Patent Application
20050032230
This invention relates to a method and apparatus for analyzing a gas for trace amounts of oxygen and more particularly to a method and apparatus for analyzing trace amounts of oxygen in a relatively pure gas using gas chromatography and a flame ionization detector.
For many years, the gas industry has been faced with ever more stringent requirements for purity in industrial gases for research and for the electronic fabrication industries. For example, the need for higher purity was recognized in U.S. Pat. No. 5,612,489 of Ragsdale et al. As recognized in the Ragsdale et al. patent, oxygen is one of the contaminant gases for whichever more stringent requirements are needed particularly for inert gases used for prevention of oxidation. Typical gases include nitrogen and argon. As taught by Ragsdale et al. a low level of interactive gas is doped into the carrier gas carrying the sample gas which is to be analyzed with a known low level typically less then 10 parts per million (PPM) of doping oxygen into the carrier gas. By so doing, the detection limit for the trace oxygen is improved. The patent states that this technique permits reproducible detection of trace oxygen at quantities less than 1000 ppm. It also alleges that it is useful in the range of less then 1 ppm on a volumetric basis. However, in such systems, a detector that is sensitive to the interactive gas i.e. oxygen must be used.
As stated in Ragsdale et al. a detector sensitive to the interactive gas is selected from the group consisting of a thermo conductivity detector, a discharge ionization detector, a helium ionization detector and a high frequency discharged detector.
Conventional gas chromatography cannot reach the low detection levels of oxygen required if a thermal conductivity detector is used. To be more specific, thermal conductivity detectors do not achieve even low parts ppm range for oxygen molecules. Further, flame ionization detectors had not been used as oxygen detectors, because they use oxygen as an oxidizing agent in the flame.
It is now believed that there is a commercial market for an improved method and apparatus for detecting trace amounts of oxygen in a relatively pure inert gas using gas chromatography and a flame ionization detector. There should be a demand because it has been found that such apparatus and methods are capable of detecting trace amounts of oxygen in amounts of less than 560 ppb oxygen with a conventional gas chromatograph and a flame ionization detector.
In essence, the present invention contemplates an improved method for analyzing trace amounts of oxygen in a gas mixture such as a relatively pure gas. The oxygen is separated from the gas mixture as for example in a gas chromatograph and the separated oxygen gas is converted to a carbon oxide such as carbon monoxide and/or carbon dioxide by passing the oxygen over carbon at an elevated temperature. The carbon oxides are then converted to methane by mixing the carbon oxides with hydrogen with a heated nickel catalyst. The methane is then introduced into a flame ionization detector which produces a count indicative of the amount of methane. This amount of methane is indicative of the amount of oxygen in the original gas mixture.
In a preferred embodiment of the invention, the method for analyzing trace amounts of oxygen in a gas mixture includes the step of comparing the amount of methane produced in a control sample with the amount of methane produced in a test sample as an indication of the amount of oxygen in the test sample.
The preferred embodiment of the invention also includes the steps of providing a gas chromatograph, a flame ionization detector, a mass of heated carbon material and a heated nickel catalyst. In this embodiment of the invention, a gas to be analyzed such as a relatively pure gas with a suspected trace amount of oxygen is mixed with a carrier gas such as helium and introduced into a gas chromatograph . The carrier gas and relatively pure gas mixture is passed through a column of a gas chromatograph to separate any oxygen from the relatively pure gas. The oxygen and carrier gas is then passed through a mass of carbon at an elevated temperature to thereby form carbon oxides i.e. carbon monoxide and/or carbon dioxide. The carbon oxide and a mass of hydrogen gas is then passed over a heated nickel catalyst to form methane which is then introduced into a flame ionization detector. The detector provides an indication of the amount of methane and consequently an indication of an amount of oxygen in the relatively pure gas which is being tested.
a is a graph taken from a strip chart recorder illustrating periodic injections of oxygen having 520 ppb standard gas, with retention times of 2.67, 7.66, 12.67, etc.
b is the second run of injections of 520 ppb standard gas with the same retention times as
An example of a conventional injector port 20 is illustrated in
As illustrated in
The carrier gas and sample gas pass through the column 12 and into a conventional detector 32 which indicates the composition of the sample gas and records the composition with a conventional recorder 34.
In the practice of the present invention, an inlet 40 shown in
The heated carbon may comprise most clean carbon sources. For example, charcoal or activated carbon can be used. The carbosieve used in the examples, disclosed herein, was taken from a charcoal filter from a Chrompack International (catalogue number 07942) of Middelburg, Netherlands. The particle size used ranged from 1-2 mm. The temperature used was 540° C. but could be higher.
Hydrogen gas is fed to a chamber 46 from the gas chromatograph as indicated by an arrow 45. The chamber 46 contains a heated nickel catalyst which converts the carbon oxides from the heated carbon containing tube 44 into methane.
As illustrated in
Analyzing a relatively pure gas for trace amounts of oxygen using gas chromatography and a flame ionization detector will now be described with reference to
The temperature in the thermostat zones (e.g. sampling loop, nickel catalyst, injector, oven and detectors were controlled using the control system of an HP 5880A gas chromatograph. The carrier gas which was helium in the experiments bypasses the sampling loop before injection of the sample. Upon injection, the sample is mixed with the carrier gas and enters a column (not shown in
The oxygen is separated from the other gases by the end of the column. The outlet of the column of the gas chromatograph is connected to a heated tube 44 that contains carbosieve material. The carbosieve materials is preferably activated at about 550° C. under helium before usage. The temperature of the carbosieve tube is fixed at about 540° C. during the experiments. The temperature of the carbosieve was controlled and monitored by a THERMOLYEN controller (model 12900) and with a LEYBOLD digital thermometer (model 666452) using K type thermal couples. Upon passing the heated carbosieve tube, oxygen will react with carbon to produce carbon oxides. The outlet of the carbosieve tube and a hydrogen gas 45 from the gas chromatograph are connected with a heated nickel catalyst tube 46. The nickel catalyst is maintained at a temperature of about 350° C. The catalyst could also be heated at about 400° C. With the help of a catalyst, carbon oxides react with hydrogen to give methane and water. The outlet of the nickel catalyst tube is then connected to the flame ionization detector 54 to analyze methane.
The detection limit of the flame ionization detector (10−12 g/ml) limits the detection limit of this method i.e. low ppb. Hence, the oxygen gas is analyzed by a flame ionization detector. Thus, trace analysis of oxygen can be routine work if a gas chromatograph is equipped with a carbosieve tube and the nickel catalyst. Accordingly, this method is a very economic method for trace analyses of oxygen.
To verify the oxygen was the analyzed species and analysis of high purity oxygen where the flame ionization detector (FID peak) was very large, the temperature of the carbosieve material during the run was gradually reduced. The intensity of oxygen peaks continued to decrease until it disappeared because there was no conversion of oxygen to carbon oxides. Also, when air was analyzed a large peak could be seen at an oxygen retention time. Since air does not give FID peak under normal conditions, the detected PEAK must be due to the converted oxygen. When high purity nitrogen was analyzed no peaks were detected at the nitrogen expected retention times. However, a peak appeared at an oxygen retention time. This oxygen is usually present as a contaminant in a high purity nitrogen gas. In addition, to the confirmation of the supplier, the high purity of the nitrogen cylinder was confirmed by using the thermal conductivity detector (TCD) of the gas chromatograph.
The experimental procedure protecting the method follows:
An HP 5880 A gas chromatograph equipped with thermal conductivity detector (TCD), flame ionization detector (FID) and 5.0 ml-loop-6-port-V calco valve was used in all of the experiments. The carrier gas was helium. A column of 13 X molecular sieve 4′* ⅛″ in conjunction with carboxen 1004, 6′* 1-8″ mark has been used. The oven temperature was 90° C. Flow rate of helium was 20 ml/min. Injector temperature was 200° C. and the detectors' temperatures were 250° C. The pressure of H2 and air for FID were 70 and 40 psi, respectively. The sample was introduced through a {fraction (1/16)}″ stainless steel line, see
The nickel catalyst was activated over night over hydrogen gas and 350° C. before use. The temperature of the catalyst was kept at 350° C. in the analysis. The temperature of the carbosieve was kept at 540° C.
Trace levels of oxygen in helium were prepared by passing high purity (99.999%) helium through fresh Hewlett Packard (PN 3150-0414) oxygen scrubber. This scrubber was designed to give less then 1 ppm oxygen in the mixture at a flow rate of 3 liters per minute. However, the flow rate used was 5 ml/min. This gave oxygen levels of much less then 1 ppm.
The outlet of the scrubber was connected to the sampling loop so that the flow rate was kept a 5 ml/min. About 5.0 minutes were allowed to flush the sampling loop before each injection.
Eleven samples were injected during a gas chromatograph run. The largest peaks were due to the oxygen in the outlet of the scrubber. Data acquisition and processing perimeters were:
High purity 99.999% helium gas was analyzed without passing through an oxygen scrubber. No peaks were detected when the same gas was analyzed using the thermal conductivity detector of the gas chromatograph. Oxygen peaks resulted from periodic injections of high purity 99.999% helium samples. The large peaks and their retention times were due to the oxygen in the helium gas. The data acquisition perimeter were the same as that in example 2. The first two peaks were smaller then the other peaks, since the steady state of the system had not been reached. However, the rest of the eleven samples gave reproducible results with an error of plus or minus 1.5%.
It was difficult to obtain a low ppb of oxygen gas as a standard. Many well known companies around the world apologized for not supplying any oxygen gas in the ppb range because they do not have equipment to measure such low levels of oxygen. The minimum concentration obtained was 520 ppb from the International Gasses and Chemicals Limited (INTERGAS), England it is an ISO 9002 company.
a and 5b shows oxygen peaks resulting from periodic injection of 520 ppb of oxygen in helium balance. The attenuation of the GC signal equals 28, the peak equals 0.02 min., threshold equals 4, and the chart speed was equal 0.2 cm per min. The average of the counts of the gas chromatograph in
In those cases where quantitative analysis is needed, an oxygen standard may be used. For example, for quantifying the analyzed oxygen one can use an internal standard method without the need of a calibration curve. However,
An important result is that the gas chromatograph can easily detect compounds that produce 10 GC/counts. This means that one can easily detect very low ppb oxygen gas by using the current gas chromatographs. This provides, for the first time, the opportunity for the gas industry to analyze and prepare oxygen traces in the low ppb range using gas chromatographs.
While the invention has been described in connection with its preferred embodiments, it should be recognized the changes and modifications may be made therein without departing from the scope of the claims.
