OSA's Digital Library

Virtual Journal for Biomedical Optics

Virtual Journal for Biomedical Optics

| EXPLORING THE INTERFACE OF LIGHT AND BIOMEDICINE

  • Editors: Andrew Dunn and Anthony Durkin
  • Vol. 6, Iss. 4 — May. 4, 2011
« Show journal navigation

Spectral characterization of biological aerosol particles using two-wavelength excited laser-induced fluorescence and elastic scattering measurements

Vasanthi Sivaprakasam, Horn-Bond Lin, Alan L. Huston, and Jay D. Eversole  »View Author Affiliations


Optics Express, Vol. 19, Issue 7, pp. 6191-6208 (2011)
http://dx.doi.org/10.1364/OE.19.006191


View Full Text Article

Acrobat PDF (1749 KB)





Browse Journals / Lookup Meetings

Browse by Journal and Year


   


Lookup Conference Papers

Close Browse Journals / Lookup Meetings

Article Tools

Share
Citations

Abstract

A two-wavelength laser-induced fluorescence (LIF) instrument has been developed and used to characterize individual biological aerosol particles, including biological warfare (BW) agent surrogates. Fluorescence in discrete spectral bands from widely different species, and also from similar species under different growth conditions were measured and compared. The two-wavelength excitation approach was found to increase discrimination among several biological materials, and especially with respect to diesel exhaust particles, a common interferent for LIF BW detection systems. The spectral characteristics of a variety of biological materials and ambient air components have been studied as a function of aerosol particle size and incident fluence.

© 2011 OSA

1. Introduction

2. Experimental design

The 2-SPFA optical design is shown schematically in Fig. 1
Fig. 1 Schematic of the two-wavelength excitation single-particle fluorescence analyzer, 2-SPFA showing the optical layout and an electronic timing diagram.
. An optical system previously described [8

8. V. Sivaprakasam, A. Huston, C. Scotto, and J. Eversole, “Multiple UV wavelength excitation and fluorescence of bioaerosols,” Opt. Express 12(19), 4457–4466 (2004). [CrossRef] [PubMed]

], was modified to accommodate additional spectral channels and incorporate custom data acquisition circuitry [13

13. Vtech Engineering Corporation, http://www.vtechcorp.com.

]. Aerosol particles are entrained in inlet air flow through an aerodynamic nozzle to form a collimated flow of about 0.8 mm in diameter by using a concentric column of sheath air flow. The inlet nozzle directs the aerosol-laden air flow at a right angle to the optical axis of the elliptical mirror (into the page), and positions the flow through the mirror’s primary focal point inside a sealed aerosol sample chamber. (Fig. 3
Fig. 3 The fluorescence signature in two 266 nm excited spectral bands from background aerosol particles sampled over a 2 hr period, overlaid with the signature of some of the common simulants and interferents. The background particles show substantial overlap with inorganic interferent kaolin and fungal spores.
in Sivaprakasam et al. [12

12. V. Sivaprakasam, T. Pletcher, J. E. Tucker, A. L. Huston, J. McGinn, D. Keller, and J. D. Eversole, “Classification and selective collection of individual aerosol particles using laser-induced fluorescence,” Appl. Opt. 48(4), B126–B136 (2009). [CrossRef] [PubMed]

] shows a 3-dimensional illustration of the particle flow in the aerosol chamber). A fused silica window is used to define the airtight aerosol chamber and pass the collected light from the elliptical mirror to the rest of the instrument. Interrogating laser beams enter and exit the sample chamber orthogonal to both the aerosol sample flow direction and the reflector optical axis. A small pump provides the differential pressure (nominally 5 kPa) to pull both HEPA filtered sheath air as well as the sample air into the aerosol chamber, and out through an exit tube mounted in the bottom chamber wall. The flow velocity of the particles in the sample stream is approximately 4 m/s, corresponding to a gas flow rate of about 0.3 liter/min. Light from a CW 785 nm diode laser is focused to intercept aerosol particles at the focal position of the elliptical mirror. Elastically scattered light from individual particles passing through this beam is used both to provide relative measure of particle size, and to generate a cueing signal for the pulsed excitation lasers when the scattering intensity exceeds a predetermined threshold value, T (see Timing Chart, inset, Fig. 1). Two externally Q-switched diode pumped Nd:YAG lasers (Coherent Model 501QDII and Coherent Vector 532-1000-20) are used to generate the UV excitation pulses. The first laser is frequency-quadrupled to generate a 266 nm pulse having a 20 ns pulse width, while the second laser is frequency-tripled to generate a 5 ns duration, 355 nm pulse. Fused silica beam splitters are used to pick off a fraction of each pulse to monitor the energy using a pair of photodiodes (PD).

Beams from both pulsed UV lasers and the CW 785 nm lasers are directed collinearly into the aerosol chamber. The output of the first pulsed UV laser occurs 1.2 μs after the cueing signal and the second laser pulse is generated approximately 0.4 μs after the first (see Timing Chart, inset, Fig. 1). The beams are focused using cylindrical lenses to form focal spots roughly 200 μm thick in the direction of the particle flow, and nearly 1mm wide, in order to intercept 100% of particle flow near the mirror focal point. Nearly 2π steradian of light that is either scattered or emitted from individual particles is collected by the elliptical mirror, and directed to a pair of field lenses to collimate the collected light. The collimated light passes through a series of custom dichroic beam splitters that separate the fluorescence into three broad spectral bands. Photomultiplier tubes (PMT’s) are used to detect the fluorescence signals in each of the three bands, as well as the scattered light. The first beam splitter was custom-designed to reflect light in a 70 nm wide band centered at 350 nm. This reflected light is directed to a 266 nm high reflectance (HR) mirror, which is used to reject 266 nm light from the 350 nm UV emission channel PMT. Subsequently, a Hoya U360 color-glass filter is used to further reduce any scattered 266 nm light (>107 rejection of 266 nm), and this channel also uses a gated PMT Hamamatsu H7680-01 to measure the 350 nm fluorescence signal. The gate is on only during the presence of the 266 nm pulse. Gating the PMT prevents damage due to high intensity scattered light from the subsequent 355 nm laser pulse. Reflected light from the 266 nm HR mirror is used to measure the elastic scatter signal from both the 266 nm and 355 nm lasers. The second dichroic beam splitter in the main optical train reflects light passed by the first one in an 80 nm wide band centered at 450 nm. A pair of high pass and low pass interference filters are used to isolate the light in the 450 nm band. The third and final dichroic beam splitter reflects light in a narrow band centered at 785 nm into a Hamamatsu H6780-20 PMT and monitors elastic scattering from the CW laser beam. Fluorescent emission in a 100 nm wide band centered around 550 nm passes through all of the beam splitters to a final PMT. A condensing lens is also used in each of the channels to focus the collimated light onto the PMT light-sensitive anodes. The PMT’s used for the 450 nm and 550 nm fluorescence bands are Hamamatsu models H5783-03 and H6780-02, respectively.

Custom-built electronics were incorporated for the data acquisition [13

13. Vtech Engineering Corporation, http://www.vtechcorp.com.

]. A special daughter board, TPC-100, was used to process the elastic scatter signal from the 785 nm beam. This board detects the peak of the scattered light pulse above a threshold, T, and outputs a trigger pulse after a set delay from the peak (see Timing Chart, inset, Fig. 1). Triggering from the scattering signal peak, instead of using a fixed threshold level, removes uncertainty that would arise from varying size particles, and enabled more precise timing for firing the two UV beams. The TPC board also recorded both the pulse height and the integral of the charge under the scatter pulse to provide particle size information. A dual pulse processor circuit (DPP-266) was also designed to separately integrate the two sequential signal pulses from single PMT outputs separated in time by 400 ns. These data acquisition circuits demonstrated linearity over 105 dynamic range of the input pulse charge in order to accommodate the large variations in the fluorescence signals arising from the differences in particle size and composition. The DPP-266 board could be triggered at rates up to of 10 kHz. The temporally integrated analog signals from these custom data acquisition boards were sampled with a National Instruments M-series DAQ card. A data acquisition and processing program was written in LabVIEW to process the data from the DAQ card. For each particle, the raw recorded data consisted of: a time stamp (0.00036 second resolution), peak-value and integrated 785 nm elastic scattering, 266 and 355 nm elastic scattering, 350, 450 and 550 nm band fluorescence excited by the 266 nm laser, and 450 and 550 nm band fluorescence excited by the 355 nm laser. Additionally, the 266 and 355 excitation laser pulse energies were recorded from the photodiodes (PD), and the 785 nm laser power was monitored. These measurements constitute a total of 13 values that were recorded for each particle. The digitized signals were corrected for the PMT spectral response and gain based on manufacturer’s specifications, the measured filter spectral transmission and reflection efficiencies, and collection efficiency of the elliptical mirror. The fluorescence and scattering data from the particles were then normalized to the relative UV laser pulse energy and previously measured UV laser fluence. Finally, the resulting fluorescence data and scattering values were converted into numbers of photons emitted from each particle per mJ/cm2 of incident energy density and recorded as a binary file.

3. Experimental data and results

3.1 Spectral differentiation among biological aerosols

The 2-SPFA has been used to measure and quantify fluorescence from a wide range of biological and non-biological aerosol particles generated in the laboratory, representative of typical ambient air samples. Controlled samples of aerosol particles were generated using different dissemination techniques resulting in wet or dry aerosolization including: Collison nebulizer, Sono-tekTM generator [14

14. Sono-Tek Corporation, http://www.sono-tek.com/.

], Micro-dropletTM generator [15

15. H. Microdrop Technologies Gmb, http://www.microdrop.de/.

] and Pitt generator. These lab-generated aerosols were entrained in HEPA filtered dry nitrogen, prior to introduction into the 2-SPFA. Laboratory calibration particles that were analyzed included: undoped and dye doped polystyrene latex (PSL) spheres (Thermo Scientific, Inc.) and fused silica spheres of varying sizes in the range of 0.7 to 8.0 microns. Some fluorescence measurements obtained with a similar system, have been presented previously [8

8. V. Sivaprakasam, A. Huston, C. Scotto, and J. Eversole, “Multiple UV wavelength excitation and fluorescence of bioaerosols,” Opt. Express 12(19), 4457–4466 (2004). [CrossRef] [PubMed]

,16

16. V. Sivaprakasam, A. Huston, H. B. Lin, J. D. Eversole, P. Falkenstein, and A. Schultz, “Field test results and ambient aerosol measurements using dual wavelength fluorescence excitation and elastic scatter for bioaerosols”, SPIE conference,” Proceedings 6554, R5540 (2007).

].

LIF signatures of some of the common organic and inorganic sample materials such as: kaolin, Arizona road dust (ARD) and Sporisorium Cruentum (SC) fungal spores; BW agent simulants including: bacterial spores, Bacillus atrophaeus, ssp. globigii (BG) from Dugway Proving Grounds (DPG), bacterial vegetative cells, Bacillus subtilis (BS) in tryptic soy broth (TSB) media and Brucella neotomae (BN) in Brucella broth; and proteins: albumin and ovalbumin (OV) are shown in Fig. 2
Fig. 2 Fluorescence signatures of selected biological samples and inorganic samples representative of ambient air. As described in the text, the sum of emission bands for 266 nm and 355 nm excitation are plotted as the horizontal and vertical axis respectively. The use of these two excitation wavelengths shows an ability to differentiate many of these samples into reasonably distinct groups.
. These aerosol particles were generated using a Collison nebulizer, with sufficient dilution to result in relatively small particles around 1.5 micron aerodynamic diameter. For bacterial spore samples these particles would be primarily composed of a single organism. The 2-SPFA elastic scatter threshold was set to trigger on particles greater than 0.7 μm in diameter based on PSL measurements, and size distributions were monitored simultaneously using TSI Inc, aerodynamic particle spectrometer (APS) model 3321.

After relevant normalization, the corrected aerosol particle data sets still include 8 independent channels so that analysis of discrimination involves multivariate classification methods that are not easily visualized directly. Results of that analysis will be presented separately, but for this article, we show differences that can be seen directly in 2-dimensional data plots using selected channels. In Fig. 2, the emission in the three 266 nm excited fluorescence bands are summed and plotted against the sum of the two 355 nm excited fluorescence bands to provide a general sense of the classification capability that this type of data can provide. Each point on the scatter plot represents an individual aerosol particle, and data from approximately 104 particles was plotted for each sample composition. Two contour lines are plotted for each type of particle, which helps clarify cases where there is strong overlap between samples. The kaolin and ARD samples show very little fluorescence in any channel, and form a cluster in the lower left corner of the plot, kaolin particles exhibiting slightly higher signal levels than ARD. The bacterial spore, BG, and vegetative cells, BS and BN, have higher fluorescence signals for both excitation wavelengths and show separable clustering from inorganic samples as well as each other. Note that some of the bacterial aero-sols show multimode distributions, and this feature will be discussed in Sec. 3.6. The tightness and the proximity of these distributions depend on the homogeneity and the growth conditions of the sample. The fungal spores (SC), represent naturally occurring biological aerosol particles (and potential interferents), and Fig. 2 shows a fluorescence distribution in this 2 dimensional data plot that is separable from the biological simulants as well as inorganic samples. Note that this statement would not be true if either excitation wavelength were used alone. The size distributions of the particles (as reflected in their elastic scattering) have not been taken into account in this representation. The aerosolized fungal spores were significantly larger in size, roughly 6.5 μm in mean diameter as measured with the APS unit, compared to a typical mean of 1.5 μm diameter for the bacterial samples shown in Fig. 2. If particle size information was used, even greater separation of these two sample groups could be achieved. Pure protein samples, ovalbumin (OV) and albumin, represent bio-toxin surrogates, and their signatures are also easily separated from other samples due to the strong presence of tryptophan emission in the 350 nm band from 266 nm excitation. The purer protein samples show very little fluorescence in any of the other wavelength bands or from 355 nm excitation, and hence are easily identifiable from other bio-aerosol samples.

3.2 Spectral distribution of ambient air and combustion aerosols

For sampling ambient aerosols, a 2″ diameter pipe was installed in the exterior lab wall through which outside air could be drawn. Ambient air was sampled by the 2-SPFA for extended periods of time to characterize natural outdoor aerosol backgrounds. In Fig. 3 fluorescence data from ambient air background particles (Bkgrd) measured over a 2-hour period is shown, overlaid with data from some of the samples shown in Fig. 2. Note that the axes values in Fig. 3 are specific fluorescence spectral bands (data channels), rather than the sums of emission intensities from all the bands. Data in Fig. 3 are plotted as a scatter plot to illustrate the observation that small fractions of ambient particles occur in essentially all parts of the fluorescence feature space. Although the concentrations of these naturally occurring fluorescent particles are relatively low, depending on what specific region of feature space is being examined, their presence is significant because they establish a background threshold level that a new population of fluorescent particles would need to exceed in order to be detected by an automated algorithm.

Diesel soot is known to have UV fluorescence properties that may act as an interferent for UV LIF aerosol detection systems. Consequently, the ambient outdoor air sampling port was also used to obtain combustion engine exhaust data from diesel vehicles. Ambient air background was sampled immediately before and after operating a diesel engine vehicle in the vicinity of the inlet pipe to provide a comparison population of background data of the air with diesel combustion particles present. Figure 4
Fig. 4 Shows the fluorescence signatures in two 266 nm excited spectral bands from background aerosol particles sampled (a) without, and (b) with, diesel exhaust present. The new highly fluorescent population present in the Bkgrd particles in (b) centered at coordinates (3.5 × 104, 3 × 103) is diesel exhaust particles.
illustrates data contrasting ambient air particles with and without diesel exhaust particles present. In Fig. 4(a) the same outdoor particle sample data shown in Fig. 3, just prior to starting the diesel engine vehicle, is now presented as contour plots to show the population distribution within the fluorescence feature space. Figure 4(b) shows typical sample data immediately following, when the diesel engine was running. In Fig. 4(b) a bimodal distribution is seen in the ambient air data with a population of relatively highly fluorescent particles appearing in addition to the non-fluorescent particle population observed before (and after) the diesel exhaust was present. The new population was shown to result from the diesel engine exhaust by repeated trials with, and without, the engine running.

The clear separation of these two populations as shown in Fig. 4(b) permits the obvious identification of the second, highly fluorescent particle group as diesel combustion products. As in Fig. 3, data from some of the lab-generated aerosol samples has been overlaid on Fig. 4 (also as contours). In Fig. 4(a) the kaolin sample can be seen to represent a significant subpopulation of ambient outdoor aerosol particles that are essentially non-fluorescent. The ambient aerosol sample also includes populations of particles showing significant 450 nm band fluorescence that extends their contour sufficiently to overlap the fungal spore (SC) sample region. However, in Fig. 4(b) the emission from the diesel exhaust particle population clearly overlaps with the signature profiles of the bacterial spore sample, BG DPG, and bacterial vegetative cell particles, BN. Fluorescence signatures from aerosolized NIST diesel particulate matter (DPM) (black powder) were also measured, but data from these particles was quite different compared to diesel particles measured directly from vehicle exhaust, with the NIST DPM typically having very low UV LIF values (e.g. similar to kaolin).

In Figs. 3 and 4 the data channels (plot axes) are from two emission band intensities excited by 266 nm wavelength light. Figure 5
Fig. 5 The fluorescence signature of ambient outdoor aerosol particles sampled with diesel exhaust present, showing the comparison between 266 nm and 355 nm excited fluorescence. The left graph (a) shows data plotted for two fluorescence channels excited by 266 nm laser light. In the graph on the right (b) the vertical axis is plotted for the 450 nm fluorescence band excited by 355 nm laser light. In this plot the diesel aerosol population is seen to be significantly better separated from the bacterial aerosol samples, BG and BN.
illustrates the additional discrimination provided by data from 355 nm wavelength excitation. Figure 5(a) shows outdoor aerosol data with diesel particles present using data from the 350 nm and 450 nm emission bands from 266 nm excitation as shown in Fig. 4(b) but with different lab-generated samples overlaid. In Fig. 5(b) the same particle data is presented using 450 nm emission from 355 nm excitation as the vertical axis. Visual comparison of Figs. 5(a) and 5(b) shows the diesel particle population is significantly separated from the bacterial samples by including the 355 nm excited 450 nm band feature space. The fluorescence in the 450 band is observed to be greater for BG spores for the 355 nm excitation compared to the 266 nm excitation and vice-versa for diesel, thus comparing the ratio of the 450 nm band for the two excitations result in better discrimination of diesel engine particles from bacterial sample particles. This qualitatively demonstrates one of the primary advantages of using both 355 nm and 266 nm excitation wavelengths. Other differences are also evident in the fluorescence characteristics of some of the simulant samples in comparing Figs. 5(a) and 5(b). Vegetative cells BN and BS show greater separation for 266 nm excitation (Fig. 5(a)), compared to 355 nm excitation (Fig. 5(b)) while signatures for OV and albumin show more separation in Fig. 5(b) compared to Fig. 5(a).

3.3 Linearity of fluorescent emission with incident energy

To explore the possibility of saturation effects, a power-law analysis was performed for the data presented in Fig. 6, and the results are shown in Table 1

Table 1. Linearity of fluorescencea

table-icon
View This Table
| View All Tables
. A least squares fit to a straight line was performed on the logarithm of the emission intensity values first, for all the data points for each sample (Column 2), and then repeated for a subset of the data excluding values of the incident fluence greater than 1 mJ/cm2 (Column 3). The resulting values for the fitted slopes of the log emission vs. the log excitation are shown in columns 2 and 3 below. For all samples, the exponent values in column 3 are closer to unity than their corresponding value in column 2. These results suggest that potential saturation for most of the biological material samples could be beginning to occur at higher fluences. However, additional study at fluences higher than we were able to achieve would be necessary to confirm that possibility.

Reference PSL particles appear to be an exception to the linear dependence model, showing not only potential saturation effects at higher fluences but a power-law dependence that was sub-linear (exponent less than one) even at lower incident fluences. The measured deviation is greater than our apparent experimental uncertainty. While the source of this behavior is not understood at present, it may be connected to the high absorptivity of PSL material at these wavelengths. The estimated absorption depth for the PSL material is 1.6 μm at 266 nm. The silica particles had minimal fluorescence and thus yielded signals slightly higher than stray light signal at each of the fluence settings. The exponent value obtained for silica essentially characterizes the stray light noise response of the system as linearly dependent on the incident fluence. The exponent values obtained for the fluorescence measured in the 450 nm band showed very similar characteristics to that shown in Table 1 for the 350 nm fluorescence band. A similar computation was carried out for power-law exponent values for the second excitation pulse that was held constant at 0.5 mJ/cm2. The results showed that essentially all the samples emission from the second excitation remained constant within experimental error, indicating that no measureable photobleaching or other photochemical degradation effects were seen at these fluences.

3.4 Fluorescence and elastic scatter dependence on particle size

For BW aerosol detection, particles in the “respirable range” from 1 to 10 μm have been of primary interest. For a given aerosol material, fluorescence and elastic scatter intensities from individual particles are obviously strongly dependent on particle size. Although an extremely large compilation of prior computational and theoretical model work exists for elastic scattering, relatively little prior investigation has been focused on fluorescent emission dependence on particle size or shape. As part of the study in this paper, fluorescence and elastic scattering intensities were measured for aerosol particles of a constant composition over a range of particle sizes. Aerosols with narrow size distributions were generated in a range of 1 to 7 μm in aerodynamic diameter using a MicrodropletTM [15

15. H. Microdrop Technologies Gmb, http://www.microdrop.de/.

] generator. These particles were interrogated simultaneously by the 2-SPFA, and an APS. Particles generated using this technique had a size distribution FWHM of typically less than 14% of the mean diameter for particles of homogeneous composition such as ovalbumin, while for particles of heterogeneous composition such as bacterial cells in media, the variation in size increased to as much as 22%. As examples of fluorescence dependence on particle size, Fig. 7
Fig. 7 Fluorescence signature of two bioagent simulants; ovalbumin and Pantoea agglomerans in TSB media ranging in size from 1 μm to 7 μm. Fairly tight fluorescence distribution is measured for the distinct sizes.
plots the measured values for the 266 nm excited 350 nm band emission versus the 355 nm excited 450 nm band emission band for discrete particle sizes composed of ovalbumin (Sigma-Aldrich grade VI) and vegetative bacteria sample Pantoea agglomerans (PA) grown in TSB media. The scatter plots are shown along with 2 contour lines for each sample and the modes of the various size distributions are labeled on the plot. The fluorescence signals for the individual size distributions are sufficiently tightly clustered to clearly differentiate the size increments shown. As the particle size increases from 1 μm to 7 μm, the emission signal increases, spanning 2-3 orders of magnitude in dynamic range.

In order to quantify the size response, the mean fluorescence signal measured for ≈104 particles for the various types of particles were studied as a function of particle size. Figure 8
Fig. 8 Mean fluorescence measured for a number of calibration and bioaerosol particles measured as a function of particle size. The fluorescence measured in the 350 nm band is plotted with the standard deviation plotted as the error bars. A power fit was performed to demonstrate the relationship between fluorescence and particle size and is labeled for each particle type.
shows graph of the 266 nm excited 350 nm fluorescence band intensity as a function of the APS aerodynamic particle size. A power-law fit was performed on each sample, and the exponent value is labeled for each sample type. This analysis was also performed on all emission bands for both the 266 nm and 355 nm excitations, and results are shown in Table 2

Table 2. Fluorescence dependence on particle sizeb

table-icon
View This Table
| View All Tables
. For homogeneous isotropic particles, in which the absorption coefficient is low enough to permit roughly uniform excitation throughout the particle, the fluorescence emission is expected to be proportional to the particle volume (i.e. total number of fluorophores) resulting in a power-law exponent value of 3. As can be seen from Table 2, for homogeneous particles such as ovalbumin and dye doped PSL particles (green dye doped PSL (Thermo Scientific, Inc.)) the observed exponent values range between 2.4 and 3.0 for most of the emission bands. For particles of mixed composition like bacterial cells in media, the fluorophores may not be evenly distributed throughout the particle, and this could lead to compromised particle size dependence. Power-law exponent values as low as 1.6 were observed for such inhomogeneous particles. Other factors that might influence the fluorescence yield as particle size is increased are: non-uniform absorption of the excitation radiation and/or re-absorption of the emitted fluorescence.

The fluorescence emitted in the 450 nm and 550 nm bands for both excitations of the PSL series has repeatedly shown weaker fluorescence yield for the 7 μm particle compared to the 5 μm particle, and this alters the value of fluorescence versus particle size exponent. To illustrate this point, the power-law fit was performed on PSL particles including and excluding 7 μm particles and both sets of values are listed in Table 2. As the fluorescence signal emitted in the 350 nm band doesn’t show much degradation with increasing size particles, the power-law dependence doesn’t vary much in this band as for the two above mentioned cases. The exponent values change significantly for the two longer emission bands for both laser excitation pulses when the 7 μm PSL particles are excluded, illustrating that saturation and/or re-absorption of the emitted light are likely effects for larger PSL particles

For all the measurements made using the 2-SPFA, the elastic scatter data was recorded at 785 nm, 266 nm and 355 nm wavelengths for each particle. To explore the elastic scatter dependence on particle size, a range of particle sizes composed of: silica particles, PSL particles and NIST diesel particulate matter (DPM) were studied. Similar to the fluorescence data discussed previously, a power-law fit was performed on the elastic scatter data as a function of particle size, and the results are summarized in Table 3

Table 3. Elastic scatter dependence on particle sizec

table-icon
View This Table
| View All Tables
. The integrated elastic scatter signal depends on the geometry of the collection optics, as well as the scattering parameter (2πr/λ where r is the equivalent sphere radius) and the refractive index (real and imaginary) of the particles being interrogated. Taking into account all the apertures milled into the elliptical reflector for the laser beams and particle flow access, a calculation was performed for the same set of particles at the 3 wavelengths of interest using a model based on Mie theory (assumed spherical particle shape). The index of refraction data available from literature for silica [19

19. I. H. Malitson, “Interspecimen comparison of the refractive index of fused silica,” J. Opt. Soc. Am. 55(10), 1205–1209 (1965). [CrossRef]

], thin nm layer of PSL [20

20. L. Yan, “Characterization of Engineered Nanomaterial by spectroscopic Ellipsometry,” Horiba Scientific Application Note, Biotechnology UVISEL, SE26.

], PSL particle available at 785 nm [21

21. X. Ma, J. Q. Lu, R. S. Brock, K. M. Jacobs, P. Yang, and X. H. Hu, “Determination of complex refractive index of polystyrene microspheres from 370 to 1610 nm,” Phys. Med. Biol. 48(24), 4165–4172 (2003). [CrossRef]

] and black carbon [22

22. V. A. Markel and V. M. Shalaev, “Geometrical renormalization approach to calculating optical properties of fractal carbonaceous soot,” J. Opt. Soc. Am. A 18(5), 1112–1121 (2001). [CrossRef]

] was used for these calculations. In the same way as the experimental results, a power-law fit was performed on the computed scatter intensities as a function of particle size, and the resulting slopes are also listed in Table 3. As silica particles have relatively low absorption across the visible region, we measure the highest slope values for these particles (approximately values of 2, meaning that the scattered intensity is proportional to the cross sectional area of the particle). For silica, the experimental values agree reasonably well with the model calculations. As absorption comes into play for PSL particles, we can observe less than quadratic dependence for all excitations with an almost linear relationship at 266 nm. Higher than expected elastic scatter signal has been consistently measured from 1 μm particles (size parameter of 4) for 785 nm excitation (i.e., value for 1 μm size particles falls consistently above the fit). Such an observation was also made for the fit for the theoretical Mie calculations. Perhaps for scattering parameter values of 4, more of the forward lobe is collected compared to elastic scatter from particles with scattering parameter ≥ 8, thus skewing the power-law fit. By excluding the 1 μm data points for the 785 nm excitation the calculated Mie theoretical values and experimental data fit a linear power-law dependence better, and discrepancy between experimental and theoretical values also decreased significantly. Thus computed power dependence values are tabulated in column 6 of Table 3. As an example by excluding the 1 μm silica particles, the percent difference between theory and experiment was reduced to 4% from −17%.

The elastic scatter signals for NIST DPM, which is a highly absorbing (black powder) particle shows no measurable signal dependence on particle size for all three wavelengths studied. The experimental values show no correlation to the calculated values obtained using carbon black refractive index data. Prior study measuring the angular scatter signal from DPM concluded that Mie theory cannot adequately account for the measured angular scatter signal from such particles, and a different kind of particle scattering model such as a fractal surface description would need to be considered to obtain reasonable agreements [23

23. C. D. Litton, “Studies of the measurement of respirable coal dusts and diesel particulate matter,” Meas. Sci. Technol. 13(3), 365–374 (2002). [CrossRef]

].

These measurements show that elastic scatter alone cannot be reliably used to provide size measurement of ambient particles. If an independent measure of the particle size was to be made, then variations in the elastic scatter could be used to obtain additional absorption information about the particles. Moreover, as the functional form of the particle size dependence for fluorescence and the elastic scatter vary, (depending on the composition of the particle) normalizing the fluorescence to the elastic scatter signal to de-convolve the particle size is only possible for well characterized materials.

3.5 Fluorescence dependence on aerosol generation technique

In order to study the effect of moisture on the fluorescence signature, aerosols of three of the simulants: PA, BG and OV, were generated from a dry powder using a dissemination technique known as a Pitt generator. This method produces a wide particle size distribution, as shown by the data graphed on Fig. 9(a)
Fig. 9 (a) Particle size distribution measured by APS 3321 and (b) the fluorescence of three of the simulants generated using a Pitt generator (dry aerosolization) and Collison nebulizer (wet aerosolization).
, measured using an APS. For comparison, these same materials were suspended/dissolved in water, and a Collison nebulizer was used for droplet generation. As discussed earlier, concentrations of these materials were adjusted so that the resulting dried aerosol from the nebulizer has a mean diameter of about 1 micron. In Fig. 9(b) the fluorescence values measured in the 350 nm band for 266 nm excitation are plotted versus those from the 450 nm band. The fluorescence distributions for the dry disseminated particles form long, straight strands with unit slope on the logarithmically scaled fluorescence plot. The length of these narrow strands is due to the wide size distribution of these aerosols. The unit slope results from the fact that emission in these two spectral bands has the same proportion regardless of the particle size. The fluorescence distributions for the Collison-generated particles of the same sample materials from solution droplets are also shown on the same plot. These nebulizer-generated aerosols are more tightly clustered around a central mean size. The fluorescence from ovalbumin generated using the droplet nebulizer overlaps and extends the powder disseminated distribution to lower values as would be expected from smaller particles. The two distributions from the ovalbumin aerosols show that the characteristic emission properties of the ovalbumin particles appear to be independent of the particle generation method. However, for the other two sample materials, PA and BG, the fluorescence distributions do not align or overlap in a smooth way. The fluorescence in the 450 nm emission band for PA and BG shows 2 to 4 times lower fluorescence for the wet disseminated particles compared to the dry particles. Similar differences were evident when comparing the 355 nm excited 450 nm fluorescent band. The results reported here compare dry aerosols generated using either a dry or wet generation technique. In an earlier study (Faris et al. [17

17. G. W. Faris, R. A. Copeland, K. Mortelmans, and B. V. Bronk, “Spectrally resolved absolute fluorescence cross sections for bacillus spores,” Appl. Opt. 36(4), 958–967 (1997). [CrossRef] [PubMed]

]), a wet aerosol generation technique was used, and particles were measured either prior to or after drying. They reported a factor of 4 lower fluorescence yields from dry B. Subtilis aerosol compared to wet aerosols. A plausible explanation has not been offered for either set of observations.

3.6 Fluorescence dependence on biological sample protocol

For some of the microbiological materials, fluorescence spectra displayed multi-mode distributions even though the size distribution was single mode. Presumably, these structured emission profiles are indicative of composite fluorescent materials such as media or metabolic products in addition to the organisms themselves. To test this idea, samples were investigated as aliquots in which one remained as prepared in the media, another was separated from the media material by centrifugation and washed in pure water (3X), and a third sample is formed from the leftover media supernatant. BG spores (DPG), and Bacillus subtilis (BS) and Yersenia rohdei vegetative cells in TSB media were chosen for this study. Figure 10
Fig. 10 Fluorescence signatures of 3 simulants (a) BG DPG spores in TSB mod media, (b) BS DPG Vegetative cells in TSB and (c) Yersinia rohdei in TSB, in media (blue), the washed particles (red) and the leftover growth media (green) showing the fingerprint of the components and the mixture. In some cases the fluorescence of the sum equals to the fluorescence of the components and the patterns shifts for other cases.
shows 266 nm excited, 350 nm band fluorescence values plotted vs. the 450 nm band as contour plots for (a) BG spores, (b) BS Vegetative cells and (c) Y. rohdei. In all three of the plots the sample in media is shown in dark blue, the spent leftover media in green, the washed sample in red and the pure nutrient broth in light blue. The spectral signature for Y. rohdei in media shows 3 distinct peaks, or maxima, with the signature for the spent media overlapping with two of these peaks, while that of the washed sample shows a strong correlation with the third peak. Similar results has been observed consistently with multiple Y. rohdei samples, and the growth media typically shows higher fluorescence intensities in the longer wavelength emission bands compared to the organism. However, the fluorescence signatures of the BG spores and BS vegetative cells do not show similar patterns. From the plots for the BG spores,the spores in the media show two distributions with the left over media overlapping with one of the distributions. The signature of the washed sample doesn’t overlap with the second distribution as one would expect, but instead shows a disjoint distribution from the spectrum of the sample in media showing that the fluorescence of the particle is spectrally modified in the presence of surrounding media. The washed sample shows higher fluorescence in the 350 nm band compared to the sample in media and lower fluorescence in the 450 nm band. One possible explanation for this behavior is that some of the emitted 350 fluorescence from the microorganisms embedded in the particle is absorbed by the surrounding media, which then re-emits that energy at longer wavelengths. Additionally, or alternatively, the incident intensity interior to the particle could be significantly attenuated by the presence of the media. The fluorescence distributions of the BS vegetative cells show a similar pattern. BS cells in media and the spent media are so overlapped that they are essentially indistinguishable, while the emission distribution for the washed cells are completely disjoint. In this case, the washed cells emission distribution is shifted vertically down with roughly an order of magnitude lower fluorescence in the 450 nm band, but roughly the same intensity in the 350 nm band. Similar spectral pattern is observed from the 450 nm emission band for the 355 nm excitation source as well. In summary, significant and non-intuitive changes are observed in the samples of the same microorganism depending on the sample preparation in terms of whether the growth media is present or not. Further investigation is required to determine the detailed mechanisms involved in each case. For example, for vegetative cell samples the possibility, or degree, of cellular lysing would be an important parameter to be determined. Moreover, a quantitative model of particle fluorescence is needed. Such a model would permit a direct comparison for the spectral mixing of multiple components with their own absorption and emission spectral profiles, in addition to calculating the source function of the incident excitation radiation in the particle interior.

As a final consideration, we wanted to determine the effect of gamma irradiation on at least a couple of samples in terms of the fluorescence spectra. The fluorescence spectra of Yersinia rhodei in TSB media and DPG BG, also in TSB media, were obtained for both non-irradiated and gamma irradiated samples. Figure 11
Fig. 11 Fluorescence signature of 2 simulants (a) Yersinia rohdei in TSB and (b) BG DPG in TSB, before and after gamma irradiation.
shows plots of the 266 nm excited, 350 nm band emission vs. the 355 nm excited 450 nm band fluorescence as contours for (a) Y. rohdei and (b) BG. For the Y. rohdei samples, the tri-modal distribution of the cells in media is shown in green, and the result after gamma radiation shows a single emission mode, which is roughly an average of the spent media and washed cell samples. The intensities of the two samples are strongly over-lapped. For BG samples a similar result is obtained. Overall, the main effect of gamma irradiation appears to be a noticeable reduction in the media component of the signature that results in an increased emphasis of the cellular spectral component, but no dramatic effects in the overall emission intensities are observed.

4. Summary

The fluorescence of a number of calibration standard particles and biological simulants has been studied. The employment of two UV excitation wavelengths was shown to increase discrimination among several biological materials, and especially between BW simulants and diesel exhaust. The fluorescence of aerosol particles was studied as a function of particle size, and for some cases where composition homogeneity and smaller absorption coefficients are obtained, the fluorescence of homogeneous samples show close to a cubic dependence on the particle size (proportional to particle volume). However, for inherently heterogeneous particles, this size dependence shows much greater variation. Particle absorption was seen to play a strong role in the dependence of elastic scatter on particle size. For highly absorbing particles, such as NIST diesel particulate matter, no size dependence was observed.

However, the power-law exponent for particles or wavelengths with weaker absorption coefficients could range from values approaching 2 (proportional to the cross sectional area of the particle) down to 1. This suggests the need for a particle size measurement that is not determined soley by elastic scatter. The linearity of the fluorescence yield was studied as a function of incident energy for particles of different known compositions and known sizes. No photo-bleaching effects were observed, and no saturation effects were observed for fluences of 5 mJ/cm2 or less. The presence and absence of growth media plays a strong role on the fluorescence signature. Understanding these signatures will be important in order to refine detection algorithms and defining the threat features of interest.

Acknowledgments

This work was supported by chemical and biological defense tech-base, administered by the Defense Threat Reduction Agency (DTRA). DTRA was founded in 1998 to integrate and focus the capabilities of the Department of Defense that address the weapons of Mass Destruction (WMD) threat. The mission of the DTRA is to safeguard America and its allies from WMD threat (e.g. chemical, biological, radiological, nuclear and high yield explosives) by providing capabilities to reduce, eliminate and counter the threat and mitigate its effects. Under DTRA, Department of Defense resources, expertise and capabilities are combined to ensure the United States remains ready and able to address the present and future WMD threats. For more information on DTRA, visit _www.dtra.mil/_ <file://www.dtra.mil/>.

References and links

1.

R. G. Pinnick, S. C. Hill, P. Nachman, J. D. Pendleton, G. L. Fernandez, M. W. Mayo, and J. G. Bruno, “Fluorescent particle counter for detecting airborne bacteria and other biological particles,” Aerosol Sci. Technol. 23(4), 653–664 (1995). [CrossRef]

2.

M. Seaver, J. D. Eversole, J. J. Hardgrove, W. K. Cary Jr, and D. C. Roselle, “Size and fluorescence measurements for field detection of biological aerosols,” Aerosol Sci. Technol. 30(2), 174–185 (1999). [CrossRef]

3.

F. L. Reyes, T. H. Jeys, N. R. Newbury, C. A. Primmerman, G. S. Rowe, and A. Sanchez, “Bio-aerosol fluorescence sensor,” Field Anal. Chem. Technol. 3(4-5), 240–248 (1999). [CrossRef]

4.

Y. L. Pan, J. Hartings, R. Pinnick, S. Hill, J. Halverson, and R. Chang, “Single particle fluorescence spectrometer for ambient aerosols,” Aerosol Sci. Technol. 37(8), 628–639 (2003). [CrossRef]

5.

Y. L. Pan, S. Holler, R. K. Chang, S. C. Hill, R. G. Pinnick, S. Niles, and J. R. Bottiger, “Single-shot fluorescence spectra of individual micrometer-sized bioaerosols illuminated by a 351- or a 266-nm ultraviolet laser,” Opt. Lett. 24(2), 116–118 (1999). [CrossRef]

6.

P. P. Hairston, J. Ho, and F. R. Quant, “Design of an instrument for real-time detection of bioaerosols using simultaneous measurement of particle aerodynamic size and intrinsic fluorescence,” J. Aerosol Sci. 28(3), 471–482 (1997). [CrossRef] [PubMed]

7.

P. H. Kaye, J. E. Barton, E. Hirst, and J. M. Clark, “Simultaneous light scattering and intrinsic fluorescence measurement for the classification of airborne particles,” Appl. Opt. 39(21), 3738–3745 (2000). [CrossRef]

8.

V. Sivaprakasam, A. Huston, C. Scotto, and J. Eversole, “Multiple UV wavelength excitation and fluorescence of bioaerosols,” Opt. Express 12(19), 4457–4466 (2004). [CrossRef] [PubMed]

9.

P. H. Kaye, W. R. Stanley, E. Hirst, E. V. Foot, K. L. Baxter, and S. J. Barrington, “Single particle multichannel bio-aerosol fluorescence sensor,” Opt. Express 13(10), 3583–3593 (2005). [CrossRef] [PubMed]

10.

K. Davitt, Y.-K. Song, W. Patterson Iii, A. V. Nurmikko, M. Gherasimova, J. Han, Y.-L. Pan, and R. K. Chang, “290 and 340 nm UV LED arrays for fluorescence detection from single airborne particles,” Opt. Express 13(23), 9548–9555 (2005). [CrossRef] [PubMed]

11.

Y. L. Pan, V. Boutou, J. R. Bottiger, S. S. Zhang, J. P. Wolf, and R. K. Chang, “A Puff of Air Sorts Bioaerosols for Pathogen Identification,” Aerosol Sci. Technol. 38(6), 598–602 (2004). [CrossRef]

12.

V. Sivaprakasam, T. Pletcher, J. E. Tucker, A. L. Huston, J. McGinn, D. Keller, and J. D. Eversole, “Classification and selective collection of individual aerosol particles using laser-induced fluorescence,” Appl. Opt. 48(4), B126–B136 (2009). [CrossRef] [PubMed]

13.

Vtech Engineering Corporation, http://www.vtechcorp.com.

14.

Sono-Tek Corporation, http://www.sono-tek.com/.

15.

H. Microdrop Technologies Gmb, http://www.microdrop.de/.

16.

V. Sivaprakasam, A. Huston, H. B. Lin, J. D. Eversole, P. Falkenstein, and A. Schultz, “Field test results and ambient aerosol measurements using dual wavelength fluorescence excitation and elastic scatter for bioaerosols”, SPIE conference,” Proceedings 6554, R5540 (2007).

17.

G. W. Faris, R. A. Copeland, K. Mortelmans, and B. V. Bronk, “Spectrally resolved absolute fluorescence cross sections for bacillus spores,” Appl. Opt. 36(4), 958–967 (1997). [CrossRef] [PubMed]

18.

Y. L. Pan, R. G. Pinnick, S. C. Hill, S. Niles, S. Holler, J. R. Bottiger, J. P. Wolf, and R. K. Chang, “Dynamics of photon-induced degradation and fluorescence in riboflavin microparticles,” Appl. Phys. B 72, 449–454 (2001).

19.

I. H. Malitson, “Interspecimen comparison of the refractive index of fused silica,” J. Opt. Soc. Am. 55(10), 1205–1209 (1965). [CrossRef]

20.

L. Yan, “Characterization of Engineered Nanomaterial by spectroscopic Ellipsometry,” Horiba Scientific Application Note, Biotechnology UVISEL, SE26.

21.

X. Ma, J. Q. Lu, R. S. Brock, K. M. Jacobs, P. Yang, and X. H. Hu, “Determination of complex refractive index of polystyrene microspheres from 370 to 1610 nm,” Phys. Med. Biol. 48(24), 4165–4172 (2003). [CrossRef]

22.

V. A. Markel and V. M. Shalaev, “Geometrical renormalization approach to calculating optical properties of fractal carbonaceous soot,” J. Opt. Soc. Am. A 18(5), 1112–1121 (2001). [CrossRef]

23.

C. D. Litton, “Studies of the measurement of respirable coal dusts and diesel particulate matter,” Meas. Sci. Technol. 13(3), 365–374 (2002). [CrossRef]

OCIS Codes
(010.1100) Atmospheric and oceanic optics : Aerosol detection
(290.5850) Scattering : Scattering, particles
(300.2530) Spectroscopy : Fluorescence, laser-induced

ToC Category:
Atmospheric and Oceanic Optics

History
Original Manuscript: December 2, 2010
Revised Manuscript: February 3, 2011
Manuscript Accepted: February 4, 2011
Published: March 18, 2011

Virtual Issues
Vol. 6, Iss. 4 Virtual Journal for Biomedical Optics

Citation
Vasanthi Sivaprakasam, Horn-Bond Lin, Alan L. Huston, and Jay D. Eversole, "Spectral characterization of biological aerosol particles using two-wavelength excited laser-induced fluorescence and elastic scattering measurements," Opt. Express 19, 6191-6208 (2011)
http://www.opticsinfobase.org/vjbo/abstract.cfm?URI=oe-19-7-6191


Sort:  Author  |  Year  |  Journal  |  Reset  

References

  1. R. G. Pinnick, S. C. Hill, P. Nachman, J. D. Pendleton, G. L. Fernandez, M. W. Mayo, and J. G. Bruno, “Fluorescent particle counter for detecting airborne bacteria and other biological particles,” Aerosol Sci. Technol. 23(4), 653–664 (1995). [CrossRef]
  2. M. Seaver, J. D. Eversole, J. J. Hardgrove, W. K. Cary, and D. C. Roselle, “Size and fluorescence measurements for field detection of biological aerosols,” Aerosol Sci. Technol. 30(2), 174–185 (1999). [CrossRef]
  3. F. L. Reyes, T. H. Jeys, N. R. Newbury, C. A. Primmerman, G. S. Rowe, and A. Sanchez, “Bio-aerosol fluorescence sensor,” Field Anal. Chem. Technol. 3(4-5), 240–248 (1999). [CrossRef]
  4. Y. L. Pan, J. Hartings, R. Pinnick, S. Hill, J. Halverson, and R. Chang, “Single particle fluorescence spectrometer for ambient aerosols,” Aerosol Sci. Technol. 37(8), 628–639 (2003). [CrossRef]
  5. Y. L. Pan, S. Holler, R. K. Chang, S. C. Hill, R. G. Pinnick, S. Niles, and J. R. Bottiger, “Single-shot fluorescence spectra of individual micrometer-sized bioaerosols illuminated by a 351- or a 266-nm ultraviolet laser,” Opt. Lett. 24(2), 116–118 (1999). [CrossRef]
  6. P. P. Hairston, J. Ho, and F. R. Quant, “Design of an instrument for real-time detection of bioaerosols using simultaneous measurement of particle aerodynamic size and intrinsic fluorescence,” J. Aerosol Sci. 28(3), 471–482 (1997). [CrossRef] [PubMed]
  7. P. H. Kaye, J. E. Barton, E. Hirst, and J. M. Clark, “Simultaneous light scattering and intrinsic fluorescence measurement for the classification of airborne particles,” Appl. Opt. 39(21), 3738–3745 (2000). [CrossRef]
  8. V. Sivaprakasam, A. Huston, C. Scotto, and J. Eversole, “Multiple UV wavelength excitation and fluorescence of bioaerosols,” Opt. Express 12(19), 4457–4466 (2004). [CrossRef] [PubMed]
  9. P. H. Kaye, W. R. Stanley, E. Hirst, E. V. Foot, K. L. Baxter, and S. J. Barrington, “Single particle multichannel bio-aerosol fluorescence sensor,” Opt. Express 13(10), 3583–3593 (2005). [CrossRef] [PubMed]
  10. K. Davitt, Y.-K. Song, W. Patterson Iii, A. V. Nurmikko, M. Gherasimova, J. Han, Y.-L. Pan, and R. K. Chang, “290 and 340 nm UV LED arrays for fluorescence detection from single airborne particles,” Opt. Express 13(23), 9548–9555 (2005). [CrossRef] [PubMed]
  11. Y. L. Pan, V. Boutou, J. R. Bottiger, S. S. Zhang, J. P. Wolf, and R. K. Chang, “A Puff of Air Sorts Bioaerosols for Pathogen Identification,” Aerosol Sci. Technol. 38(6), 598–602 (2004). [CrossRef]
  12. V. Sivaprakasam, T. Pletcher, J. E. Tucker, A. L. Huston, J. McGinn, D. Keller, and J. D. Eversole, “Classification and selective collection of individual aerosol particles using laser-induced fluorescence,” Appl. Opt. 48(4), B126–B136 (2009). [CrossRef] [PubMed]
  13. Vtech Engineering Corporation, http://www.vtechcorp.com .
  14. Sono-Tek Corporation, http://www.sono-tek.com/ .
  15. H. Microdrop Technologies Gmb, http://www.microdrop.de/ .
  16. V. Sivaprakasam, A. Huston, H. B. Lin, J. D. Eversole, P. Falkenstein, and A. Schultz, “Field test results and ambient aerosol measurements using dual wavelength fluorescence excitation and elastic scatter for bioaerosols”, SPIE conference,” Proceedings 6554, R5540 (2007).
  17. G. W. Faris, R. A. Copeland, K. Mortelmans, and B. V. Bronk, “Spectrally resolved absolute fluorescence cross sections for bacillus spores,” Appl. Opt. 36(4), 958–967 (1997). [CrossRef] [PubMed]
  18. Y. L. Pan, R. G. Pinnick, S. C. Hill, S. Niles, S. Holler, J. R. Bottiger, J. P. Wolf, and R. K. Chang, “Dynamics of photon-induced degradation and fluorescence in riboflavin microparticles,” Appl. Phys. B 72, 449–454 (2001).
  19. I. H. Malitson, “Interspecimen comparison of the refractive index of fused silica,” J. Opt. Soc. Am. 55(10), 1205–1209 (1965). [CrossRef]
  20. L. Yan, “Characterization of Engineered Nanomaterial by spectroscopic Ellipsometry,” Horiba Scientific Application Note, Biotechnology UVISEL, SE26.
  21. X. Ma, J. Q. Lu, R. S. Brock, K. M. Jacobs, P. Yang, and X. H. Hu, “Determination of complex refractive index of polystyrene microspheres from 370 to 1610 nm,” Phys. Med. Biol. 48(24), 4165–4172 (2003). [CrossRef]
  22. V. A. Markel and V. M. Shalaev, “Geometrical renormalization approach to calculating optical properties of fractal carbonaceous soot,” J. Opt. Soc. Am. A 18(5), 1112–1121 (2001). [CrossRef]
  23. C. D. Litton, “Studies of the measurement of respirable coal dusts and diesel particulate matter,” Meas. Sci. Technol. 13(3), 365–374 (2002). [CrossRef]

Cited By

Alert me when this paper is cited

OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.


« Previous Article  |  Next Article »

OSA is a member of CrossRef.

CrossCheck Deposited