Detector for Airborne Biological Particles
Field of the Invention
This invention describes a method and apparatus for the detection of airborne biological particles.
Background to the Invention
In a wide variety of environmental, occupational, military and industrial scenarios, fine particles, typically within the size range from a few tenths of a micrometre to a few hundred micrometres, play an important role. Environmental airborne particles, usually comprising mineral dusts, combustion products and biological particles, which are carried by winds and other air movement, can result in breathing difficulties, allergic reactions a possible degradation of the body's immune system. In the military field, the deliberate generation of hazardous aerosols has posed a major threat since their first substantial use in World War I, and today a wide variety of biological and chemical weapons is believed to be possessed by both national governments and terrorist organisations.
The in-situ characterisation of airborne particles has therefore become an important objective in both civilian and military fields, and considerable effort has gone into developing techniques which can analyse certain particle parameters and provide some degree of identification or classification. Moreover, since even brief exposure to some of the aforementioned aerosols can damage health and may even prove fatal, the speed of response of the measurement technique has been an important consideration.
A potentially powerful technique of airborne particle analysis involves the introduction of individual particles into a near vacuum where they are fragmented using an intense laser light pulse. The resulting atomic and molecular fragments are then measured using a time-of-flight mass spectrometer or similar, yielding a detailed assessment of the material content of the particle. (See for example, Marijnissen J et al, 'Proposed on-line aerosol analysis combining size determination, laser induced fragmentation, and time-of-flight mass spectrometry', Journal of Aerosol Science, volume 19, pages 1307-1310, 1988). Such methods offer a high
degree of particle discrimination but remain expensive and cumbersome to implement and, because they are comparatively slow in terms of the rate at which individual particles can be analysed, they do not offer the real-time aerosol analysis capability (i.e.: response to a change in aerosol composition within a few seconds) desired in monitoring applications.
Of other possible particle characterisation techniques, those based on elastic optical scattering have become popular because they offer genuine real-time nondestructive particle analysis. Here, the term elastic denotes that the scattered light is at the same wavelength as the illuminating light. In their simplest form, optical scattering instruments are designed to draw ambient airborne particles through a measurement space. A light source, usually a laser, illuminates the measurement space and the particles scatter some radiation to an appropriately positioned detector. The magnitude of the scattered radiation may, to a first order, be used to determine particle sizes and number illuminated at any instant. Whilst comparatively straightforward to implement, simple light scattering techniques such as these do not yield sufficient information about the particles to provide anything other than a very superficial overview of the ambient aerosol. They do not, for example, provide any indication of the material nature of the particles; whether the particles are of solid or liquid form; or whether the particles are of biological or non-biological origin.
In order to discriminate more effectively between airborne particles of different types, a number of methods have been developed which measure multiple parameters from individual particles in addition to their (optical scattering) size. For example, analysis of the spatial distribution of light scattered by individual airborne particles passing through the measurement space of an optical scattering instrument has proved to be an effective method of improving particle discrimination. This is because the spatial pattern of scattered light contains information relating to the shape of the scattering particle. Examples of instrument geometries which embody this approach to spatial scattering analysis are described in: 'Portable Particle Analysers', Ludlow, I. K. and Kaye P H. European Patent EP 0 316 172, July 1992; 'Particle Asymmetry Analyser', Ludlow, I. K. and Kaye, P. H. European Patent EP 0 316 171 , Sept. 1992.; 'Apparatus and Method for the Analysis of Particle Characteristics using Monotonically Scattered Light', Kaye, P.H. and Hirst, E. US Patent 5,471 ,299. Nov. 28, 1995; and 'Hazardous Airborne Fibre Detector'. Hirst, E. and Kaye, P.H. UK Patent Application No: 9619242.2; filed 14th September 1996.
However, light scattering analysis instruments of the type described above cannot discriminate particles on the basis of their material structure. For example, a non- biological silicate-based particle may yield an essentially identical spatial scattering pattern to a biological cell of similar size and shape. In order to discriminate particles on the basis of their material structure it is necessary to employ other techniques such as an analysis of light which is scattered inelastically by the particle. Such light is manifest as either a fluorescence emission or, far more weakly, a Raman emission. Since useful Raman signals from individual microscopic particles in flow have, to date, proved unattainable, they will not be discussed further here. In contrast, several workers have demonstrated successful measurement of fluorescent spectra from single airborne particles and have used this technique to attempt particle discrimination on the basis of fluorescence.
For example, Pinnick et al ('Fluorescent Particle Counter for Detecting Airborne Bacteria and Other Biological Particles' Pinnick R G et al., Aerosol Science and Technology, volume 23, pages 653-664, 1995) developed an instrument in which a stream of airborne particles passes through a measurement space and is illuminated with light at 488nm wavelength from an Argon-Ion laser. The light excites some naturally occurring fluorophores within each individual particle in turn and the fluorescence emission spectrum from that particle between about 500nm and 800nm wavelength is recorded and analysed. Based on the fact that biological particles such as spores produced measurable fluorescence, the authors proposed the technique as a possible means of discriminating biological from other non- biological particles that may be present in an environment. Furthermore, it is known that the fluorescence spectrum from a particle will be a function of the excitation wavelength illuminating the particle, thus, if a particle is illuminated by one wavelength and, following the capture of the fluorescence spectrum, is illuminated instead by a second excitation wavelength, then the information available in these two discrete spectra to identify the material of the particle is significantly enhanced.
Other workers, (for example see Hairston P P et al, 'Design of an instrument for real-time detection of bioaerosols using simultaneous measurement of particle aerodynamic size and intrinsic fluorescence', Journal of Aerosol Science, vol. 28, no. 3, pages 471-482, 1997), have combined a measurement of the magnitude of fluorescence from a particle with a measure of its size, in this case the aerodynamic size of the particle. This dual-parameter measurement approach provides a greater
degree of particle discrimination than measurement of particle fluorescence alone. This method has been extended by Kaye P H et al 'Simultaneous light scattering and intrinsic fluorescence measurement for the classification of airborne particles', Applied Optics, volume 39, number 21 , pp 3738-3745, to incorporate a method of determining the shape of individual particles from an analysis to the spatial distribution of light scattered by the particle.
However, all of the methods described above involve the analysis of individual particles at normally high processing rates. Whilst offering a high degree of particle discrimination, the methods all suffer the same problem of being expensive and complex to implement. This high cost is a result of the requirement for an intense and well-collimated light source(s), usually a laser operating in the ultra-violet at 266nm or 355 nm wavelength, the requirement for precision optical systems, the need for complex and high-speed data processing electronics, and, normally, the need for an independent power generator to supply the instruments with electrical power over extended time periods. Because of the high cost of implementing the methods, the deployment of monitors based on them is normally limited to very small numbers of discrete monitors. In some cases, especially outdoor environments or areas of military conflict, the biological threat may appear anywhere across a large area, and the deployment of small number of discrete monitors is of limited value in rapidly detecting the threat, should it arise.
What is required, therefore, is a monitor which is of sufficiently low cost and small size that it may be manufactured and deployed in very large numbers across wide areas of potential risk, or even that it could be worn or carried by every individual person in the area who may be exposed to the biological hazard. Such a monitor would ideally meet the following specification:
1. Low cost.
2. Hand portable or person wearable. 3. No reagent requirements, (i.e. no requirement for recharging chemical or biochemical assay systems).
4. Unattended operation.
5. Typically 48-72 hours continuous operation using built-in battery power supply or similar.
6. Maximum response time of typically 10s i.e.: will detect the presence of biological particles in an environment within a time period short enough to prevent prolonged exposure of individuals to the hazard.
In one aspect, therefore, the present invention seeks to provide a low cost detector that can affordably be deployed in large numbers over a wide area to be monitored.
In another aspect the present invention seeks to combine the advantages of single particle illumination (ie: improved discrimination of biological from non-biological particles), with the advantages of multiple particle illumination (ie: simpler optical and mechanical design and lower cost).
Summary of the invention
According to a first aspect of the present invention there is provided an apparatus for the detection of airborne biological particles which comprises :
a zone through which air to be analysed flows, in use;
a source of illumination to illuminate airborne particles present in said zone and to excite the particles to fluoresce, said source being a source of non-monochromatic light; and
a detector to detect fluorescence from the particles as an indicator of the presence of biological particles.
The source is especially suitably a pulsed light source. The pulses of this preferably are at intervals that are linked to the rate of airflow through the zone whereby the air is substantially wholly replaced before the next pulse occurs.
Particularly preferably the source is a high intensity flash source such as a Xenon. Suitably a low pass filter is used between the source and the zone to allow only the lower wavelengths that are in the appropriate range to excite fluorescence to pass. Preferably these allow only radiation under 350 nm to pass.
A simple collimating means may also be used to help to direct the light from the source into the zone.
Particularly preferably the apparatus further comprises a fan as the means to cause airflow through the zone. The rate of operation of this fan suitably being synchronised to the rate of pulsing of the light source.
In one preferred embodiment the apparatus further comprises:
a) a detector to detect light scattered elastically by the particles and generate a first signal ; b) a said detector to detect fluorescence from the particles arranged to do so contemporary with detection of the scattered light and to generate a second signal ; and c) a processor to compare the magnitude of the first signal to that of the second signal as an indicator of the presence of biological particles.
In this aspect of the invention, the magnitudes of the fluorescence and elastically scattered light from an ensemble of airborne particles illuminated (particularly preferably by a pulse of exciting radiation such as that from a xenon flash lamp), are independently collected and compared.
Suitably, here no fluorescence spectral information is recorded / recordable , simply the magnitude of the fluorescence from all the particles present in the measurement space at the time of illumination by the exciting radiation. By comparing this magnitude with that of the elastically scattered light from the illuminated particles, a broad indication can be obtained as to the likely presence of biological particles within the ensemble of particles.
In a further preferred embodiment of the present invention the apparatus does have a fluorescence detector capable of detecting fluorescence spectra and which is configured to detect multiple spaced apart spectra, each spectrum from a respective airborne paticle.
Such detector suitably comprises an array of many detector elements, suitably in a matrix form. The detector is preferably a CCD array but may alternatively, for
example, comprise a CMOS array. An integrating light detection array is particularly preferred - ie an array where the individual pixels / elements accumulate light over a period of exposure time. This is of particular value to enable multiple spectra to be captured for each particle, for example when exposing the particles to two or more different wavelength bands of light in quick succession.
The apparatus suitably for this purpose has two or more light sources arranged to deliver different wavelengths / wavelength bands from each other. Each may , for example , be fronted by a different filter or one have a filter and another not . Alternatively or additionally one or more light source may have a rapidly switchable filter.This could potentially , for example, comprise an optical mask of LCD type or the like.
In the embodiment generally the apparatus suitably has a dispersive optical element, eg a diffraction grating or prism, between the zone and the fluorescence detector in order to image a spectrum from each particle onto the fluorescence detector.
A collimating means is suitably further provided between the zone and the dispersive optical element to direct light from particles in the zone onto the dispersive optical element.
Particularly preferably this embodiment of apparatus has a processor configured to receive a captured image from the detector array and scan it to identify the location of each of any one or more groups of spectra present , suitably providing a digital pattern of signal intensities for each spectrum. A processor for receiving and processing the captured image may be configured to compare any spectra or groups of spectra in the image against pre-recorded spectra or groups of spectra and to identify any substantial match.
According to a second aspect of the present invention there is provided an apparatus for the detection of airborne biological particles which comprises
a zone through which air to be analysed flows, in use;
a source of illumination to illuminate airborne particles present in said zone and to excite the particles to fluoresce; and
a detector to detect fluorescence from the particles as an indicator of the presence of biological particles, the detector being an optical array detector and the apparatus further having a dispersive optical element between the zone and the fluorescence detector in order to image a spectrum from each particle onto the fluorescence detector.
According to a further aspect of the invention there is provided a method of detecting airborne biological particles which comprises :
providing a zone through which air to be analysed flows, in use, illuminating airborne particles present in said zone to excite the particles to fluoresce, and detecting fluorescence from the particles contemporary with detection of the scattered light and comparing the detected fluorescence signal to the detected scattered light signal to provide a ratio as an indicator of the presence of biological particles.
According to a yet further aspect of the invention there is provided a method of detecting airborne biological particles which comprises :
providing an apparatus of the second aspect of the invention and first illuminating airborne particles present in said zone with a first wavelength or band of wavelengths to excite the particles to fluoresce, and detecting fluorescence spectra from the particles with the detector array and immediately thereafter illuminating the airborne particles present in said zone with a second wavelength or band of wavelengths to excite the particles to fluoresce, and detecting fluorescence spectra from the particles with the detector array to provide a pair of spectra for each particle on an image captured by the array detector.
Brief Description of the Drawings
Preferred embodiments of the present invention will now be more particularly described, by way of example, with reference to the accompanying drawings, wherein:
Figure 1 is a schematic perspective view of the preferred embodiment of the fluorescence detection system of the second aspect of the invention;
Figure 2 is a diagram of an example image generated by the detector array of the fluorescence detection system showing the spatially separated spectra received from a plurality of fluorescing air-borne particles;
Figure 3a is a diagrammatic view of a pair of the spectra each spanning eight pixels;
Figure 3b is a graphical representation of the pair of spectra; and
Figure 4 is a schematic perspective view of a first developed embodiment of the airborne biological particle detection system of the first aspect of the invention;
Description of the Preferred Embodiments
In the first developed embodiment of the first aspect of the invention, shown in Figure 4, air containing particles is drawn from the ambient environment through a tube 21 by an electrically driven fan 22. Typically the airflow through the tube 1 would be 100ml per second.
The use of a fan is important as it requires far less electrical power than does an electro-mechanical pump to move an equivalent volume of air, providing no substantial impedance to the flow is presented. Most of the single particle analysis methods described as prior art have to employ pumps because they require clean particle-free air into which to inject the flow of the incoming sample airborne particles. This clean air has to be derived by drawing ambient air through fine filters. Such filters offer a high impedance to the airflow and therefore require pumps rather than fans to generate the necessary airflow.
To the side of the tube 21 is an illumination source 23 comprising a xenon flash tube 24, a collimating lens 25, and an optical filter 26. The flash tube would be of a similar type to those widely used in contemporary disposable cameras. These tubes are very low cost but are capable of very high optical output powers, typically hundreds of milliJoules per flash. The spectral output of the emitted light extends from the ultraviolet (~100nm wavelength) to the infrared (~1 um wavelength).
Approximately 30% of the emitted optical power is in the band 200-350nm wavelength which covers the radiation band required to excite fluorophores found in most biological particles. The optical filter 26 is therefore selected to allow only radiation of wavelength less than 350nm to pass through to the particle-laden airflow carried in the tube 21.
Typically the electronics 27 used to drive the flash-tube will be configured to generate flashes at approximately 0.5 second intervals. (This is non-critical, but the flash period should be such that the air volume illuminated by a flash has been completely replaced in the tube 1 prior to the subsequent flash. Also, the slower the flash rate, the lower the power consumption of the device. The cross-section of the collimated beam of ultra-violet radiation and the airflow containing the particles is referred to as the scattering volume. Typically several particles or even tends of particles in the size range of interest (typically 1-10 urn) may be present in the scattering volume at the instant of a flash.
Lying orthogonal to the axis of the illumination system and to the axis of the tube 1 is a detection assemble comprising an assembly 28 for the detection of light elastically scattered by particles in the scattering volume and an assembly 29 for the detection of fluorescent light emitted by particles' in the scattering volume. The elastically scattered light detector 28 comprises a focussing lens 30 and a low-cost silicon carbide detector or similar capable of detecting the ultra-violet light scattered from the illuminated particles. The magnitude of the signal from the detector will be proportional to the number and mean size of the particles illuminated within the scattering volume, and this relationship may be determined either empirically using known test particles or from conventional established light scattering theory.
The fluorescence detector 29 comprises an optical filter 32, a collimating lens 33, and a sensitive optical detector such as a small photomultiplier tube 34. The optical filter 32 is selected so as to block the elastically scattered ultra-violet radiation from the illuminated particles and only allow to pass the fluorescence light at longer wavelengths of typically 400-700nm. The collimating lens 33 then focuses this light onto the photomultiplier tube. Again, the magnitude of the fluorescence signal will be dependent both on the number and sizes of illuminated particles and on the intrinsic fluorescent properties of those particles.
The electrical signals from the silicon carbide detector and the photomultiplier tube are captured, recorded, and processed using suitable electronic circuitry 35. The nature of the processing is described below. For protection, the monitor system (together with re-chargeable battery supply) would be housed in a robust case. The size of the case would be typically 12cm by 8cm by 8cm, with a weight of typically 800g. This is far smaller and of lower weight than any of the prior art bioaerosol detection systems referred to earlier. In addition, its cost would be typically between 2% and 0.5% of the cost of any of the prior art systems.
Data Processing
Each time the xenon tube flashes, the electronic system 35 records a value of elastic scatter and a value of fluorescence. The ratio of the fluorescent to elastic scatter signals will be characteristic of the nature of the particle material contained within the scattering volume at the time of the flash and will, to a first order, be independent of the number and size of those particles. Thus, whilst the elastic scatter signal yields information on the concentration and sizes of aerosol particles, the ratio signal is dependent only on the particle material. The data processing electronics 35 is configured in such a way as to determine, from the combination of the ratio and elastic scatter data, whether the particles within the scattering volume are likely to be (or contain a significant proportion of) biological particles. This decision would typically be cased on the measurements from a series of flashes such that the response time of 10 seconds maximum was achievable with an acceptable degree of confidence. A positive detection result would elicit an alarm signal of some kind so as to give time to persons present to leave the area or to take protective measures to avoid exposure/inhalation of the biological threat posed. Subsequent to the alarm, it would be envisaged that samples of the airborne
particles in the environment would be captured for conventional bio-assay analysis to confirm the identity of the particles. Such test may take several hours to several days to complete.
The monitor as described above would be intended for continuous operation for periods of typically 48 to 72 hours, after which re-charging or replacement of the battery pack would be required. Further additions to the monitor could include a means of counting and sizing a subset of single particles flowing though the tube 1 so as to provide confirmatory data relating to the sizes and concentration of particles present, or an additional detector assembly which could establish the shapes of some of the individual particles within the airflow via analysis of the spatial distribution of the light scattered by the particle (as per the prior art described earlier). Such additions would improve the particle discriminating capabilities of the monitor but are not essential and would be implemented at the expense of greater instrument complexity and hence manufacturing cost.
Referring now to Figures 1 to 3, a preferred embodiment of the second aspect of the present invention, the improved fluorescence detection system, will now be discussed. Here, airborne particles 1 are drawn into the instrument from the ambient environment by a pump or similar device. The particle flow is arranged to be in the form of a thin sheet (the reason for this is explained later) and this may be achieved by sandwiching the flow between two transparent plates (not shown) or by emitting the flow through a narrow slit orifice. To either side of the particle flow are emitters of electromagnetic radiation, 2 and 3. In the example shown, only two such emitters are indicated, although conceivably a single emitter or more than two emitters could be employed.
The emitters are capable of illuminating the particle flow with pulses of electromagnetic radiation of a suitable wavelength to excite fluorophores within the particles. In the preferred embodiment, the emitters are xenon flash lamps with suitable collimating optics (for example, model RSL3100 produced by Perkin-Elmer Inc., Santa Clara, California, USA). A pulse of radiation from emitter 2 passes through an optical band-pass filter such that the radiation falling on the particle flow is suitable to excite a particular set of biological fluorophores. For example, if the pass band was 200-250nm wavelength, then tryptophan and chlorophyll would be
excited to fluoresce if present in any of the illuminated particles. The total number of particles illuminated by the pulse of radiation may vary from zero to several tens, depending on the concentration of particles in the sampled air and the geometry of the illuminated airflow.
For convenience, consider just two of these illuminated particles, labelled 5 and 6 in Figure 1. Some of the elastically scattered light and the inelastically scattered light (fluorescence) from these two particles will be captured by the lens assembly 7 which collimates the light from the particles and directs it to a high-pass optical filter 8.
For collimation of this light by lens 7 to be effectively achieved, all the illuminated particles should suitably lie in or close to the focal plane of the lens. This is why the particle flow should be in the form of a thin sheet co-planar with the focal plane of the lens 7. An alternative would be to restrict the illumination from the flash lamp 2 to a thin sheet and allow the particle flow to extend beyond this sheet on either side. However, this has the disadvantage of allowing the possibility of some particles being illuminated at the extremities of the light sheet, thus experiencing a lower irradiance, producing an artificially low level of scattered light and fluorescence, and resulting in an erroneous assessment of the volume of fluorophores present within the particle.
A filter 8 prevents the elastically scattered 200-250nm radiation passing through but allows the higher wavelength fluorescence radiation from the particle to pass.
Collimated fluorescent light from particle 5 passing through the optical filter 8 falls in turn onto a dispersing optical element 9 such as a diffraction grating, prism, or similar device. In this way the spectral components within the light are spread in the horizontal plane by an amount dependent on their wavelengths. The light is subsequently focused by a second lens assembly 10 onto an optically sensitive detector array 11. If the dispersing element 9 was not present, the light scattered by a single particle, say particle 5, would be focused to a spot on the detector array 11. The presence of the dispersing element 9 causes instead a linear spectrum to be incident onto the detector array. The magnification of the optical system 10 and the size of the individual pixels on the array 11 will determine the number of pixels covered by the spectrum. Typically this would be arranged to be about eight or ten
pixels, sufficient to show the major features of the fluorescence spectrum of the particle. (The detector array itself would be typically 1024 by 1024 pixels in size).
In a similar way, fluorescent light from particle 6 would be imaged as a spectrum on the detector array, but because particle 6 is physically displaced from particle 5, so the spectrum from particle 6 will be displaced on the array from the spectrum of particle 5, as indicated in Figure 1. Because the magnitude of fluorescence from an individual particle will be small, even when the particle is illuminated with intense ultra-violet radiation from a xenon flash lamp, the detector array 11 must be extremely sensitive. A preferred embodiment would be a second-generation intensified charge-coupled device camera such as those manufactured by Photek Ltd, St Leonards-on-Sea, U.K.).
The flash duration from a xenon lamp such as envisaged for this application would have a duration of typically 10 microseconds. If the particles in the sample airflow are all moving with a typical downward velocity of 0.1 m/s, the particles will have moved typically 1 micrometre during the period of illumination. This is small compared to the typical size of the pixels on the detector array (~10 micrometres square), so that even allowing for some magnification by lens assembly 10, the spectrum recorded by the array will be essentially horizontal (a function of the orientation of the dispersing element).
Following the pulsing of xenon source 2 and the recording of the particle fluorescence spectra on the detector, a second xenon 3 is pulsed. In the preferred embodiment, this emitter would also be a collimated xenon source, but with its output filtered by band-pass filter 12 to allow a different wavelength band of radiation to fall on the particles. Typically this would be 350-400nm wavelength, sufficient, for example, to excite chlorophyll fluorophores but not those of tryptophan. The fluorescence spectrum of each illuminated particle would be recorded on the detector array in the same way as when the particles were illuminated by xenon source 2. However, because the particles will have moved downwards during the interval between the two flashes, the two spectra from each particle will be vertically separated on the detector array as in figure 2. This figure illustrates the capture of ten pairs of spectra arising from ten particles present within the measurement space at the time of illumination by the two xenon sources. The camera shutter exposing
the detector array is arranged to open prior to the flashing of xenon source 2 and to close following the flashing of xenon 3.
Once the camera shutter has closed, the image recorded by the detector array is transferred to a computer for processing. The camera is then ready to repeat the cycle of spectra capture. Typically this cycle would be repeated every 20ms, allowing the spectra from typically 500 particles to be captured each second, assuming an average of ten illuminated particles per captured image. This rate of data capture is commensurate with that required for fast-response detection of biological particles in an ambient environment.
The computer is programmed to scan the recorded image and identify the locations of pairs of spectra. Figure 3a illustrates such a pair of spectra covering 8 pixels horizontally. Typically the spectra will cover the wavelength range 400-700nm. The intensity of light falling on each pixel is represented by an electrical charge, and this is converted to a digital pattern of relative intensities as indicated in Figure 3b. Analyses of these spectra against pre-recorded spectra from known samples in the laboratory will allow matching of the particle type to one of these known samples. Thus the invention will allow the continuous monitoring of an ambient environment for particles of known interest.