Method and apparatus for determining concentration of NH-containing species

Abstract

A method and apparatus for determining ammoniacal species concentration in a gas sample. In one embodiment, trace concentration of ammonia in an air sample is determined by monitoring emission intensity from an excited radical species (NH*), which is produced in a reaction between ammonia and fluorine. The observed emission intensity is compared with calibration data obtained from previously analyzed gas samples containing ammonia. The method and apparatus can also be adapted to detect ammoniacal species concentration in other NH-containing gas samples.

Description

CROSS REFERENCES AND RELATED APPLICATION

[0001] This application is a divisional of co-pending U.S. patent application Ser. No. 09/624,118, filed Jul. 24, 2000, which also claims priority to U.S. provisional patent application Serial No. 60/147,017, entitled “Method and Apparatus for Determining Concentration of NH-Containing Species,” filed on Aug. 3, 1999, which are herein incorporated by reference.

BACKGROUND OF THE DISCLOSURE

[0002] 1. Field of the Invention

[0003] The invention relates generally to a method and apparatus of determining gas phase species concentration, and more particularly, to a method and apparatus for detecting concentration of NH-containing species.

[0004] 2. Description of the Background Art

[0005] The use of ammonia (NH3), a corrosive and toxic gas, in industrial processes is wide spread. Trace amount of NH3 has also been shown to adversely impact the use of chemically activated deep ultraviolet photoresists in advanced semiconductor fabrication. The need for worker protection, from either acute exposure to high NH3 concentrations or long term exposure to very low concentration levels, has resulted in the development of sampling methods for the detection and quantitative measurement of NH3 in ambient air. Some existing analytical techniques for NH3 detection are briefly described below.

[0006] a. Electrochemical Method

[0007] In this method, gaseous NH3 is absorbed into an electrochemical sensor assembly with a resultant change in the electrical conductivity of the sensor cell. The increased current flow allowed by the sensor is fairly linear over the concentration range of 1-50 ppm. A lower detection limit is about 500 ppb, but reproducibility of the sensor to periodic exposure of NH3 is only fair.

[0008] b. Ozone Method

[0009] This method uses a reaction between ozone (O3) and ammonia, in which NH3 is first converted to NO2, followed by a chemiluminescent reaction between NO2 and O3. The reaction with O3 results in the formation of excited state NO2 molecules, denoted as NO2*, and the intensity of emission from NO2* is used to determine the original NH3 concentration. However, difficulties in quantitative measurement result from side reactions during the conversion from NH3 to oxides of nitrogen (forming NO and, perhaps, NO3 or HNO), and also from non-stoichiometric side reactions between NO2 and O3. In addition, the emission from excited NO2 species (NO2*—the asterisk “*” is used in this disclosure to designate an excited state of a species) extends from the near UV into the yellow-green region of the visible spectrum (this emission is the well known “air afterglow” in the night sky, and results from the reaction: NO+O2→NO2*+O). Detection of this very diffuse emission over a broad spectral region is susceptible to interference from other emitting species, and may pose difficulties in accurate concentration determination.

[0010] c. Air Sampling Method

[0011] In this method, air samples are collected via a carefully prepared evacuated sampling ampoule and injected into a gas chromatograph (GC) for comparison against analyzed standards by well known methods. Careful selection of the GC column and temperature settings are necessary in order to obtain reliable results. A number of detectors are available for this method. One very sensitive detection method is mass spectrometry, but calibration for quantitative work is very difficult. Additionally, the GC/MS method is very expensive, and it is difficult to configure in a continuous sampling mode.

[0012] d. Laser Induced Emission

[0013] This method has the potential for great sensitivity, but requires great expertise and expense due to its sophistication. NH3, or a fragment thereof, is electronically (or vibrationally) excited by a pulsed, tunable dye laser, thereby creating observable fluorescence. However, non-linear optical effects and saturation effects tend to make quantitative measurements extremely complex, if at all possible.

[0014] Each of these prior art techniques has its own limitation and varying degrees of experimental complexities. Therefore, a need exists in the art for alternative analytical methods that allow continuous on-line determination of low level of ammonia in ambient air or gas samples.

SUMMARY OF THE INVENTION

[0015] Embodiments of the invention generally provide a method and apparatus for determining the concentration of an NH-containing species in a gas sample. The method comprises detecting radiation from excited imidogen radicals (NH*) generated from the gas sample, and determining the concentration of the NH-containing species from calibration data correlating detected NH* radiation intensity with concentration of the NH-containing species. In one embodiment, the NH-containing species is ammonia (NH3), and the NH* radiation is generated by reacting NH3 with a gas sample containing fluorine. Using a bandpass optical device, NH* radiation around 336° nm can be selectively transmitted and detected, with minimal interference from other emitting species.

BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The teachings of the present invention can readily be understood by considering the following detailed description in conjunction with the accompanying drawings, in which:

[0017] FIG. 1 depicts a schematic diagram of an apparatus for determining ammonia concentration according to one embodiment of the invention; and

[0018] FIG. 2 is an illustration of a calibration plot that can be used for determining concentration of ammoniacal species.

[0019] To facilitate understanding, identical reference numerals have been used, where possible, to designate identical elements that are common to the figures.

DETAILED DESCRIPTION

[0020] The present invention generally provides a method and apparatus for determining concentration of an ammoniacal (ammonia-like, or NH-containing) species in a gas sample. In one embodiment, the ammoniacal species is ammonia (NH3). It is known to those skilled in the art of molecular spectroscopy that gaseous NH3 and molecular fluorine (F2) will spontaneously react, typically at sub-atmospheric pressures. As is the case with many gas phase reactions, several chemical reaction pathways are possible, giving rise to different reactive or non-reactive intermediate or product species. It is also known that light emission accompanies this spontaneous reaction, and that the emission is characteristic of energetic, or excited state, species generated in the reaction. Among these electronically excited species is the diatomic imidogen free radical (NH*), which has a spectral emission in an unusually narrow wavelength region around 336° nm (due to the NH A3Π-X3&Sgr;− transition). This light emission, also known as fluorescence, is the predominant emission in the visible and ultra-violet (UV) region of the optical spectrum from the spontaneous reaction between NH3 and F2. When the emission is generated from a chemical reaction, it is sometimes referred generally as chemiluminescence.

[0021] Embodiments of the invention provide a method and apparatus by which a trace concentration of NH3 can be determined from a functional relationship between the NH3 concentration and the observed NH* emission intensity, where NH* is used to denote generally an excited state of the NH species. In particular, the method relies on two assumptions: 1) that the detected NH* emission intensity (INH*) is proportional to the concentration of NH* species; and 2) that the concentration of NH* is in turn correlated with the initial NH3 concentration prior to the reaction with F2.

[0022] The first assumption can be expressed as:

INH*°=°k*[NH*]  Eq.(1);

[0023] where k* is a proportionality constant related to a variety of factors specific to the experimental setup, including light collection efficiency, detector sensitivity, and the like; and [NH*] is the concentration of excited NH species present in the detection volume.

[0024] The second assumption can be expressed as:

[NH*]°=°k1f([NH3])  Eq.(2);

[0025] where k1 is a proportionality constant, and f([NH3]) denotes generally a function of the concentration of NH3. Again, k1 is an experimental constant which depends on a variety of factors related to the reaction kinetics between NH3 and F2. This, along with Eq. (1) above, leads to:

INH*=kf([NH3])  Eq.(3);

[0026] where k=k*k1.

[0027] According to the method of the invention, the concentration of NH3 present in a gaseous sample can be determined by experimentally measuring the intensity of emission from NH*, and determining the NH3 concentration [NH3] from the functional relationship of Eq.(3). The exact functional relationship f([NH3]) can be obtained by a calibration procedure to be described below. The method is particularly suited to the determination of trace level of NH3 in a gas sample.

[0028] In general, the concentration of an intermediate species in a reaction, such as an excited state of a reactive radical (NH*), is very low, and one may encounter difficulties in detecting emission from such a species. However, one can take advantage of the fact that the predominant visible and UV emission from the NH3+F2 reaction originates from NH*. By using a suitable bandpass optical device, such as an optical interference filter or monochromator, one can selectively transmit and detect the NH* emission around 336° nm to the exclusion of background signal from other emitting species. Any background emission, if not properly excluded, may interfere with (i.e., contribute to) the observed light emission intensity and thus affect the accuracy of the determination of NH3 concentration.

[0029] Since the reaction between NH3 and F2 occurs in the absence of heating or other external energy sources (i.e., as a “dark reaction”), the resulting fluorescence can be measured against a dark background. This allows the use of extremely sensitive light detection methods, such as photon-counting, to detect and quantify trace amounts of NH3 present in a gas sample, such as an ambient air sample containing NH3. Hence, the invention has superior sensitivity over existing methods.

[0030] An apparatus suitable for practicing the present invention is illustrated schematically in FIG. 1. The apparatus 10 comprises a vacuum system 100 and an optical detection system 160. The vacuum system 100 further comprises a reaction vessel 102 (or reactor) connected to a pressure-reducing device such as a vacuum pump 180 and other gas flow and pressure regulating components. As shown in FIG. 1, the gas flow and pressure regulating components may illustratively comprise vacuum valves 104, 106, 108 and 110. The valve 104 controls gas flow at the inlet 124, while the valve 108 controls gas flow at the outlet 128. At least one of these valves 104, 108 should have an adjustable orifice for variable gas flow control, such as that provided by a needle valve. Different needle valves with varying sizes of orifices can result in a fine control of the gas flow up to a range of, for example, 500° sccm. The exact flow range, however, is not critical to the practice of the present invention.

[0031] The valve 110 is a throttle valve connecting the outlet 128 to the vacuum pump 180 via a vacuum line 184. For example, the vacuum pump 180 may be a mechanical pump with inert fluorocarbon oil having a 2 CFM pumping capacity. The exhaust gases, including carrier gas, reactant and product gases, are evacuated through the vacuum line 184. The pumping capacity (or speed) provided to the vessel 102 can be varied by adjusting the throttle valve 110. The adjustment of valves 104 and 108, in conjunction with the throttle valve 110, allows control of the gas flows through the vessel 102. Thus, a partial vacuum in the range of about 0.1° mbar to about 50° mbar can be achieved inside the vessel 102. A pressure transducer 182 is also provided for pressure measurement. It is preferable that more than one pressure gauge be used for pressure monitoring at different pressure ranges. For example, capacitance manometers available from MKS Instruments, Inc., Andover, Mass., are suitable for this purpose.

[0032] The reaction vessel 102 also comprises a second inlet 126. A valve 106 is used to control the gas flow through the inlet 126, which extends into the interior 102I of the vessel 102, and terminates in an inlet tip 127. An optical window 152 is provided on one side 103 of the vessel 102 at close proximity to the inlet tip 127. The reaction vessel 102 is preferably made of glass or quartz, but other materials such as stainless steel are also acceptable, as long as it is compatible with the chemicals or gases used. The optical window 152 should be made of a material which can transmit radiation around 336° nm. In general, any ultra-violet (UV) transmitting materials such as different grades of quartz will suffice.

[0033] To perform the NH3 concentration measurement according to embodiments of the invention, the vessel 102 is evacuated with the throttle valve 110 and valve 108 fully open. After a base pressure of about 0.1° mbar or below is reached, the valve 108 is closed to some appropriate intermediate position while a gas sample to be analyzed, e.g., air containing an unknown concentration of NH3, is introduced into the vessel 102 via the inlet 124. The flow rate of this NH3/air sample through the vessel 102 can be controlled by adjusting the valves 104 and 108. Asteady flow of the gas sample may be established within a range of about 100-500° sccm, preferably at about 300° sccm. An operating pressure in the range of about 0.1° mbar to about 50° mbar, preferably at about 10° mbar, may be used.

[0034] With the NH3/air flow rate and pressure established, a second gas sample containing fluorine—e.g., a dilute mixture of F2 in a carrier gas such as argon (Ar) or helium (He), is then is introduced into the vessel 102 through the inlet 126 by the valve 106. This fluorine-containing gas sample, also referred to as a reactant gas, is used to initiate a reaction between NH3 and F2. The reactant gas is preferably a highly diluted mixture of F2 in a non-reactive carrier gas such as ultra-high purity (UHP) Ar or UHP He. Of course, other similarly non-reactive or inert gases, e.g., nitrogen (N2), may also be used as a carrier gas, provided that they do not substantially interfere either with the NH3+F2 reaction or the detection of the NH* emission. The F2/carrier gas mixes with and reacts with the flow of air containing NH3 (or generally, the gas sample to be analyzed) just down-stream of the gas inlet tip 127. This counter-flow reaction method and apparatus design is well known in experimental gas kinetics.

[0035] A reaction zone 150, where F2 and NH3 reaction occurs, is generally defined in the vicinity of the reactant gas inlet tip 127 inside the vessel 102. By controlling the flow rate of the F2 gas into the air/NH3 flow stream, one can confine the reaction zone 150 to within a relatively small, well-defined volume. A better defined reaction zone 150 is preferable because it allows an efficient collection and detection of the chemiluminescence.

[0036] As the reactant gas reacts with NH3 in the NH3/air sample, emission from excited NH* species is detected using the optical detection system 160 to be described below. The reactant gas flow should be adjusted so as to maximize the NH* emission intensity INH* detected by the optical detection system 160. That is, at a fixed flow rate of NH3/air, the reactant gas flow should be sufficiently high such that additional F2 (or reactant gas) will not result in an increase of detected INH* signal for a given configuration of the optical detection system 160.

[0037] It is understood that the process parameters disclosed herein are meant to be illustrative, and other gas flow rates and operating pressures may be adjusted as appropriate to different reaction vessels. In general, the choice of the operating pressure may be based on several considerations—e.g., a higher operating pressure tends to favor a higher reaction rate between NH3 and F2. However, a higher pressure also results in increased collisions between the excited NH* and other gas molecules. These collisions may lead to “quenching” of the NH* emission, and thereby reduce the amount of detectable optical signal. Therefore, an optimal operating pressure may involve balancing these competing considerations, and one can experimentally arrive at the desired operating pressure by establishing initial flows of the NH3 and F2 gases, and adjusting valves 104, 106, and 108 to maximize the NH* signal. Such optimization technique is well-known to one skilled in the art of chemical kinetics.

[0038] If the gas sample to be analyzed is being used as a process gas in a certain process application, the apparatus 10 may also be used for continuous on-line measurement of NH3 concentration in the process gas. For example, the apparatus 10 may be connected (e.g., at its inlet 124) to a reactor (not shown) used for the particular process application, and a relatively small flow of the process gas may be diverted from the reactor into the reaction vessel 102 via the inlet 124. The NH3 concentration may then be continuously monitored according to embodiments of the invention, without interfering with the particular process application.

[0039] Optical Detection System

[0040] The light emission from the reaction of NH3 and F2 (due to the NH A3Π-X3&Sgr;−. transition), is transmitted through a suitable optical window 152 and a bandpass optical device 162, and detected by a detector 164. A lens 161, or similar imaging optics, may also be used to facilitate the collection and imaging of light emission from a sample volume (e.g., the reaction zone 150) onto the detector 164.

[0041] The bandpass optical device 162 preferably has a bandpass that is sufficiently narrow as to transmit the NH* emission near 336° nm, while substantially rejecting emissions from other species that may interfere with the detection of the NH* emission (i.e., selectively transmitting the desired NH* emission). In one embodiment of the invention, a narrow bandpass filter 162, e.g., an interference filter having a bandpass of about 10° nm (i.e., full-width bandpass at half-maximum intensity, or FWHM), with a peak transmission of about 10-50% around 336° nm may be used. Due to the “piling-up” of the Q-branch of the NH A3Π-X3&Sgr;− electronic transition, most of the NH* chemiluminescence can be transmitted through the interference filter 162, which also effectively blocks other undesirable or background emissions, thus facilitating the detection of NH* emission. Such an interference filter is available from commercial optics supply vendors. The optical characteristics of the interference filter cited herein are meant to be illustrative. It is understood that filters with different optical characteristics (i.e., FWHM bandpass, peak transmission percent and peak wavelength) may also be used to transmit the NH* emission for practicing embodiments of the invention. For example, if measures are taken to eliminate interfering emissions (e.g., by eliminating species having interfering emissions), a wider bandpass filter may be tolerated.

[0042] In other embodiments, the bandpass optical device 162 may comprise a combination of different filters that is effective for selectively transmitting the desired NH* emission, while blocking interfering emissions from other species. For example, the combination may include a longpass filter and a shortpass filter with appropriate cut-off wavelengths, or a bandpass filter having a FWHM bandpass larger than about 10° nm and a suitable cut-off filter. One example of a possible interfering emission originates from OH radicals, which may arise from the presence of moisture or other reactions in the reactor. It is known that an excited state of the OH radical has a strong emission around 306° nm. If a short wavelength cut-off filter (or longpass filter) is used to block the 306 emission from excited OH radicals, then a filter having a FWHM bandpass larger than about 10° nm may be used. Other bandpass optical devices such as a monochromator or similar equipment with wavelength selection capabilities can also be used in place of an interference filter.

[0043] The light emission that passes through the bandpass device or filter 162 is incident upon the detector 164, which is selected to be sensitive to the transmitted emission. For example, the detector 164 may be a RCA 1P28 photomultiplier tube operating at about 800V DC. The photocurrent generated by the emission can be detected using commercially available detection and amplification equipment 166. Suitable detection and amplification equipment 166 may include picoammeters or photon-counting devices with dynode pulse discrimination electronics, among others. In general, various combinations of detectors and amplification equipment may be used to detect the emission through the optical device 162 and convert it to a radiation intensity parameter that correlates with the intensity of the NH* emission. The apparatus 10 should preferably include a device 168 for monitoring and/or recording of the amplified optical signal, or more generally, the radiation intensity parameter. The device 168 may illustratively be a computer that interfaces with the detection and amplification equipment 166 and provides for data storage and retrieval.

[0044] Calibration of the apparatus 10 is accomplished with dilute, analyzed samples containing NH3—e.g., NH3 in N2 or in air, or other suitable carrier gases. A calibration plot, for example, is constructed by plotting the chemiluminescent intensity during reaction with excess F2 against known NH3 concentrations [NH3] from analyzed, calibration gas samples.

[0045] Calibration Procedure

[0046] In order to determine the concentration of NH3 in an unknown gas sample, a calibration procedure is performed in the reaction vessel 102 to generate calibration data which correlate detected NH* emission intensities with known NH3 concentrations in calibration gas samples. The calibration gas samples, e.g., NH3/air mixtures, can be analyzed to obtain known NH3 concentrations by conventional analytical methods, or prepared by successive dilutions from more concentrated mixtures that are amenable to conventional analytical techniques, or procured as analyzed mixtures from any number of industrial gas suppliers.

[0047] The calibration procedure involves experimentally measuring the NH* emission intensities (INH*) from reaction with F2 for several calibration gas samples with known NH3 concentrations [NH3], using the procedure previously described. For example, NH* emission measurements can be performed for each of several calibration gas samples containing NH3 concentrations between a few hundred to a few thousand parts per billion (ppb) by mixing each of the calibration gas samples with a reactant gas containing fluorine. Although the fluorine-containing reactant gas used in the calibration procedure may be different from that used in the reaction with the gas sample having the unknown NH3 concentration, it is preferable and more convenient to use the same reactant gas (e.g., similar F2 concentration and/or carrier gas).

[0048] The calibration data comprising detected emission intensity INH* (or a calibrated radiation intensity parameter correlating with INH*) and its corresponding NH3 concentration [NH3] may be represented in a calibration plot, such as that illustrated in FIG. 2, which can be extrapolated to lower concentrations. More generally, a functional relationship between INH* and [NH3] can be derived from the calibration data. The NH3 concentration in a gas sample with an unknown [NH3] can then be determined by comparing the observed NH* emission intensity (from a reaction between the NH3-containing gas sample and the reactant gas containing fluorine) with the calibration data, or from the functional relationship correlating NH* emission intensity INH* with NH3 concentration.

[0049] In one illustrative embodiment, a mixture of 0.5% F2 in UHP Ar can be used as a reactant gas for calibration. In practice, it is not possible to maintain a stable concentration of highly diluted F2 gas in a vessel (gas cylinder) for an extended period of time due to the corrosive or reactive nature of F2. However, this will not affect the calibration procedure because F2 is introduced in “excess” to produce a maximum detected INH* for the given detection system 160. The reaction vessel 102 is first evacuated to a pressure of about 0.1° mbar. After a steady flow of one calibration NH3/air sample (i.e., previously analyzed, with known NH3 concentration) is established, e.g., at a pressure of about 10° mbar and a flow rate of about 300° sccm (or generally, the same pressure and flow rate as used for the reaction with the gas sample having unknown NH3 concentration), the F2/Ar reactant mixture is introduced into the vessel 102 at the inlet tip 127. To ensure that F2 is in “excess”, the F2/Ar reactant gas is introduced until the observed NH* emission intensity no longer increases with additional F2/Ar reactant mixture. The NH* emission from the reaction zone 150 is detected as described above using detector 164.

[0050] The observed intensity INH*, along with the known concentration of the previously analyzed NH3/air sample, may be recorded or stored in a suitable medium, e.g., a computer. The calibration procedure is repeated for the remainder of the calibration gas samples (that have previously been analyzed to obtain known NH3 concentrations). A calibration plot such as that shown in FIG. 2 can be recorded, showing the observed intensity INH* vs. the known NH3 concentration. In general, a functional relationship derived from the calibration data will be used for the determination of NH3 concentration in gas samples. The calibration results should ideally be recorded by electronic means, for example, a computer or processor having a storage device, to facilitate data storage and retrieval.

[0051] Other NH-Containing Samples

[0052] Although the present embodiment focuses on the measurement of NH3 in a gas sample, this invention can be extended to the determination of other ammoniacal, or ammonia-like species, e.g., molecular species with a N-H bond, such as organic amines, imines and other NH-containing species. Although the detailed chemical reactions may differ, it is anticipated that reactions between different ammoniacal species and fluorine (contained in the reactant gas sample) will lead to the formation of excited NH species, or NH*. Depending on the specific ammoniacal species, the reaction may not be “dark”, as previously explained for the case of NH3. However, the use of a narrowband interference filter should still suffice to isolate the emitted radiation from NH*, for example, around 336° nm, to allow for a determination of the ammoniacal species concentration. Of course, a separate calibration procedure has to be performed as previously described for a number of the gas samples containing known concentrations of the species of interest. The choice of pressure and flow parameters used in the F2 reaction can readily be arrived at through experimentation which is known to those skilled in the art of chemical kinetics.

[0053] Although one embodiment which incorporate the teachings of the present invention has been shown and described in detail herein, those skilled in the art can readily devise many other varied embodiments that still incorporate these teachings.

Claims

1. An apparatus for determining concentration of an ammoniacal species in a first gas sample, the apparatus comprising:

a reactor having a first inlet for introducing said first gas sample into a reaction zone inside said reactor;

an optical detection-system to detect radiation arising from said first gas sample inside said reactor;

a second inlet for introducing a second gas sample containing fluorine into said reactor, wherein said radiation arising from said first gas sample is generated from a reaction between said first gas sample and said second gas sample within the reaction zone inside said reactor; and

a data acquisition and storage system to convert said detected radiation into a radiation intensity parameter, wherein said radiation intensity parameter is used to determine concentration of said ammoniacal species in said first gas sample.

2. The apparatus of claim 1, wherein said reactor further comprises a pressure measuring device, first and second gas flow controllers to control said first and second gas sample flows into said reactor and a vacuum pump, wherein said pressure measuring device, said first and second gas flow controllers and said vacuum pump cooperate with each other to maximize said radiation generated from said reaction between said first and second gas samples.

3. The apparatus of claim 2, wherein a reaction between said first and second gas samples is performed in a pressure range of about 0.1 to about 50 mbar.

4. The apparatus of claim 2, wherein said optical detection system comprises an optical device which selectively transmits radiation originating from said reaction between said first and second gas samples and a photodetector capable of detecting said transmitted radiation.

5. The apparatus of claim 4, wherein said optical device transmits radiation with a full-width half-maximum bandpass of between about 331 nm and about 341 nm.

6. The apparatus of claim 1, wherein said first gas sample comprises ammonia.

7. The apparatus of claim 5, wherein said radiation arising from said first gas sample originates from NH* radical.

8. The apparatus of claim 6, wherein said concentration of ammonia in said first gas sample is determined by comparing said radiation intensity parameter with at least one provided set of calibration data regarding ammoniacal species concentration or a functional relationship correlating detected radiation from excited imidogen radicals with ammoniacal species concentration.

9. The apparatus of claim 8, wherein said calibration procedure is performed inside said reactor by reacting each one of a plurality of calibration gas samples comprising known concentrations of ammonia with a reactant gas comprising fluorine, detecting radiation from NH* radicals generated from each reaction, converting said detected radiation into a calibrated radiation intensity parameter for each of said plurality of calibration gas samples, and forming calibration data associating said calibrated radiation intensity parameter with its corresponding said known concentration of ammonia.