WO2014132077A1

WO2014132077A1 - Apparatus and method of breath volatile organic compound analysis and calibration method

Info

Publication number: WO2014132077A1
Application number: PCT/GB2014/050600
Authority: WO
Inventors: Grant Andrew Dedman Ritchie; Graham Hancock; Robert Peverall
Original assignee: Isis Innovation Limited
Priority date: 2013-03-01
Filing date: 2014-02-28
Publication date: 2014-09-04
Also published as: GB201303687D0

Abstract

An apparatus and method for analysing breath, in particular volatile organic compounds such as acetone in breath and for calibrating breath analysis measurements using a breath- derived background sample. Breath is sampled directly through a mouth piece where flow rate and carbon dioxide are monitored, and is in part drawn through a cold trap into a spectroscopic measurement cell, such as for carrying out cavity-enhanced absorption spectroscopy, and also in part drawn through a molecular sieve trap into a holding cell or bellows of flexible volume. Measurements are then made on the sample in this spectroscopic cell and subsequently on the background sample derived from the holding cell. The cold trap is designed to reduce the water content of the breath sample to a predetermined level by reducing the temperature of the breath to a predetermined level. Knowing the quantity of water allows the absorption cross section of water also to be known, providing a direct calibration of the response of the instrument. The molecular sieve is selected to remove water and the analyte of interest, this resulting in a breath- derived background sample which includes exhaled methane and carbon dioxide and thus provides an appropriate background level for the individual subject against which the absorption measurement on the dried breath sample can be compared to quantify the analyte of interest.

Description

APPARATUS AND METHOD OF BREATH VOLATILE ORGANIC COMPOUND

ANALYSIS AND CALIBRATION METHOD

The present invention relates to an apparatus for measuring, i.e. detecting and quantifying, volatile organic compounds (VOCs) in breath, to a method of detecting and quantifying breath VOCs using such an apparatus, and to a calibration method for use in breath analysis. In particular, it can allow the detection and quantification of ketones such as acetone in breath and uses a technique that eliminates the effects of variation in exhaled amounts of non-analyte such as carbon dioxide and methane.

The analysis of a human or animal subject's breath has been used and proposed for many years as a way of measuring various aspects of the subject's condition. It can be used for disease detection or for measuring the subject's performance and condition (e.g by analysing breath exhaled during exercise). For example, it has long been suggested that the level of acetone in exhaled breath, which is a marker of blood ketones, can be used as a possible marker for changing blood glucose levels in type I diabetics. Breath acetone levels are also sensitive to diet and exercise, and thus monitoring them can assist with assessment of diet and exercise regimes. Type I diabetes sufferers must continually measure their blood glucose levels with checks several times a day. It is also

recommended that diabetics who are feeling ill, or those at diabetes onset, also measure their blood ketones in order to prevent diabetic ketoacidosis (DKA) - this is especially relevant for juvenile sufferers. Currently, the most common way of measuring blood glucose levels involves finger lancing and blood testing, and ketones can be measured both by blood and urine testing. However, a non-invasive method for monitoring blood glucose levels and more convenient ways of testing for blood ketones would be extremely useful. Although measurement of breath acetone appears to offer that possibility, current methods of measuring breath acetone rely on mass spectrometry, optical techniques or fuel cell methods, all of which have individual practical difficulties. For example, although mass spectrometric techniques are accurate, they require the use of large and expensive mass spectrometers, and are thus unsuitable for widespread use. Lower-cost techniques of measuring breath acetone have been proposed based on absorption spectroscopy, but these have been too bulky to be realised in a convenient device. They can also suffer from selectivity problems. For example, the article "A New Acetone Detection Device Using Cavity Ringdown Spectroscopy at 266 nm: Evaluation of the Instrument Performance Using Acetone Sample Solutions" by C Wang and A Mbi (Measurement Science and Technology, 17 July 2007), examines the possibility of using cavity ringdown spectroscopy to measure acetone, but did not produce a commercially-usable device and did not operate on breath (instead testing using samples of acetone in deionised water). A later paper measured breath samples indirectly from bags (Wang et al. IEEE SENSORS

JOURNAL Volume: 10 Issue: 1 Pages: 54-63 DOI: 10.1109/JSEN2009.2035730 Published: JAN 2010). More compact methodologies, such as chemical conversion followed by fluorescence spectroscopy, chemical conversion followed by multipass absorption spectroscopy, fuel cell methods or fibre-based spectroscopy suffer from calibration problems or lack of sensitivity. Thus, although the need for an effective breath VOC analyser has been recognised, none of the currently proposed techniques have delivered one.

Although absorption spectroscopy is a promising technique for breath analysis, the nature of exhaled breath makes it difficult to analyse. Breath contains large quantities of water vapour, and possibly also water in aerosol form, and variable quantities of carbon dioxide and methane and all of these can interfere with the signal from the analyte of interest. For example absorption by water vapour consists of both a structured discrete spectral line component, and a continuum component. While careful choice of the light source wavelength can mitigate the effects of the discrete spectral line component, the continuum component cannot be avoided in this way, which is a particular difficulty for measuring VOCs in breath which can exhibit broad spectral absorption features. It is particularly problematic where breath sample pressures approach atmospheric pressure in the measurement cell, and at infrared wavelengths where all VOCs absorb light. Aerosol formation also results in a continuum absorption, which is problematic across all wavelength regimes, but especially within the ultraviolet.

The quantity of carbon dioxide and methane in exhaled breath also varies significantly between individuals, and indeed can vary with time even for one individual. Failure to take into account the quantity of carbon dioxide and methane can lead to erroneous results for measuring breath VOCs, particularly at certain detection wavelengths (e.g. the near-infrared, or 1.6 - 1.8 microns, which is suitable for acetone). There is therefore a need for improved breath VOC analysers and methods and also for improved calibration techniques for use in breath analysis.

Accordingly a first aspect of the present invention provides an apparatus for measuring the quantity of at least one volatile organic compound (VOC) in breath, comprising: sample handling means for receiving a sample of breath and adapted to process it: a) to remove water therefrom to produce a dried sample of breath having a predetermined water content; and b) to remove water and the at least one VOC therefrom to produce a breath-derived background sample; a spectroscopy cell, light source and detector for performing spectroscopy on a sample in the spectroscopy cell; a control and delivery mechanism adapted to deliver the dried sample of breath to the spectroscopy cell and to control the apparatus to make a spectroscopic measurement thereon, and to deliver the breath-derived background sample to the spectroscopy cell and control the apparatus to make a spectroscopic measurement thereon; and a data processor for obtaining from the two spectroscopy measurements the VOC content of the sample of breath.

Preferably absorption spectroscopy is used for the spectroscopic measurement. In a preferred embodiment the spectroscopy cell is an optical cavity for performing cavity- enhanced absorption spectroscopy on a sample in the optical cavity.

The use of a dried sample of breath, by which is meant a sample of breath with a known (lowered) water content, is that if the water content is known, its spectroscopic response, e.g. absorption coefficient, is also known. In particular, a spectroscopic transition of water with a known absorption cross-section, can be measured to determine the response of the analysis device. The known residual amount of water in the dried sample of breath therefore provides for absolute calibration of the device.

Alternatively if the absorption measurement uses a light source in a spectral region where there are no discreet water absorption features, then light scattering such as Rayleigh scattering can be used as a calibration.

Preferably the water content of at least a portion of the sample of breath is reduced to a predetermined level by use of a drier, such as a cooler or chiller which cools the portion of the sample of breath to a predetermined temperature. The Antoine equation relates the vapour pressure of water to temperature, so with a known temperature a known amount of water is present. In one embodiment of the invention two chillers are provided for alternate use, allowing one to be cooling a sample while the other is, for example, being purged.

The drier may include a heater for heating the drier to remove condensation after the dried sample of breath has exited the drier and the drier may also include a temperature sensor to allow monitoring of the sample temperature.

Preferably the sample handling means also include a filter/drier for removing VOCs and water from at least a portion of the sample of breath to produce the breath- derived background sample. The filter/drier may comprise a molecular sieve which can remove water and all or most of the VOCs, including the one (or more) to be measured. The resulting breath-derived background sample provides a baseline for the individual as it contains the usual air constituents (oxygen, nitrogen, argon etc.) and the carbon dioxide and methane that had been exhaled. The comparison of the absorption measurement on the dried sample of breath with an absorption measurement on the breath-derived background sample therefore gives a more accurate quantification of the amount of VOC in the breath. The filter/drier may receive a breath sample directly or, more preferably it receives part of the dried breath sample.

The molecular sieve can be a type 3 A, type 5A or calcium sulphate molecular sieve. The molecular sieve can be selected to include a particular VOC in the background measurement if this is required. This would allow the quantification of a particular selected VOC without the interfering effect of other VOCs (which are included in the breath-derived background sample).

The light source is preferably an infrared light source and the apparatus can be adapted to measure acetone as the VOC with the absorption measurements being conducted in the acetone continuum wavelength region.

The invention also provides a corresponding method of measuring the amount of at least one volatile organic compound (VOC) in breath, comprising: receiving a sample of breath and processing it: a) to remove water therefrom to produce a dried sample of breath having a predetermined water content; and b) to remove water and the at least one VOC therefrom to produce a breath-derived background sample; delivering the dried sample of breath to a spectroscopy cell and making a spectroscopic measurement thereon; delivering the breath-derived background sample to a spectroscopy cell and making a spectroscopic measurement thereon; and obtaining from the two spectroscopic measurements the VOC content of the sample of breath.

The above aspects of the invention are directed to the measurement of VOCs in breath. However the idea of producing a background sample from the breath sample itself is applicable in other breath analysis techniques. Thus another aspect of the invention provides a calibration method for use in an analysis of breath comprising: receiving a sample of breath and processing it: a) to remove water therefrom to produce a dried sample of breath having a predetermined water content; and b) to remove water and analyte therefrom to produce a breath-derived background sample; and analysing both the dried sample of breath and the breath-derived background sample.

The removal of water and the removal of water and analyte can be achieved by corresponding techniques to those discussed in the first aspects of the invention.

The invention will be further described by way of examples with reference to the accompanying drawings in which: -

Figure 1 is a schematic block diagram of an embodiment of the invention;

Figure 2 is a schematic timing and measurement sequence diagram; and

Figure 3 illustrates the results of comparing measurements in accordance with an embodiment of the invention to measurements made by a mass spectrometer.

Figure 1 illustrates an apparatus for analysing VOCs in breath in accordance with an embodiment of the invention. A mouthpiece 1 including an antibacterial filter is provided into which a subject can exhale into the apparatus. The flow rate and carbon dioxide content are monitored by a flow sensor 3, e.g. of differential pressure type, and a carbon dioxide monitor 5, such as a non-dispersive infrared absorption sensor and the sample of breath, after processing by the sample handling means 7 is transported for analysis to a cavity-enhanced absorption spectroscopy arrangement 9 which includes an optical cavity 10, light source 11, detector 13 and controller and data processor 15.

The controller and data processor 15 preferably also receives signals from the flow sensor 3, carbon dioxide monitor 5 and pressure transducers 20 in the sample handling means 7 and controls the pump 24, valves and other components of the sample handling means 7. The sample handling means 7 includes two chillers 26 and 28 which are provided for alternate use and which receive a portion of the breath sample. The chillers are adapted to cool the breath samples to a predetermined desired temperature to reduce the water vapour pressure to a preset level. In this embodiment, for spectroscopic measurements in the 1.7 micron wavelength region, a temperature of -20 °C is suitable. The chillers 26 and 28 are, for example, in the form of hollow metallic cuboids with entrance and exit channels for the breath sample. This provides a large internal surface area which maximises water condensation and thus the speed of cooling. The temperature of the chillers is preferably controlled with a Peltier device with a fan assembly and a temperature sensor linked to a controller such as a proportional-integral-differential (PID) controller. The Peltier device can also be used as a heater to remove residual moisture from the chillers 26, 28 between measurements in conjunction with flushing or pumping, or both, with the chiller temperature raised above ambient temperature. Dried breath samples from the two chillers can be directed to the optical cavity 10 via gas conduits 27 and 29 and via a particle filter 30.

To produce the breath-derived background sample, a portion of the dried breath sample from the chillers is passed into a holding cell 32 via a filter/drier 34. The holding cell 32 is preferably an expandable volume such as an expansion bellows of the edge- welded variety so that higher quantities of sample can be measured, enabling higher sensitivity. The filter/drier 34 is a desiccant, preferably a molecular sieve, which is selected to remove both water and VOCs from the breath. The molecular sieve can be of the type 3 A or type 5A, or other chemical substances such as calcium sulphate can be used. The filter/drier must remove the analyte of interest, but it can be designed not to remove other substances or VOCs, thus making them part of the background and eliminating their effect from the results.

The breath-derived background sample resulting from passing the breath through the filter/drier 34 can be directed to the optical cavity 10 via the particle filter 30.

Movement of the samples through the sample handling means 7 is achieved by use of a pump 24 with suitable valves and monitoring is provided by pressure transducers 20 and 22.

In this embodiment the spectroscopy arrangement 9 is a cavity-enhanced absorption spectrometer using an optical cavity which gives high sensitivity. The optical source 11 is a laser or light emitting diode and the preferred detector 13 a photodiode, with the wavelength chosen to optimise absorption for a particular VOC, e.g. acetone. Optical cavities which provide cavity enhancement of the signal differ from single pass or multipass optical cells in that the mirrors which define the cavity are arranged so that the light in the cavity retraces its path rather than tracing a single or single zigzag path through the cell.

Figure 2 illustrates the timing and measurement sequence of this embodiment.

Firstly the apparatus, including the chillers 26, 28 and holding cell 32, is evacuated of most of the gas within it. At least one of the chillers, chiller 26 in Figure 2, is set at -20 °C. In Figure 2 the second chiller 28 is illustrated as being at a higher temperature implying that it is being purged from a previous measurement cycle.

The subject exhales through the mouthpiece 1 and the signals from the flow sensor and carbon dioxide sensor 3 and 5 are used to select a certain portion of the exhaled breath stream which is directed into the cool chiller (26 in Figure 2). A portion of the breath sample is directed from the chiller into the holding cell 32 via a filter/drier 34. If a bellows type holding cell 32 is used, this will inflate as it receives the sample of breath from the chiller 26.

A sample of breath from the chiller 26 (or chiller 28 in an alternate measurement cycle) is passed via an exit line 27 through a particle filter into the optical cavity 10 and a spectroscopic measurement is made on the dried breath sample, in this embodiment at a wavelength between 1.6 and 1.8 microns. Preferably the pressure of the dried breath sample in the optical cavity 10 is of the order of several hundred millibars (more preferably greater than 600 millibars). This measurement results in an absorption for breath VOC (e.g. acetone), carbon dioxide and methane combined).

The optical cavity, chiller 26 and connecting lines are then purged and the optical cavity 10 is filled from the holding cell 32, with the gas from holding cell 32 passing again through the filter/drier 34 and entering the optical cavity via the particle filter 30. During this time the chiller 26 can be heated for its purge process and the chiller 28 cooled ready to receive the next breath sample in the next measurement cycle.

An absorption measurement is then made in the optical cavity on the dry VOC-free background sample from the holding cell 32. This gives an absorption signal which includes the exhaled methane and carbon dioxide, but not the VOC analyte. Thus this signal can then be used in the data processor 15 in the usual equations for absorption. In the case of cavity enhanced absorption spectroscopy the background signal therefore constitutes loin the following equation: - (I₀-I)/I = oNL/(l-R)

where I is the absorption measurement on the sample from the chiller 26, L is the length of the cavity, R is the reflectivity of the mirrors, σ is the absorption cross-section of the analyte and the object is to determine N, the absolute number density (or concentration) of the VOC (e.g. acetone) in the sample.

The dried breath sample can also be used to calibrate the instrument using water spectroscopy as its water content is reduced to a predetermined amount by its pass through the chiller 26 or 28. The absorption cross-section as a function of water vapour content is known from standard spectroscopy references and thus measuring the absorption of the dried breath sample provides an absolute calibration allowing the output of absolute VOC levels.

The idea of producing a breath-derived background sample by producing, on the one hand, a dried sample of breath and on the other hand a dried analyte-free background sample, using two chambers such as chiller 26, 28 and holding cell 32 can be applied to breath analysis techniques other than the CEAS technique illustrated with reference to Figures 1 and 2. Because the background sample is derived from the same sample of breath as the measurement, it allows a more accurate calculation of the amount of the analyte because the background sample includes the species other than the analyte.

Figure 3 illustrates a comparison of measurements of breath acetone as measured by a CEAS embodiment of the invention as illustrated in Figure 1 and as measured by a mass spectrometer. In this case to provide for accuracy in both measurements breath was collected in bags for delivery to both instruments and 42 measurements on breath were made together with four calibration points using an artificial gas mixture (these being the four highest points in the plot). It can be seen that the agreement between the instrument of the invention and the mass spectrometer is good, the standard deviation being 140 ppb.

Claims

1. Apparatus for measuring the quantity of at least one volatile organic compound (VOC) in breath, comprising:

sample handling means for receiving a sample of breath and adapted to process it: a) to remove water therefrom to produce a dried sample of breath having a predetermined water content; and b) to remove water and the at least one VOC therefrom to produce a breath-derived background sample;

a spectroscopy cell, light source and detector for performing spectroscopy on a sample in the spectroscopy cell;

a control and delivery mechanism adapted to deliver the dried sample of breath to the spectroscopy cell and to control the apparatus to make a spectroscopic measurement thereon, and to deliver the breath-derived background sample to the spectroscopy cell and control the apparatus to make a spectroscopic measurement thereon; and

a data processor for obtaining from the two spectroscopy measurements the VOC content of the sample of breath.

2. Apparatus according to claim 1 wherein the sample handling means comprises a drier for reducing the water content of at least a portion of the sample of breath to a predetermined level.

3. Apparatus according to claim 2 wherein the drier comprise a chiller for cooling at least a portion of the sample of breath to a predetermined temperature thereby to reduce the water content to a level dependent upon the predetermined temperature.

4. Apparatus according to claim 3 wherein the drier comprises first and second chillers for alternate use.

5. Apparatus according to claim 2, 3 or 4 wherein the drier further includes a heater for heating the drier to remove condensation after the dried sample of breath has exited the drier.

6. Apparatus according to any one of the preceding claims wherein the sample handling means is adapted to produce the breath-derived background sample from a portion of the dried sample of breath.

7. Apparatus according to any one of the preceding claims wherein the sample handling means further includes a filter/drier for removing the at least one VOC and water to produce the breath-derived background sample.

8. Apparatus according to claim 7 wherein the filter/drier comprises a molecular sieve which removes both water and the at least one VOC.

9. Apparatus according to claim 8 wherein the molecular sieve is selected from a type 3A, type 5A, zeolite or calcium sulphate molecular sieve.

10. Apparatus according to any one of the preceding claims wherein the spectroscopy cell is an optical cavity for performing cavity-enhanced absorption spectroscopy on a sample in the optical cavity.

11. Apparatus according to any one of the preceding claims wherein the light source is an infra red light source.

12. Apparatus according to any one of the preceding claims adapted to measure acetone as the VOC.

13. Apparatus according to claim 12 wherein the light source and detector are adapted to perform the absorption measurement in the acetone absorption wavelength region.

14. A method of measuring the amount of at least one volatile organic compound (VOC) in breath, comprising:

receiving a sample of breath and processing it: a) to remove water therefrom to produce a dried sample of breath having a predetermined water content; and b) to remove water and the at least one VOC therefrom to produce a breath-derived background sample; delivering the dried sample of breath to a spectroscopy cell and making a spectroscopic measurement thereon;

delivering the breath-derived background sample to a spectroscopy cell and making a spectroscopic measurement thereon; and

obtaining from the two spectroscopic measurements the VOC content of the sample of breath.

15. A calibration method for use in an analysis of breath comprising;

receiving a sample of breath and processing it: a) to remove water therefrom to produce a dried sample of breath having a predetermined water content; and b) to remove water and an analyte therefrom to produce a breath-derived background sample; and

analysing both the dried sample of breath and the breath-derived background sample.

16. A calibration method according to claim 15 wherein the step of removing water to produce a dried sample of breath having a predetermined water content comprises cooling at least a portion of the sample of breath to predetermined temperature thereby to reduce the water content to a level dependent upon the predetermined temperature.

17. A calibration method according to claim 15 or 16 wherein the step of removing water and an analyte therefrom to produce a breath-derived background sample comprises removing an analyte from a portion of the dried sample of breath.