOL).
CCSEM (R.J. Lee Group, Monroeville, PA) analysis is expense relative to PCM and LPC techniques; therefore, we used CCSEM for analysis of material collected in stage 4 of the aerosol cyclone only to confirm that these small particles were not missed during sizing with PCM . Using CCSEM, particle projected area diameter was determined with the rotated chord technique (length of 16 chords traced across the particle at equiangular intervals through the centroid of the particle) and ZepAPA computer software program algorithm (R.J. Lee Group). For analysis, a dilute suspension of each beryllium material in distilled and deionized water was deposited onto a polycarbonate filter using vacuum filtration and coated with a thin layer of gold/palladium; 1000 particles were sized from a representative section of each filter. Images of particles were obtained by depositing an aliquot of suspension onto a 300-mesh gold electron microscopy grid (Ted Pella Inc.) and analyzed using SEM (JEOL).
Results and discussion
Geometric mean diameter of beryllium powders and process-sampled particles
Geometric mean, μm (GSD)A
Be metal powder
The size of each BeO material appeared similar among sizing techniques (see Figures 1, images B1 to B3); however, values determined using PCM and the LPC were statistically different (p < 0.001), with the exception of stage 4 BeO powder (p = 0.49). For BeO, these small differences in GM particle size may be attributed to differences in particle size definition. For these materials, aerodynamic cluster size will dictate deposition in the lung, but primary particle size may influence biological activity.
The GM Feret diameter of particles sampled from the alloy production line became slightly smaller with decreasing aerodynamic size fraction; however, when suspended in liquid for CCSEM and LPC analyses, GM particle size decreased by up to a factor of two (p < 0.001). Thus, these beryllium-containing alloy particles probably also detach in liquid, thereby shifting the size distribution smaller.
The detachment of particles suspended in a liquid vehicle has significance for the expected versus actual size (and concomitant number) distribution of particles delivered to an exposure subject. In the current study, beryllium metal and beryllium alloy particles, but not beryllium oxide, detached in deionized water (in the absence of external agitation) and in physiological saline mixed with an electrolyte solution (after brief ultrasonic agitation). Maier et al.  report that in the absence of ultrasonic agitation, nanostructured TiO2 powder suspended in artificial lung fluid did not disaggregate, but when subjected to brief ultrasonic agitation an ultrafine particle size fraction formed. Sager et al.  report TiO2 nanoparticles and ultrafine carbon black exhibit varying degrees of dispersability in liquids such as physiologic saline, artificial lung fluid, and rat and mouse bronchoalveolar lavage (BAL) fluid. Thus, on a material-by-material basis, detailed knowledge of the effect of the sample preparation conditions, including the use of a liquid vehicle (water, physiological saline, artificial lung surfactant, BAL fluid, etc.), on measured physicochemical characteristics of aerosols generated for toxicology studies is essential.
We conclude that careful attention is needed when characterizing exposure aerosols for pulmonary toxicology studies. A suite of complimentary analytical techniques may be necessary for understanding particle size distribution. When estimating size distribution, consideration must be given to the use of a liquid vehicle, which may alter the expected aerosol size distribution.
The authors thank A.F. Hubbs at NIOSH for critical review of this manuscript and W.E. Miller for statistical support. This project was supported by NIOSH Research Grant 1R03 OH007447-01. ABS received support from NIEHS Training Program in Environmental Health Sciences ES07141. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the National Institute for Occupational Safety and Health.
- Stefaniak AB, Hoover MD, Day GA, Dickerson RM, Peterson EJ, Kent MS, Schuler CR, Breysse PN, Scripsick RC: Characterization of physicochemical properties of beryllium aerosols associated with prevalence of chronic beryllium disease. J Environ Mon 2004, 6: 523–532. 10.1039/b316256gView ArticleGoogle Scholar
- Oberdörster G, Maynard A, Donaldson K, Castranova V, Fitzpatrick J, Ausman K, Carter J, Karn B, Kreyling W, Lai D, Olin S, Monteiro-Riviere N, Warheit D, Yang H: Principles for characterizing the potential human health effects from exposure to nanomaterials: elements of a screening strategy. Part Fibre Toxicol 2005, 2: 8. 10.1186/1743-8977-2-8PubMed CentralPubMedView ArticleGoogle Scholar
- Rao GVS, Tinkle S, Weissman DN, Antonini JM, Kashon ML, Salmen R, Battelli LA, Willard PA, Hoover MD, Hubbs AF: Efficacy of a technique for exposing the mouse lung to particles aspirated from the pharynx. J Toxicol Environ Health A 2003, 66: 1441–1452. 10.1080/15287390306417PubMedView ArticleGoogle Scholar
- Hoover MD, Castorina BT, Finch GL, Rothenberg SJ: Determination of the oxide layer thickness on beryllium metal particles. Am Ind Hyg Assoc J 1989, 50: 550–553.PubMedView ArticleGoogle Scholar
- Stein F: Particle size measurement with phase contrast microscopy. Powder Tech 1969, 2: 327–334. 10.1016/0032-5910(69)80025-2View ArticleGoogle Scholar
- Maier M, Hannebauer B, Holldorff H, Albers P: Does lung surfactant promote dissaggregation of nanostructured titanium dioxide? J Occup Environ Med 2006, 48: 1314–1320. 10.1097/01.jom.0000215405.72714.b2PubMedView ArticleGoogle Scholar
- Sager TM, Porter DW, Robinson VA, Lindsley WG, Schwegler-Berry DE, Castranova V: Improved method to disperse nanoparticles for in vitro and in vivo investigation of toxicity. Nanotoxicology 2007, in press.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.