GB2486495A - Flame technique for the analysis of samples using infra-red absorption spectroscopy - Google Patents

Abstract

Detecting chemical species of interest using a FLAIR (flame infra-red) system 100. An infra-red optical beam 109 which may be produced by a Fourier Transform Infra-red Spectrophotometer (FTIR Spectrophotometer) 120 is passed through a flame 101 produced on a burner head 103. A chemical sample is introduced into the flame 101 via aspirator 107. The absorption spectrum of the infra-red radiation is measured after the radiation has passed through the flame 101, wherein the spectrum is indicative of certain species. The flame 101 may be a cool flame. The spectrum may be measured by measuring the intensity of the infra-red radiation at a specific polarization angle. The aspirator 107 may introduce the sample into a mixing chamber 105 which contains three ports positioned for a fuel and oxidant provided by supply member 113, as well as a liquid drain. A computer 130 may be operable for data capture and analysis.

Description

I

New Flame Technique for the Analysis of Samples by their Molecular Absorptions The present invention relates to an analytical instrument and technology for the analyses of liquids, gases, and solids which can be made gaseous by detecting their infra-red absorption spectra resulting from the molecular species which form following their introduction to a flame. This technique is referred to as FLAIR, an acronym for "flame infra-red".

Flame spectroscopy is amongst the oldest of all analytical techniques preceding the well know early studies using the blowpipe by the Swedish scientist in the 17th and 18th Centuries and the German scientists in the 19th Century. The flame photometer developed by H. Lundegardh in 1928 was the first to have all the components of the modern instrument, namely a flame, a means of nebulizing a liquid sample for reproducible and consistant introduction of a sample, and way of selecting the portion of the light spectrum of interest, and a detector I transducer which gave a electric signal proportional to the intensity of the emission in the flame.

Following this landmark development came the first major commercialization of an analytical instrument, with over 50,000 photometers being sold by Evans Electroselenium Ltd.. These early photometers and many which are currently manufactured today, are most suitable for the measurement of elements which are easy to form atoms in the flame and excite, such as potassium, sodium, lithium, rubidium, etc. They use relatively cool hydrocarbon flames, mainly composed of methane, butane and propane with air, and having temperatures around 1850 to 1950 degrees Centigrade. In an effort to measure elements which form more refractory molecules in the flame by ensuring they break down to their atomic form hotter flames were employed, such as Hydrogen I Oxygen, and Acetylene I Nitrous Oxide, having temperatures around 2800 deg C. The range of elements that could be analysed and the number of useful applications grew with the use of hotter flames and with the introduction of Atomic Absorption Spectroscopy from the mid 1950's onward. The introduction of plasmas in the 1970's, having effective temperatures of 10000 deg C and higher, further enhanced the usefulness of this technique.

The overwhelming success of flame photometry in all its forms is undoubtedly due to the fact that there is not a lot of sample preparation involved, the analysis is quick and easy to achieve, it is sensitive and relatively inexpensive.

By its nature, flame spectroscopy is primarily applicable to inorganic cations that are broken down to their atomic form in the flame. The atomic emissions and absorptions for these analytes occur in the visible and UV portion of the electromagnetic spectrum. The term flame spectroscopy is synonymous with the term atomic spectroscopy.

There are exceptions to this since some refractory molecules do give relatively broad banded emissions usually from their oxides and hydroxides which can be used quantitatively. These molecular emissions are more likely to occur with the cooler flames rather than the hottest flames and plasmas and normally are avoided where possible by the analyst who understandably prefers the advantages obtained from highly resolved atomic resonance lines.

Elements having their atomic resonance lines in the vacuum ultraviolet region are not suitable for flame analysis. There are a number of exceptions where a cool flame, such as hydrogen diluted by nitrogen using diffused air (less than 1000 deg C) has been used for the purpose of promoting molecular emissions which occur in the visible and ultraviolet regions of the spectrum. Employing a cool hydrogen flame, sulphur has been determined by monitoring a S2 emitting species at 384 nm, and phosphorus by an HPO species at 528 nm, and halides by an Indium halide, lnX, emitting species, in the region of 360 to 410 nm, etc. All these emissions require a reducing cool hydrogen flame and the molecules quickly break down and the emissions disappear with the addition of oxygen to the flame. There have been over a 1000 publications on the technique known as Molecular Emission Cavity Analysis which started in 1971, but it appears that the work that continues is primarily of academic interest and no commercial instrument is currently available. The major difficulty is that this technique is dependent on introducing the sample into the cool flame via an aperture at the end of a probe which is inserted into a flame. This probe changes with use which affects the analytical performance. Although it is possible to analyse discretely for inorganic sulphates, suphites, and sulphides in admixture with a range of cations (by adding excess phosphoric acid) it is difficult to analyse a range of organo suphur compounds which present in a mixture without prior separation or pre-treatment.

The most commercially successful examples of the use of molecular emissions are found in its use as a chromatographic detector. The suphur and phosphorus "specific" detectors utilise a hydrogen diffusion flame to generate their blue and green emissions respectively. This is following the separation of the compounds on a chromatographic column.

If we consider molecular emissions outside the visible I UV region, i.e. the infra-red, we find even less applications. Infra-red emissions have been used to measure the body of the flames for C02 and for H20 molecules in hydrocarbon flames and exhaust gases but have not shown a great deal of sensitivity. The number of emitting species in the infra-red region appears to be limited judging by the cited references. It is apparent from these studies on flames that blackbody radiation is a potential major interference in the infra-red region particularly when solid particles are heated to higher temperatures.

In summarizing, the major advantages of flame photometry are not realised for the large number of elements which have their atomic resonance lines in the vacuum ultraviolet region.

The infra-red absorption using the wavelength region from I to 20 microns has been shown to be particularly useful for the qualitative and quantitative analysis of organic compounds and many of the molecular species for which atomic flame spectroscopy does not easily apply. The usefulness of IR Absorption on the measurement on gases is also well established. The absorption bands are related to the stretching and bending vibrations of atoms rather than their electronic transitions and their wavelength is very much determined by the bond strength between atoms.

In the molecular absorption mode of infra-red spectrophotometry, a continuous light source, generated by a Nernst glower or other similar source, produces infra-red light at all wavelengths which are then passed through the sample, here a flame containing a sample to be analysed. The light leaving the sample is then spectrally separated into its component wavelengths and its intensity is measured. In experiments a Fourier Transform Infra-red Spectrophotometer has been used to achieve the light separation. The intensity of light measured gives information on the number of molecules absorbing, i.e. quantitative information while the wavelength of the absorption provides qualitative information on the particular bond responsible for the absorption.

The potential advantages offered by a hybrid analytical instrument consisting of a flame photometer with an infra-red absorption detector are evident. In the ideal, the flame will break down the sample into its molecular constituents and the infra-red spectrophotometer will quantitatively and qualitatively measure their presence. The potential advantage offered to the operator is a quick analysis with minimal sample pre-treatment.

The patent describes the development of FLAIR aimed at producing an instrument which would retain the analytical advantage as flame photometry for a new range of organic and inorganic samples.

In accordance with a first aspect of the present invention, a method for detecting chemical species of interest is provided, comprising; (b) introducing a chemical sample into a flame; (c) directing infra-red radiation through said flame; and (d) measuring the spectrum of the infra-red radiation after said radiation has passed through said flame, wherein the spectrum is indicative of certain species.

Preferably the flame is a cool flame which is important for the analysis of molecules such as HCI, HS and C-C. However, on the addition of oxygen a hotter oxidising flame can be produced which is useful for the analysis of species such as 002, P-0 amongst others The method of the present invention has many advantages over the prior art. For example, the flame selection allows molecular species to remain intact, produces a relatively simple background spectrum on which a sample spectrum could be measured, and provides a reproducible means of introducing a liquid or gas sample independent of altering the flame composition. It is also realised that the composition of a flame in going from a fuel rich reducing zone to an oxygen rich oxidising zone will produce different molecular species, a feature which can be exploited advantageously through careful control of the flame gases throughout measurement.

Further advantages of the present invention include not having to use a solid probe as in Molecular Emission Cavity Analysis; and allowing measurement of species with atomic resonances in the vacuum UV by using their molecular absorptions in the infra-red region.

Preferably, the spectrum is measured by measuring a property of the infra-red radiation. In one embodiment said property is its intensity. In another embodiment the property is its intensity at a specific polarization angle.

Preferably, the sample is introduced into the flame through an aspirator, although other options are envisaged. For example, headspace analysis taking the vapour space above a liquid sample directly, purging a liquid sample with a gas or using an ultrasonic vibrator to create a mist.

In one embodiment the sample is aspirated with an inert gas to control the uptake rate of the fluid sample. Said inert gas is also preferably infra-red transparent.

In a preferred embodiment, the sample and inert gas are mixed with the flame fuel and oxidant. Here the uptake rate of the sample is advantageously independent from the fuel and oxidant.

Advantageously, varying amounts of oxidant are provided to the flame to vary the characteristics of the flame. This provides additional means of obtaining selectivity of the analysis, along with wavelength selection.

In one embodiment the sample is a fluid. In other embodiments the sample is dissolved in a solvent or made gaseous by processes such as pyrolysis.

In one embodiment where the sample is a liquid, the sample is a liquid and is weighed during aspiration and analysis. This allows the calculation of the amount of sample entering the flame and can be used to improve quantitative analysis when dealing with samples that have a variable aspiration rate.

In a preferred embodiment, the method of the present invention further comprises the initial step of; (a) measuring a background spectrum; and subsequent step of (e) subtracting the background spectrum from the spectrum measured in step (d).

Where the sample is solid dissolved in solvent, the solvent is used to measure

the background spectrum.

Preferably, the method prevents the ingress of air into the flame. If air is allowed to enter the flame, the flame forms an oxidising flame very quickly and very close to (approximately 1cm) the base of the burner head. The optical beam produced on an infra-red spectrophotometer is broad and focussed starting from a few centimetres and reaching a minimum of just less than a centimetre at its minimum, the focal point in the middle of its path length. To ensure a broad enough reducing zone to suit the optical beam it is preferable to introduce diffusion barriers to stop the ingress of air.

Other advantages realised by isolating the reducing zone is that much of the noise normally occurring at the edge of the flame is eliminated and it is possible to maintain a temperature under 750 degrees Centigrade.

In a second aspect of the present invention, an apparatus for detecting chemical species of interest is provided, comprising; a burner head; a burner system to provide a flame to said burner head; supply means for introducing a chemical sample into said cool flame; an infra-red radiation source adapted to pass infra-red radiation through said cool flame; and a detector for measuring the spectrum of said infra-red radiation after it has passed through said cool flame when in use.

Preferably, the flame is a cool flame.

Preferably a mixing chamber adapted to provide gas to said burner head is further provided; along with a supply system for the introduction of a fuel and an oxidant to said mixing chamber, where the supply of said fuel and the supply of said oxidant are independently controllable. The oxidant is preferably oxygen although others may be used.

In one embodiment the mixing chamber comprises a port for liquid drainage. In another embodiment the mixing chamber comprises impellor blades for removing large droplets of a liquid sample.

Preferably the supply means is an aspirator adapted to provide the sample to the mixing chamber, and in one embodiment the aspirator is further adapted to provide an inert gas in order to control the flow rate of said sample. Said inert gas is also preferably infra-red transparent.

Advantageously, in one embodiment the apparatus further comprises a reflective optical element adapted to reflect said infra-red radiation so as to increase the path length through said flame. This increases sensitivity of the measurements and lowers the limit of protection.

In one embodiment the burner head comprises a first diffusion barrier mounted on a first side along its length so as to prevent ingress of air to said flame along said length, when in use. Preferably, the burner head comprises a second diffusion barrier mounted along a second side of said burner head so as to face the first diffusion barrier and prevent the ingress of air to the flame from said second side when in use and the burner head further comprises an inert gas boundary layer at at least one end.

Advantageously the diffusion barriers remove the need for the burner to have a diffusion flame. This allows the composition of the flame to be controlled only by the composition gases. For example, the amount of oxygen can be gradually varied to create both a fuel rich reducing environment and an oxidising environment.

The solid and inert gas barriers also create an enlarged zone suitable for the size of the infra-red light source.

In one embodiment the diffusion barrier(s) are cooled. This assists in reducing the flame temperature which can affect the rotational spectra.

In a further embodiment the burner head further comprises a structure to fully enclose the flame, in use.

In another embodiment the burner head is mounted on a movable table.

In yet another embodiment means for cooling said gases before reaching the mixing chamber are provided. Similarly to the cooling of the diffusion barriers, this assists in reducing the temperature of the flame. In one embodiment the cooling is done using cooling coils although other embodiments such as dry ice and electronic cooling are envisaged.

In one embodiment an externally cooled surface is provided to the flame in use.

In a preferred embodiment, the surfaces of the first and second diffusion barriers facing each other are optically reflective. This advantageously allows the effective path length through the flame to be increased by introducing at least one optical element adapted to direct said infra-red radiation at an angle towards said optically reflective surfaces, thus increasing the sensitivity of the measurement.

In one embodiment the apparatus further comprises a filter element which measures a specific portion of the full infra-red spectrum. In one embodiment said filter element also controls the supply of gases to the flame. In designing a suitable filter element that is inexpensive, one must know: -the restricted wavelength range which will measure the molecule of interest.

-the potential analytical interferences which may occur with the sample -the optimal flow rates of all the gases forming the flame and aspirating the sample.

The burner design had to compensate for several problems. The sensitivity of an absorption measurement is directly dependent on the optical path length (Beer's Law), and, therefore, an atomic absorption burner head which 10 centimetres long has been used, although other lengths and configurations are envisaged.

For example the optical elements and reflective surfaces of the absorption barriers described above which allow multiple reflectances of said infra-red radiation through the flame increase the effective path length.

Embodiments of the present invention will now be compared and contrasted with the prior art with reference to the following drawings, in which; Figure 1 is a schematic drawing of an apparatus for detecting species of interest according to one aspect of the present invention; Figure 2 is a schematic drawing of the burner head feature of the apparatus; Figure 3a is a schematic drawing of a first configuration of the burner head and opticalbeam; Figure 3b is a schematic drawing of a second configuration of the burner head and optical beam; Figure 3c is a schematic drawing of a third configuration of the burner head and optical beam; Figure 4 is a flowchart outlining the method steps of in the analysis of an unknown sample; and Figure 5 provides data produced by the new flame technique for analysis of samples according to the present invention.

Considering first Figure 1, a schematic drawing of the FLAIR system 100 is shown, where an infra-red optical beam 109 produced by a Fourier Transform Infra-red Spectrophotometer (FTIR spectrophotometer) 120 is passed through a flame 101 produced on a burner head 103.

A conventional Venturi aspirator 107 is used having an adjustable sample uptake rate which can uniformly and reproducibly deliver a preset liquid sample at a preset rate, from 0.5 to 5 millilitres of solution per minute. A scale and timer (not shown) is provided to measure the flow rate of sample delivered to the flame at the time of measurement. In dealing with organic/aqueous mixtures one can find large changes in aspiration rates between samples which must be accounted for in quantitative analysis.

The Venturi aspirator 107 introduces the sample into a mixing chamber 105, having several stationary impellor blades (not shown) which are designed to remove largest sample droplets of liquid sample from going into the flame. The mixing chamber 105 contains three additional pods positioned appropriately for a fuel and oxidant provided by supply member 113, as well as a liquid drain.

Nitrogen was introduced via the Venturi, the rate of which directly affected the rate of sample uptake. (Nitrogen can be substituted by other inert gases, such as argon, neon, helium, etc.) A computer 130 is operable for data capture and analysis.

The burner 103 is designed to fit the mixing chamber 105 and positioned in the optical beam 109 of an Infra-red Absorption Spectrophotometer 120. The burner 103 has been adapted in the following ways: A 10 centimetre long path length atomic absorption head was used in experiments. However this is not limiting and other path lengths may be used.

A solid air diffusion barrier 114 was attached along the length of both sides of the burner 103 to stop the ingress of air into the flame 101 from the sides. A nitrogen boundary layer was created at the ends of the burner to stop the ingress of air. The nitrogen was supplied by conduits 116.

A means of cooling the incoming gases and the diffusion barrier was provided. In one embodiment the cooling is provided by passing the gases through cooling coils, or alternatively dry ice or electronic cooling may be used. An additional externally cooled surface was provided by a metal container which could be lowered into the burner assembly between the barrier plates. The final modification to this approach would be to totally enclose the flame pulling a vacuum to collect the exhaust gases (not shown).

Some molecular emissions in a cool flame are enhanced when in contact with a cool surface of particular solids. This enhancement is believed to be due to molecular and atomic fragments of substances in the flame sticking momentarily to these surfaces whereby they come in contact with other fragments in the flame, and recombine to form an emitting molecule or free radical. Since this mechanism makes certain molecular collisions and recombinations more likely, there is an enhancement of the molecular emission. Although not visible, this mechanism can generate certain molecular formations which would show a similar enhancement of molecular species which would only show themselves in a infra-red absorption measurements. The cool surfaces on the present instrument are intended to fully exploit the enhancement of measurement by the "Salet Phenomenon" effect.

In one embodiment an optical reflective surface 115 is provided on the inside surface of the barrier walls.

Flow controllers were provided to supply the following gas flow rates ranges: Nitrogen 0.5 to 10 litres per minute Hydrogen 0.1 to 10 litres per minute Oxygen 0.01 to 10 litres per minute The burner head 103 and mixing chamber assembly 106 is positioned on a moveable table (not shown) so that the height of the burner head 103 can be adjusted in the optical path 109. The burner 103 can also be rotated at right angles to the optical beam 109.

Different configurations are possible of optical beam 109 and burner 103, as shown in Figure 3. The first configuration (Figure 3a) positions the burner along the optical path, with the beam preferably lying within 1 cm of the burner head, although this measurement is not limiting.

The second configuration (Figure 3b) aligns the burner head perpendicular to the optical beam, and with the aid of optical reflectors (118) at either end of the flame, a multiple reflectance mode can be obtained for enhancing sensitivity by Beer's Law.

In the third configuration (Figure 3c) the diffusion barriers 114 have an optical reflective surface 115. Here, the burner is aligned at right angles to the optical path, and the beam of incoming light is reflected on the inner surface of the solid diffusion barriers, again with the purpose of increasing the effective path length.

Extra optical elements (118) ensure that the beam is directed to the detector after going through the flame.

Other possible configurations are envisaged.

Observations The use of the nitrogen diluted hydrogen flame provides the simplest background of any flame for the IR measurement and therefore is preferred. (In some cases, other flame gases may be possible for specific samples and analytes.) The use of the inert gas to drive the Venturi aspirator 107 is preferred in that it has the least effect on the conditions of the flame, as compared to the effects of hydrogen and oxygen. This makes sample delivery rate more independent of flame conditions.

Air cannot be used to supply the oxygen since it affect the amount of C02 in the flame. It is possible to use an Oxygen enrichment instrument if sufficient C02 is removed from the air.

Selectivity The selectivity of measurement is affected by the following parameters, in decreasing order: The wavelength of the measurement The composition of flame gases The resolution of measurement The temperature of gases The wavelength is the most critical parameter in the determination of which molecules are being measured. The wavelength is very much determined by the bond strength which affects the stretching and vibrating energies of the bond.

The flame gas composition determines which molecular species are formed as the constituents of the sample break down. In turn, the wavelength will change according to which species are present.

The resolution of measurement is determined mainly by the length of time that the sample is aspirated into the flame and measured. Starting times for in initial experiments were 20 seconds, although it will be appreciated that this is not a limiting time.

Finally, control of the temperature of gases and baffles (not shown) directly affect and shift the rotational peaks present in the spectrum.

Measuring a sample Figure 4 shows a flowchart outlining the method on analysing an unknown solid material. The solid is firstly dissolved in a suitable solvent (step 401). A wide range of solvents can be used for dissolution of many different samples and substances with the only provision that the solvent doesn't interfere spectrally with the absorption wavelength of the substance being measured.

A typical analysis starts by producing a non reactive zone along the measurement path near the burner 103. This is accomplished by turning off the oxygen supply and igniting the gases above the barriers (step 402). Operating with these conditions produces an infra-red spectrum showing absorptions similar to that of the solvent (step 403).

The dissolved sample is then introduced to the flame (step 404). The addition of a small amount of oxygen to the above flame, and measuring along the same optical path near the burner head, the flame now produces a reducing zone which starts to break down the sample constituents and produce new molecular species.

Increasing the oxygen in increments finally produces an oxidising flame. For most hydrocarbons, only C02 and H20 species are shown. The increase in oxygen is done incrementally, and the spectrum is measured and analysed at each increment (step 405). Analysis of the spectra is then carried out (step 406).

Cooling the gases and diffusion barriers can occurs at any of the above stages of having the above measurement at different gas compositions to examine the effect, if any, on resolution and the size of absorption present and types of molecular species in the flame. This can be regarded as a means of fine tuning the analytical response of Flair.

Modern FTIR Infrared Spectrophotometers offer the ability to store spectra and use any stored or freshly acquired spectra as a background to be subtracted from future sample measurements. Solutions containing several components can have each subtracted from a sample compositcomposite spectrum. These are most useful features for analysing FLAIR spectra and removing persistent absorptions which are always present such as those from water and C02. (The CO2 absorption found in hydrocarbon flames, however, is too large to correct effectively by background correction methods and import analyte responses in this region of the spectrum would be lost.) If the sample has been previously assessed by Flair analysis, is known then the flame can be set straight to the desired mode. For example, in analysis of organo halide pesticides, the flame is set to reducing mode; and in analysis of carbonaceous material in soil the flame is set to oxidising mode. In both cases the background spectrum of pure solvent is still required.

It is believed that all organo halides will easily and quickly form H-X under reducing flame conditions, and therefore offer a quick analysis of organo halides in extracts of soil samples taken in estuary and farmland for example. This would be a quick measure of pollutants which could precede more extensive tests such as gas and liquid chromatography.

Sensitivity and Limit of Detection The sensitivity and LOD can be improved mainly by increasing the analysis time which reduces flame noise and by increasing the path length.

Once the flame conditions and instrument configuration is established for a particular application, it will result in a simple single wavelength filter instrument that can offer an inexpensively and practical solution to a difficult application.

Example spectra

FigureS shows spectra demonstrating the method and apparatus of the present invention.

Figure 5a shows the spectra of water (501), acetone (502), ethanol (503) and a mixture of the three (504). These were taken without using FLAIR on the pure samples directly and in admixture. The manner the samples were introduced to the instrument was by using attenuated total reflectance (ATR). The water spectrum was initially takes and used as a reference. The pure was measured as a sample and gave the expected straight line spectrum.

Figure Sb shows the spectra measured from a FLAIR experiment starting from a flame without any oxygen being added (511), ie a non reactive zone. The above mixture was aspirated into the flame to obtain a spectrum (511) which was similar to that found on the original ATR spectrum not using the flame. (Use of the flame does result in a noisier spectrum which could be improved on in the future) The other spectra were obtained by gradually increasing the oxygen addition until none of the organic bonds are absorbing, such as C-OH, C=O, C-H, and only CO2 is shown to be present by its absorption peak at around 2400 WaveNumber (514).

Spectrum 512 shows partial combustion with 1 litre/minute of oxygen being added; spectrum 513 shows more combustion with 2 litres/minute of oxygen being added and spectrum 514 shows total combustion (3 litres/minute of oxygen being added) with only C02 and water absorption as described above.

Figure Sc shows an organic halide. 4-chloro-3,S-dimethylphenol which is a solid dissolved in methanol first by an ATR spectrum taken without using the flame (531), and second using a fuel rich reducing flame on FLAIR (532).

Pure methanol was used as a reference.

The 532 spectrum shows the formation of HCI absorptions at 1300 (533); 3700 (534) and 2700 to 3000 (535) wavenumbers.

Claims (7)

1. CLAIMS: 1. A method for detecting chemical species of interest, comprising; (b) introducing a chemical sample into a flame; (c) directing infra-red radiation through said flame; and (d) measuring the spectrum of the infra-red radiation after said radiation has passed through said flame, wherein the spectrum is indicative of certain species.
2. 2. The method of claim 1, wherein the flame is a cool flame.
3. 3. The method of claim I or claim 2, wherein the spectrum is measured by measuring a property of the infra-red radiation.
4. 4. The method of claim 3, wherein the property of the infra-red radiation is its intensity.
5. 5. The method of claim 3, wherein the property of the infra-red radiation is its intensity at a specific polarization angle
6. 6. The method of any of the preceding claims, wherein the sample is introduced into the flame through an aspirator.
7. 7. The method of claim 6, wherein the aspirator is driven using an inert gas to control the flow rate of the sample.8 The method of claim 7, wherein the inert gas is further infra-red transparent.9. The method of claim 7 or claim 8, wherein the sample and inert gas are mixed with the flame fuel and oxidant.10. The method of any of the preceding claims, wherein varying amounts of oxidant are provided to the flame to vary the characteristics of the flame.11. The method of any of the preceding claims, wherein the sample is a fluid.12. The method of claim 11 when dependent on claim 6, wherein the sample is a liquid and is weighed during aspiration and analysis.13. The method claim 12 wherein the sample is a liquid and large droplets of said fluid are removed from the sample before introduction to the flame.14. The method of any of claims I to 11, wherein the sample is a solid dissolved in a solvent.15. The method of any of claims Ito 11, wherein the sample is a solid made gaseous.16. The method of claim 15, wherein the solid is made gaseous by pyrolysis.17. The method of any of the preceding claims, further comprising the initial step of;(a) measuring a background spectrum;and subsequent step of (e) subtracting the background spectrum from the spectrum measured in step (d).18. The method of claim 17 when dependent on claim 14, wherein the background spectrum is measured when said solvent alone is introduced into the flame.19. The method of any of the preceding claims further comprising inserting a cool surface into the flame.20. The method of any of the preceding claims, wherein the fuel is hydrogen.21. The method of any of the preceding claims, wherein the oxidant is oxygen.22. The method of any of the preceding claims, wherein ingress of air to the flame is prevented.23. An apparatus for detecting chemical species of interest, comprising; aburnerhead; a burner system to provide a flame to said burner head; supply means for introducing a chemical sample into said cool flame; an infra-red radiation source adapted to pass infra-red radiation through said cool flame; and a detector for measuring the spectrum of said infra-red radiation after it has passed through said cool flame when in use.24. The apparatus of claim 23, wherein said flame is a cool flame.25. The apparatus of claim 23 or claim 34, wherein the burner system comprises; a mixing chamber adapted to provide gas to said burner head; and a supply system for the introduction of a fuel and an oxidant to said mixing chamber, where the supply of said fuel and the supply of said oxidant are independently controllable.26. The apparatus of claim 25, wherein the mixing chamber comprises a port for liquid drainage.27. The apparatus of claim 26, wherein the mixing chamber comprises impellor blades for removing large droplets of a liquid sample.28. The apparatus of any of claims 23 to 27, wherein said supply means is an aspirator adapted to provide the sample to the mixing chamber.29. The apparatus of claim 28, wherein the aspirator is further adapted to provide an inert gas in order to control the flow rate of said sample.30. The apparatus of claim 29, wherein the inert gas is further infra-red transparent.31. The apparatus of any of claims 23 to 30, further comprising a reflective optical element adapted to reflect said infra-red radiation so as to increase the path length through said flame.32. The apparatus of any of claims 23 to 31, wherein the burner head comprises a first diffusion barrier mounted on a first side along its length so as to prevent ingress of air to said flame along said length, when in use.33. The apparatus of claim 32, wherein the burner head comprises a second diffusion barrier mounted along a second side of said burner head so as to face the first diffusion barrier and prevent the ingress of air to the flame from said second side when in use.34. The apparatus of claim 33 wherein the surfaces of the first and second diffusion barriers facing each other are optically reflective.35. The apparatus of claim 34 further comprising at least one optical element adapted to direct said infra-red radiation at an angle towards said optically reflective surfaces in order to increase the path length of said radiation through said flame.36. The apparatus of any of claims 32 to 35, wherein the diffusion barrier(s) are cooled.37. The apparatus of any of claims 23 to 36, wherein the burner head further comprises an inert gas boundary layer at at least one end.38. The apparatus of any of claims 23 to 37 wherein the burner head further comprises a structure to fully enclose the flame, in use.39. The apparatus of any of claims 23 to 38, wherein the burner head is mounted on a movable table.40. The apparatus of any of claims 25 to 39, further comprising means for cooling said gases before reaching the mixing chamber.41. The apparatus of claim 40, wherein said means for cooling the gases is cooling coils.42. The apparatus of any of claims 23 to 41, wherein an externally cooled surface is provided to the flame in use.43. The apparatus of any of claims 23 to 42, further comprising a filter element which measures a specific portion of the full infra-red spectrum 44. The apparatus of claim 43, wherein the filter element further controls the supply of gases to the flame.