JPH04274728A - Method and apparatus for preliminary enrichment for analyzing minute amount of component in gas - Google Patents

Abstract

PURPOSE: To obtain a method and device for preconcentrating for analyzing trace components in gas. CONSTITUTION: After sample gas A is introduced into a sorbent confinement pipe 10, a preconcentration device for analyzing the trace components discharges the gas by a vacuum pump 18, and trace components separated when low pressure carrier gas passes through the sorbent pipe 10 is carried to a low pressure operation detector like a mass spectrometer 14 by the carrier gas. The sample gas is heated to accelerate elimination or the gas acting with trace components is heated to enhance detection. A plurality of valves between sample inlets A, A', carrier gas outlets D, D' and vacuum pumps 18, 18' of the device and a detector 14 secure a sample gas stream to carry the trace components to the detector.

Description

Description: FIELD OF THE INVENTION This invention relates to the detection of trace components in air or other gases, and more particularly, to the detection and measurement of trace components in air or other gases. An improved method and apparatus for preconcentrating components. The present invention
Although it can be used with any detector that basically measures the relative concentration of a gas whose pressure is reduced below atmospheric pressure for analysis, its primary use is for mass spectrometry measurements. It is something. BACKGROUND OF THE INVENTION Preconcentration is a well-known technique in analytical chemistry and is used for the detection of trace contaminants in both gases and liquids. In gas analysis, gases, usually at or near atmospheric pressure, are drawn through or past an adsorbent that selectively adsorbs or captures trace components. This adsorption generally occurs on the surface of solid sorbents, which are often powders, "wool" or sheets. Sorbents are selected on the basis of their ability to adsorb the compounds of interest and their adsorption capacity. The amount of trace component adsorbed is proportional to the amount of time the gas is drawn through or otherwise in contact with the sorbent (the "sampling time") before the sorbent becomes saturated. do. Following the time period during which this sampling occurs, the sorbent releases its trace components, usually by rapidly heating the sorbent. The temperature of the sorbent can be raised sufficiently in one step so that all of the analyte is desorbed at the same time (“flash desorption”), or different species can be desorbed at different temperatures. It can also be ramped up more slowly so that it is released at different times ("temperature programmed desorption"). In either case, the trace components are reintroduced into the carrier gas stream within a much shorter desorption time than the period during which sampling is completed, so that the concentration of the analyte plug in the carrier gas is less than that of the original sample. It is more advanced than its concentrated form in gas. An example of a device that operates in such a manner is provided by Envirochem, Kemblesville, Pennsylvania.
ville, Pennsylvania. ), there is a universal automated concentrator. [0004] The sorbent is in the form of a powder confined within a thin-walled tube that is sufficiently granular to the gas passing through the tube, and during the desorption period, the carrier gas passing through the sorbent is in the form of a powder. It is common for the flow to be in the opposite direction to the flow through it during the sampling period. An example of a preconcentrator used in this method is the Automatic Chemical Agent Monitor System (Automati
c Chemical Agent Monitor
System) (ACAMS), which is the U.S. Army's Southern Research Institute (South).
Hern Research Institute
) and was developed in Lewisburg, West Virginia.
ABB Process Analytics (ia)
Previously Combustion Engineerin
g, Inc. ), and Birmingham, Alabama (B
C.M.Irmingham, Alabama)
S Research Corporation (CMS Res)
Manufactured by EACH Corporation. Preconcentrators are also used in the analysis of volatile compounds in liquids, usually water, by using a "purge and trap" technique. In contrast, an inert gas bubbles through the liquid, entraining volatile contaminants and transporting them to a similar sorbent for the gas sample. After the sampling (purging and trapping) is completed, the analyte is desorbed from the sorbent as described above. Techma, Inc., Cincinnati, Ohio (
Tekmar Company of Cinc
Tekmar LSC 2000 (Innati, Ohio)
0) is a “purge and trap” available on the market.
This is an example of a device. Preconcentrators are commonly used in conjunction with gas chromatographs (GC) to separate and confirm analytes. The flow rate of the carrier gas through the preconcentrator (or final trap) during the desorption phase is generally determined by the characteristics of the GC separation tube. For example, capillary separation tubes use lower mass flow rates than conventional filled separation tubes. moreover,
In order to improve the resolution and sensitivity of the GC, the G
At or near the inlet of the C separation tube, the desorbed analyte is often re-trapped, usually by cryogenic means.
From this, the analyte is rapidly desorbed and within a short period of time, the sample is injected into the separation tube. The above Tekuma LSC2000 "Purge and Trap" instrument is a filling-megabore (
The Tekma Model 1000 capillary interface uses cryotechnical refocusing to connect the Tekma LSC2 "Purge and Trap" instrument to a capillary separator gas chromatograph using cryotechnical refocusing. to be able to work together with the tograph. The following publications are cited for a more comprehensive description of the above prior art devices and techniques and are incorporated herein by reference: 1. S.W. 846, “Test Methods for Evaluation of Solid Wastes” 3rd Edition, EPA Publishing OS.W. 0000846; Methods 0030 and 5040
．． (SW-846, “Test Methods
for Evaluation of Soli
d Waste,”Third Edition,
EPA Publication OSW0000
846;Methods0030and5040. ) 2. D. L. Fox. Analytical Chemistry 61, 1R (
1989), and references (D.L.Fox.Anal.
Chem. 61, 1R (1989), and refer
ences within. ) 3. D. L. Fox. Analytical chemistry 59,280
R (1987), and references (D.L.Fox.Ana.
l. Chem. 59, 280R (1987), and
references within. )4. R.A. Kagel and S.O.Fauer, Analytical Chemistry 58, 1197 (1986). (R.A.Kage
l and S. O. Farwell, Anal.
Chem. 58, 1197 (1986)) 5. P.A. Studler and Duffry Kijowski, Analytical Chemistry 56, 1432 (1984). (P.A. Steud
ler and W. Kijowski,Anal
．． Chem. 56, 1432 (1984)) 0008
]6. R. Otson, G.M. Leach, and L.T.K. Analytical Chemistry 59, 58 (
1987). (R. Otson, J.M. Leach,
and L. T. K. Chung, Anal. Che
m. 59, 58 (1987). 7. T. Beller and J. Lichtenberg, Journal American Water Works Association 66, 739 (1974).
(T. Bellar and J. Lichten
Berg, J. Am. Water Works A
ssoc. 66, 739 (1974). ) 8. United States Environmental Protection Agency, Capillary Separator Gas Chromatography/Mass Spectrometry Purge and Trap Method 524.2 for Volatile Organic Compounds in Water, August, 1986. (
United States Environme
ntal Protection Agency,
“Method 524.2. Volatile
Organic compounds in W
ater by Purge and Tra
p Capillary Column Gas
Chromatography/Mass Spec
trometry,” August, 1986.) 9. N. Kirsien American Laboratory
16(12), 60 (1984). (N. Kirsh
en, Am. Lab. 16 (12), 60 (1984)
．． )10. L.D. Landau and E.M.
Lifshitz's Fluid Mechanics, translated by G. B. Saikes and W. H. Reid. Pergamon Press. London, 1959. P. 59. (L.D.Landau
and E. M. Lifshitz, “Flui
d Mechanics,”Translated
by J. B Sykes and W. H
．． Reid, Pergamon Press, Lon.
Don, 1959, P. 59. ) SUMMARY OF THE INVENTION In existing devices, the carrier gas entering the analyte is generally at atmospheric pressure, if not always near or above atmospheric pressure, with respect to the high pressure side of the final trap. It will be detached. One important feature of the invention is that during the desorption period, in order to increase the concentration of analyte in the carrier gas,
It consists in using substantially reduced pressure on both sides of the trap and also low volume flow for short periods of time. Considering the theory of preconcentration, the total number (N) of trace component molecules taken from a sample of gas and adsorbed in an ideal preconcentrator is expressed as: N=fs ns Ws ts (
1) Here, fs is the fractionated molecular concentration of trace components in the gas, ns is the number density of air molecules (cc-1), which is proportional to the pressure of the gas flowing into the preconcentrator, Ws is the volumetric flow rate (cc/se) entering the preconcentrator
c), and ts is the sampling time (sec
). The same total number (N) of molecules is released from the preconcentrator during the desorption phase, i.e.: N=fd nd Wd td (
2) where the parameters are those appropriate for the desorption phase and where nd and Wd are the number density and volumetric flow rate of the gas leaving the preconcentrator, respectively.
This has the same meaning as equation (1). The subscripts “s” and “d” used here are respectively “sa” in the equation.
mple (sample)" and "desorption". The combination of equations (1) and (2) yields the preconcentration factor, or "gain", as the ratio of the molecular concentrations of the trace components during the desorption and sampling phases.
(gain)' (G): [Formula 1] In an ideal gas, the following equation applies. [Equation 2] Here, P is pressure, K is Boltzmann constant,
T is "K" Kelvin temperature, p is mass density (gm/cc)
, n is the number density and m is the molecular mass. Therefore, equation (3) can also be expressed separately as follows:
The mass flow of gas (Q) is the product of mass density and volumetric flow rate, ie, Q=pW. G can therefore also be written as: ##EQU4## where the subscripts "s" and "d" again indicate the period of sampling and desorption. In viscous flow, the mass flow of a compressible gas through a constriction, such as a tube or channel, is determined by the high and low pressures (Po and Pf, respectively) at the ends of said constriction, so as to obtain a good approximation under isothermal conditions.
), and thus: In equations (6) and (7), Po
-Pf is the pressure difference between the ends of the tube, which is denoted as "ΔP" in equation (8). Also equation (8)
<P> is the algebraic average pressure inside the tube, and Equation (7)
is [Equation 6]. K in equations (6), (7) and (8) is a factor due to the tube geometry and the viscosity coefficient of the gas. Since the viscosity coefficient of a gas is independent of its pressure, the factor K is a constant. The combination of equations (5) and (8) is
The average pressure in the tube and the pressure drop across the tube during the sampling and desorption phases provides an indication of the gain (G). ##EQU00007## To maximize the gas flow rate during the sampling phase, the gas is pumped at a sufficient rate so that the pressure at the outlet end of the preconcentrator is virtually zero. In this case, Ps is again the pressure at the inlet of the sorbent during the sampling period. Substituting these into equation (8) gives the following equation: [Equation 9] Therefore, the combination of equations (5) and (10) gives a relatively simple expression for the gain: number 10
] As an example, if sampling is performed at atmospheric pressure (Ps
= 760 torr) and the pressure difference between both ends of the preconcentrator during the desorption phase is 100 torr, the pressure at the low pressure end will be 40 torr, so
Even if the above sampling and desorption times are equal,
The preconcentration factor G is 32. During the desorption phase, if there is a pressure drop in the tube,
ΔPd is much higher than the pressure <P>d at the low-pressure end of the sorbent, so the expression for the gain is further simplified:
2] From the above equations, it can be seen that by providing different pressures and gas flow rates such that a high pressure and large volume flow rate occurs in the sampling phase of the gas, and a much lower pressure and small volume flow rate in the desorption phase, substantially It will be appreciated that significantly greater gas and vapor preconcentration is achieved, making it suitable for analysis by mass spectrometry. The present invention takes advantage of this concept by exposing the active sorbent to a sample of the gas or vapor to be analyzed at or near atmospheric pressure so that the sorbent within the confined space is It is sealed and its internal pressure is reduced. The sorbent is then heated to desorb the previously adsorbed components into a low-pressure gas flowing into the confined space and transport these components to the ionizing chamber of the mass spectrometer or other low-pressure detector. , and then analyze these components. If the desorption time is less than the time required for adsorption, further concentration of the gases and vapors involved will occur. As this description progresses, the suitability and potential of the present invention will become apparent.
It will be more fully understood by those skilled in the art upon reference to the accompanying drawings. DESCRIPTION OF THE PREFERRED EMBODIMENT Referring to FIG. 1, valves V1 and V2 are variable leak valves. Valve V2 is used to regulate the mass flow rate of said gas through the sorbent 10 during the desorption phase and in relation to the pressure in the low pressure gas source, the desorption period pressure Pd. Valve V1 is a variable leak valve that leads to the mass spectrometer 14 or other low pressure detector and is adjusted to send a sufficiently small sample of the desorbed gas stream from valve E to valve V1 that the vacuum pump of the mass spectrometer 14 17 or other low pressure detector. Valve A, B, C
, D and E are high conductance valves that are opened or closed as described below, valves A and B are open;
Valves C, D and E are then closed. During the sampling phase, valves A and B are open and valves C, D and E are closed. Gas, normally at atmospheric pressure, is drawn through the sorbent 10, shown here as a granular powder inside the tube, towards the vacuum pump 18. After a predetermined sampling time ts, valve A is closed, valve C is opened and valve B remains open, and the vacuum pump 18 reduces the pressure in the sorbent tube 10. The purpose of the bypass, including valve C, is to provide a rapid pump-down of volume to the left side of the sorbent tube 10 (as shown in FIG. 1). This bypass is connected to the sorbent tube 10.
All gas to the left of the sorbent tube 10 must be pumped through the sorbent, which is an obstruction to the gas flow, although it is not essential that it will eventually be pumped down. Without the bypass, the period to complete the preconcentration sequence would be slightly longer. When there is sufficient vacuum in the sorbent tube 10, valves B and C are closed and valve D is opened to allow gas from the low pressure gas source 11 to enter the sorbent tube 10. Valve E is opened at the same time as valve D, or slightly later, to allow low pressure gas to flow through the sorbent 10. After gas flow is established, the sorbent tube is rapidly heated by heater 12,
The components adsorbed on the sorbent 10 are released into the low pressure gas stream. A part of this flow is cut off and the variable leak valve V
1 and an ionizer 15 before entering a mass spectrometer 14 or other low pressure detector. The remaining gas passes through valve V2 and enters pump 18. Heater 12 includes an electrically energized heater ribbon wrapped around sorbent tube 10;
Other means of heating will be apparent to those skilled in the art. Mass spectrometer 1
4 is evacuated by a vacuum pump 17. Ionizer 15
can also be placed in a chamber that is evacuated by another vacuum pump (not shown) in a different pump arrangement. The desorption time td is the period required to desorb all trace components adsorbed during the sampling time. This period can be controlled within limits by varying the rate of heat flow to the sorbent tube 10. More rapid heating yields less td and therefore equation (
As explained in 5), a high gain factor G is produced. The period during which the trace components enter the ionizer 15 is independent of the desorption gas flow rate. For example, using a slow desorption gas volumetric flow rate at a given pressure results in a high number density (or partial pressure or fractional concentration) of trace components within the desorption gas stream because they are This is because it is distributed over a short distance, resulting in a large signal. However, this short length of the trace components in the gas flow is due to the slow linear flow rate of the valve V1 within the same length of time td.
accepted through. Nevertheless, the above volumetric flow rates from the sorbent 10 to the inlet of the mass spectrometer 14 with valve V1 are:
It does not determine the period required to transport detached specimens. The linear velocity of the carrier gas, which determines the transport time, is proportional to the volumetric flow rate. Under low pressure desorption conditions, high linear velocity and low mass flow rate occur simultaneously. The response of the mass spectrometer 14 to the desorbed analyte shown in FIG.
is influenced by the pumping speed of the aperture 16 separating the . For example, consider a case where a desorbed sample enters the ionizer 15 and immediately diffuses throughout the ionizer 15. The analyte concentration fi(t) in the ionizer 15 increases exponentially (0<t<td) during the period when the analyte enters the ionizer: [Equation 13] Here, fd is the desorption gas is the analyte concentration in the stream; and τ is the time constant characteristic of the ionizer. Once the slug flow of gas containing the analyte passes the inlet of valve V1 (t>td), the concentration in the ionizer increases exponentially with the same time constant τ from its peak value fi (td). The time constant τ in equations (13) and (14) is the ratio of the ionizer volume Vi to the pumping speed Si (in units of volume per unit time) leaving the ionizer 15, and thus : τ=Vi/Si (1
5) In FIG. 1, the ionizer 15 is pumped only through an opening 16 that separates the ionizer 15 from the mass spectrometer 14. If the ionizer time constant τ is small compared to the desorption time td, the analyte concentration in the ionizer 15 rapidly approaches and follows the analyte concentration in the carrier gas fd. Conversely, if the time constant τ is much larger than the desorption time td, the analyte concentration in the ionizer 15 will rise to only a fraction of the concentration in the carrier gas during the period in which the analyte enters the ionizer 15. The concentration then decreases with the characteristic time constant τ according to equation (14). Ionizers used for electron impact (EI) and chemical ionization (CI) usually have a small volume and a sufficiently large aperture so that their time constant τ is smaller than the typical desorption time (2-3 seconds) and 1
The arrangement shown in is sufficient. However, some discharge ionization sources require large volumes and moderately high pressures for limited recombination of their plasmas. In such cases, the arrangement shown in FIG. 2 is preferred, where all of the desorption gas is directed through the ionizer and the gas flow rate through this ionizer 15 is increased. A suitable valving sequence for sampling air or other gases and concentrating trace elements for subsequent analysis by a mass spectrometer is set forth in Table I for FIG. [Table 1]
Table I Valves A, B, C, D
and E are on/off valves. Using "1" to indicate an open valve and "0" to indicate a closed valve, the valve adjustment sequence is as follows: Valve
sorbent
phase
A B C D E Heater 12 Process
1. 0 1 0 0 0
Off initial pump down
2. 1 1 0 0 0
Off Air sampling 3. 0 1 1 0 0
Off Decompression
4. 0 0
0 1 1 off
Establishment of desorption gas flow 5. 0
0 0 1 1 off
Desorption of Analyte In the arrangement of FIG. 2, valves A, B, C, D and E operate as in FIG. 1, and valve V1
and V2 are adjusted to produce the desired desorption mass flow as well as the appropriate pressure within the ionizer. In either arrangement (FIGS. 1 and 2), the low pressure desorption gas need not be the same as the initial gas entering the preconcentrator. In fact, it is often desirable to use a low pressure desorption gas that is chemically different from the gas being sampled. For example, trace components in some air samples may react with oxygen at elevated temperatures. If air is also used as the desorption gas, the desorbed analyte may react with oxygen in the heated air during the desorption process and create different species. An inert gas such as argon or a fully oxidized gas such as carbon dioxide may be desirable in such cases. Conversely, a reactive carrier gas is used to detect derivatives of the initial analyte, or to identify compounds with similar mass spectra based on reactivity or reaction pathway with the desorption gas, or to identify more easily ionized species. It is conveniently used to create. Furthermore, when chemical ionization is used as the ionization method, the chemical ionization reagent gas may become the desorption gas. The use of low pressure preheated desorption gas, preheated by preheater 22, provides heat for raising the temperature of the adsorbent in sorbent tube 10, thus shortening the period during which desorption occurs therefrom. or reduce the heating heat amount of the heater 12. The heat capacity of the low pressure desorption gas is typically much smaller than the heat capacity of the sorbent material within the sorbent tube 10. Nevertheless, due to the low thermal conductivity of many sorbent materials, the preheated desorption gas becomes thermally active, caused by the high temperature desorption gas that is concentrated on their surface from molecular desorption; Provides support for a more effective desorption process than expected. From the arrangements of FIGS. 1 and 2, it can be seen that the desorption mass flow is large enough and the capacity of the mass spectrometer 14 or other low pressure detector is such that only a small portion of the desorption gas can be admitted through the analysis aperture 16. Assuming, the gas handling capacity of the mass spectrometer or other detector becomes the limiting characteristic. If,
If the mass flow of desorption gas is low enough and the pumping capacity of the mass spectrometer vacuum pump is high, the entire desorption gas flow will flow into mass spectrometer 14 through opening 16 and ionizer 15 . In such a case, valve V2 may be closed or replaced with a different arrangement as shown in FIG. In the arrangement shown in FIG. 3, valves A, B,
The opening and closing order of C, D and E is the same as for FIGS. 1 and 2. Vacuum pump 17 is a high speed pump (typically several hundred liters per second) that maintains the mass spectrometer pressure within the range of typically 10-5 torr. The difference here is the total desorbed gas and therefore the total trace components adsorbed by the sorbent that enters the ionizer 15 of the mass spectrometer 14 and then accepted into the mass spectrometer 14 itself through its opening 16. The actual mass flow rate entering a typical mass spectrometer is 10-3 torr/l/sec or about 0.1 atm.
-cc/min, and the same mass flow rate of desorption gas is therefore the maximum amount that can pass through the sorbent 10. In this case, the pressure provided by the adjustable valve V1 and the low pressure desorption gas source controls the desorption gas pressure and the actual mass flow. For very low desorption pressures as well as very low desorption gas mass flows, valve V1 should either be wide open or removed as seen in FIG. If an electron impact (EI) or chemical ionization (CI) ionizer 15 is used, both of which normally operate at pressures below 1 torr, if the minor components are present in the conduit between the sorbent 10 and the ionizer 15, 4 is acceptable if there is no tendency to stick to the inner wall of the valve E or to the inner wall of the valve E. However, if trace components tend to become stuck to these surfaces, the arrangement of Figure 3 is preferred because the low pressure gas source and valve V1 maintain a high pressure between the sorbent and valve V1. It can be set to
This is because it inhibits the diffusion of trace components into the conduit. According to equation (5), the relative gain in concentration is due to the ratio of the mass flow during the sampling phase to the mass flow during the desorption phase. Diffusion losses and the flow requirements of the detection method ultimately determine the minimum acceptable desorption mass flow. Therefore, the mass flow during the sampling period Qs must be as large as possible in order to maximize the gain. The absolute mass flow during the sampling period is determined by the factor K in equation (6), which is the preconcentrator mass flow Q for the difference in the squares of the pressures at the ends of the preconcentrator. This factor obviously depends on the geometry of the length and internal diameter in the case of tubes. To obtain high mass flow during sampling, the preconcentrator tube must be short. However, the use of short tubes leads to the possibility that the adsorption sites on the sorbent are completely filled during the sampling phase and that the sorbent no longer adsorbs all trace components in the sampling phase. In such cases, the sorbent is saturated and trace components in the sampling gas that then flows through the sorbent are within the analyte concentration during the desorption phase, even though they are within the sampling mass flow and sampling time ts. is not reflected. One solution is to reduce the sampling time until G becomes proportional to Ts. Alternatively, if one strives to keep G high, the conduction of thermal energy from the preheater 12 or 22 to the center of the tube is affected by the geometric changes, so there is an acceptable loss of increase in the desorption time td. With this, the preconcentrator can be replaced with one of the larger geometries (length and diameter in the case of a tubular preconcentrator). A major problem with commonly used preconcentration techniques is that they rely primarily, and often exclusively, on the ratio of sampling time to desorption time. Since some period of time is required for measurements on the preconcentrated sample to determine the minimum desorption time, the length of time required to achieve a particular preconcentration factor must be relatively large. It turns out. Mathematically this is shown as ts = Gtd. For example, consider that multiple scans over a range of several hundred Daltons require approximately 0.1 seconds. It usually takes about 2 to evacuate the adsorbed gas through a small sorbent tube.
Since the time required is seconds, the mass spectrometer described above can perform multiple scans over the time the desorbed gas is present. With a conventional preconcentrator and a concentration gain factor of 120, the sampling time ts must be approximately 120 x 2 seconds = 240 seconds = 4 minutes. In contrast, for the present invention, using a desorption pressure of 40 torr and a desorption pressure drop of 100 torr (for the calculations above), and for a desorption time td = 2 seconds, to achieve the desired gain factor of 120 Equation (11) shows that the sampling time is only 7.5 seconds, so the total measurement time is (sampling time plus desorption time plus 1 or 2 seconds of valve switching): Instead of the total measurement time of 4 minutes in the case of a conventional preconcentrator, it is approximately 10 to 12 seconds, ignoring the pressure transition time after the valve opens and closes. The total length of time required to analyze a given sample is determined by the changes that occur when the gas flows required for the process are modified by starting, stopping, or reversing, or by changing their pressure or velocity. It is important that In general, when a large amount of gas is discharged through a tube, a viscous flow occurs and the internal volume pressure as a function of time P(t) is theoretically given by: where where Po is the initial pressure, a is the inner diameter of the tube, L is the length of the tube, V is the volume of the gas, and η is the viscosity coefficient of the gas. Diameter 1/8" (a~0.15cm)
Atmospheric pressure (Po = 760t) is passed through a 30cm long tube.
orr~1×106 dynes/cm2) to air (η
~10-4 poise), 100c
The time to pump down from a volume of m3 to 0.01 atm is: [Formula 17] This calculation shows that the pump down and let up times are
must be on the order of seconds or less, demonstrating the need to keep unnecessary volumes small and conduit lengths short. One of the most time-consuming aspects of the measurement cycle is the cool-down period after the trace components have been thermally desorbed. Forced cooling is accomplished by flowing cooled gas through the sorbent tubes while blowing cooling air along the exterior surface or circulating cooling liquid around the tubes. However, it is well known to those skilled in the art that other means such as Peltier cooling can also be used. However, whatever means are used, they are time consuming and complicate the structure of the device. Another approach is to install two or more preconcentrators in parallel and switch between them. One is used while the other is cooled. An apparatus for doing this is shown in FIG. 1a. Valves F and F' allow one preconcentrator to operate while the other is cooled, and vice versa. Some equipment is common, such as low pressure gas supplies 11 and 11', vacuum pumps 18 and 18' and high pressure intakes. Separation valves V1 and V1' and V2 and V2'
are both present, valves F and F' can be omitted, and valves E and E' alternately switch the flow from preconcentrators I and II to mass spectrometer 14 via opening 16 and ionizer 15. Can be used for. The concept of the preconcentrator according to the invention was developed in 1990.
In April 2016, Extrel Corporation, Pittburg, Pennsylvania,
onin (Pittsburgh, PA) mass spectrometer. The test apparatus was similar to that shown in FIG. 1, in which the preconcentrator was of a sorbent tube configuration as shown in FIG. The sorbent tube 30 used here is a conventional microhematocrit capillary (Fischer Scientific Company, Pittsburgh, PA, Fischer Sci).
entific Co. ,Pittsburgh,P.
A. Catalog No. 02-668-68), length 37m
m, internal diameter approximately 1.5 mm and wall thickness approximately 0.1 mm. The extremely thin walls of this tube allow rapid heating and cooling of the sorbent 31. The capillary tube was made of mesh 60/80 Chromasorb with a length of 1 cm.
) 106 sorbent (trademark Manville, Deirfield, IL, Institute of Applied Science, purchased by Alltech Associates, Inc.; TM.Manville.purchase
d from Alltech Associa
tes, Inc. , Applied Science
Labs, Deerfield, IL), the sorbent has a length of 3 mm on each end thereof;
It is held in place by a piece of tightly wrapped silanized glass wool 32. A nichrome ribbon (not shown) wrapped around the sorbent tube serves as a resistance heater 12 to desorb the adsorbed analyte. A J-shaped thermocouple joint spot welded to the nichrome ribbon measures the temperature of the ribbon and is connected to a temperature control device (Fenwall, Model 550, Model 550).
550, Fenwall) adjusted its temperature during the desorption period. The sorbent tube is sealed to the rest of the apparatus by a 1/16" Swagelok fitting 34 (trademark Crawford Fitting Co.) and a Teflon ferrule. As shown in FIG. The remaining piping is 1/8" tubing (stainless steel and Teflon), 1/8" tubing (stainless steel and Teflon),
8" Swagelok fittings, Teflon solenoid valves (General Valve Corporation, Fairfield, N.J., Part 2-10-900 and Balcoen Engineering Corporation, Springfield, N.J., Model 20-1-3; P
art2-10-900, General Valv
e Corporation, Fairfield,
NJ; and Model20-1-3, Valco
r Engineering Corp. , Spr.
ingfield NJ), and stainless steel needle valve (Hoytei, Model SS22RS2, Model
s SS22RS2 Whitey). Hitachi 160VP Cuteback (174L/min) is shown in Figure 1.
1, and the pressure of the gas at each end of the sorbent tube 10 of FIG.
Baratron form of S Instruments, Parakeet
122AA-01000DB; Baratron t
ype122AA-01000DB, MKS Ins
truments, Inc. , Burlington,
MA). This manometer is communicated to the preconcentrator via a stainless steel toggle valve (not shown in Figures 1-4), which generally remains closed during the complete preconcentration process to avoid overvolume within the preconcentrator. be. Pressure measurements are usually carried out only during preparation for a particular preconcentration procedure. Low pressure gas supply for the desorption period is provided by two needle valves connected in series between the ambient air and the vacuum pump 18. The area between the two needle valves is connected to valve D, which is adjusted to produce the desired pressure with the sorbent. Sufficient gas flow through the needle valve is maintained to ensure that gas flow through the sorbent tube 10 does not affect the pressure on the sorbent. This arrangement is similar to a stiff voltage divider in electronic circuits. Vacuum regulator (e.g. Matheson
Model 3491) consists of other means that may be used. The mass spectrometer 14 used in these experiments was a single, differentially pumped, four-stage instrument (Spectrum
EL, Pittsburgh, Pennsylvania, Extrel Corporation;
The first vacuum chamber of the mass spectrometer houses the ion source and is 450 L
/S turbomolecular pump, the pumping rate of which was limited to approximately 150 L/S by the connection to the chamber. Pump operated in rear vacuum chamber, 3
The 20L/S diffusion pump houses a 4-stage mass filter. All of the experiments described below were performed using an electron impact ion source operating at 70 eV electrons. External timer (Chron
Trol, Lindbergh Enterprises, Inc., San Diego, California. ;LindburgE
enterprises, Inc. , San Diego,
CA. ) controlled the operation of the solenoid valve in the preconcentrator. The timing procedure used to investigate the operation of the preconcentrator described above is described in Table II below, using a "1" to indicate a valve that is open again and a "0" to indicate a closed valve. do. [Table 2]
Table II
valve sorbent
Phase A B C D E
heater process
Period 1. 0 1 0
0 0 Off Initial pump down 2 seconds 2. 1 1 0
0 0 Off Air sampling 3 seconds 3. 0 1 1
0 0 Off Decompression 2 seconds4. 0
0 0 1 1 off
Establish gas flow 3 seconds5.
0 0 0 1 1
Off Sample desorption 10 seconds

Total 20 seconds In this sequence, the sampling time was 3 seconds, and the total cycle time (excluding the time required for sorbent cooling after desorption) was 20 seconds. Experiments using this timing sequence were conducted with air spiked with various amounts of carbon tetrachloride (CCl4). Figure 6 shows the mass spectrometer signal recorded as a function of time during a preconcentration cycle, where air was sampled containing a total CCl4 concentration of 3.4 parts per million (ppm). Valve E (Figure 1) to the mass spectrometer was opened for about 7 seconds into the scan, at which time a slight increase in the signal was seen. The sorbent 10 is heated for about 10 seconds during scanning, and the adsorbed analyte is rapidly desorbed. Data recorded at masses 117, 119, 121 and 123 Daltons show that the main peak in the mass spectrum of CCl4 shows a significant response shortly thereafter, and the desorption shape is approximately 2 seconds in duration (1/2 (measured as full width in intensity). For comparison, the reaction at 140 Daltons shows only a slight response with background mass not associated with CCl4. The mass peaks at 119, 121 and 123 Daltons correspond to fragment ions of CCl4 molecules containing one or more 37Cl atoms. Based on the isotopic abundance of 37Cl, a signal at 123 Daltons effectively arises from the analyte present at a concentration of about 50 parts per billion. The desorption profile shown above was recorded using low pressure and low flow desorption conditions to increase the concentration of desorbed analyte in the carrier gas. These desorption profiles indicate that the concentration of CCl4 in the desorption phase is approximately 550 times its concentration in the air sample. Since the sampling and desorption times are approximately equal in this case (3 seconds and 2 seconds, respectively), the gain in concentration arises almost entirely from the difference in pressure and flow rate during the sampling and desorption phases. This gain is measured by comparing the peak intensity of the desorption shape to the signal generated by introducing a 3.4 ppm sample directly into the mass spectrometer. Each signal is corrected to the background level created by a blank air sample, and a unique valve is used to ensure that the mass spectrometer is
The pressure is adjusted to 0-5 torr. Since the total pressure within the mass spectrometer is the same for the two measurements, the comparison of the background corrected signals is a direct comparison of relative concentrations. The observed gain in this example is within two factors of the gain G, calculated from a comparison of desorption pressure, flow rate and time, as shown in Table III below: 3]
Table III
sampling
Detachment
Average pressure <P> 369torr
18.5torr
Pressure difference ΔP 668torr
21 torr
time
3 seconds 2 seconds (full width at 1/2 intensity) Calculated gain = 950 In this case, the difference between the expected gain and the observed gain is largely due to the saturation of the sorbent 10 as described above. . Figure 7 shows the peak ion signal during the desorption period as a function of CCl4 concentration for each monitor ion mass, with +/- error bars indicating saturation results. Each CC
The desorbed signal for l4 isotope fragmentation does not increase linearly with sample concentration. Finally, equation (9) predicts that the gain G from which the ionic strength during the desorption phase is inversely proportional to the average pressure in the sorbent tube during the desorption period. In FIG. 8, the dependence of the peak intensity at mass 69 on desorption pressure for a sample of perfluorotributylamine in air is shown. The signal was measured as a function of the average pressure during the desorption period for constant sampling conditions (sample concentration, mass flow and time) and a constant 30 torr pressure difference across the sorbent during the desorption period. The above ion signal increases with reduced pressure as expected (increase 1/<P>d), but the signal increases by 1/<P>d
does not increase linearly with respect to This discrepancy is due, at least in part, to the fact that the shape of the observed transient desorption peak is pressure dependent. This is because the observed peak width that increased with increasing pressure probably heats up slowly at higher pressures and flow rates, contributing to the nonlinear response of the signal to 1/<P>d. Having described preferred embodiments of the invention, it should be understood that other adaptations and variations may be made within the scope of the appended claims.

[Brief explanation of the drawing]

FIG. 1 shows a preferred embodiment of the invention, a tube containing a sorbent material that coldly adsorbs trace components and hotly releases them into the gas phase, low pressure detection shown as a mass spectrometer. 1 shows in schematic form a preconcentrator consisting of a vessel;

FIG. 2 is similar to FIG. 1 except that two preconcentrator arrays are joined to one mass spectrometer;

FIG. 3 shows an alternative arrangement of a preconcentrator in combination with a mass spectrometer according to the invention;

FIG. 4 is another schematic arrangement of a preconcentrator in combination with a mass spectrometer, in which all desorbed concentrated components are received by the mass spectrometer;

FIG. 5 is yet another schematic representation of the invention in which electron bombardment or chemical ionization occurs within the ionizer and the desorption pressure is matched to the ionizer pressure;

FIG. 6 is a detailed cross-sectional view of a sorbent tube as used in the present invention;

FIG. 7 is a graph of the desorption distribution of different isotope fragment ions from airborne trace carbon tetrachloride;

FIG. 8 shows peak isotopic fragment ion signals during the desorption period as a function of airborne carbon tetrachloride; and

FIG. 9 shows the dependence of the peak intensity at mass 69 on desorption pressure for a perfluorotributylamine sample in air.

[Explanation of symbols]

10, 10' Sorbent tube 11 Low pressure gas source 12, 12' Heater 14 Mass spectrometer 15, 15' Ionizer 16 Opening 17 Vacuum pump 18, 18' Vacuum pump 19 Opening 22 Heater, i.e. preheater 25 Ionizer 30 Adsorbent tube 31 Sorptive agent 32 Glass wool 34 Joint

Claims (25)

[Claims] 1. A method for preconcentrating trace components in a gas for continuous analysis, comprising: A. adsorbing at least one of the minor components on the confined sorbent material from a sample gas moving at a predetermined gaseous mass flow rate for a predetermined period of time; B. B. confining said sorbent material within a sealed space that includes an inlet of a detection device that receives said one trace component; and B. desorbing a trace element from the sorbent material for a predetermined period of time into a carrier gas moving at a predetermined gaseous mass flow rate that is substantially less than the gaseous mass flow rate of the sample gas; As a result, the amount of the one trace component desorbed from the carrier gas containing the desorbed component is less than the same amount of the one trace component relative to the portion of the sample gas containing the trace component. There are also substantially more steps involved. 2. The method of claim 1, wherein said carrier gas containing said one trace component is introduced into said detection device through said inlet. 3. The detection device includes a mass spectrometer.
The method according to claim 2. 4. The method of claim 1, wherein the carrier gas pressure is substantially lower than the sample gas pressure. 5. The method of claim 4, wherein the sample gas is at or near atmospheric pressure. 6. The method according to claim 1, wherein the sample gas is air. 7. The method of claim 1, wherein the sorbent material consists of particles of solid material confined within a glass tube. 8. The method of claim 1, wherein said one minor component is desorbed from said sorbent material by heating said sorbent material. 9. The method of claim 8, wherein said carrier gas is preheated to facilitate desorption of said one minor component from said sorbent material. 10. The method of claim 1, wherein the carrier gas is air. 11. The method of claim 1, wherein said carrier gas does not chemically react with said one minor component. 12. A derivative of said one minor component is created wherein said carrier gas chemically reacts with said one minor component to facilitate its detection by said detection device.
The method described in. 13. A plurality of entrapped sorbent materials are provided, each parallel to a sealing device, such that one or more of the plurality of sorbent materials are present for analysis by the detection device. 2. The method of claim 1, wherein the method of claim 1 can be selectively used to accept the sample gas and desorb the carrier gas. 14. The method of claim 3, wherein the mass spectrometer includes an electron impact ionization ion source operatively coupled to the inlet. 15. Claim 3, wherein the mass spectrometer includes a chemical ionization ion source operatively coupled to the inlet.
The method described in. 16. The method of claim 15, wherein the carrier gas is a reagent gas for chemical ionization. 17. The method of claim 3, wherein the mass spectrometer is operatively coupled to the inlet to operate at subatmospheric pressure. 18. A portion of the carrier gas that enters the gas emitting ion source is removed therefrom by a vacuum pump, and the remainder of the carrier gas enters the high vacuum of the mass spectrometer through an opening. 18. The method of claim 17, wherein the method is drawn. 19. An apparatus for preconcentrating trace components in a gas to facilitate analysis, the device being: held within a confined space and selectively adsorbing or adsorbing the trace components onto the sorbent material. the sorbent material associated with the desorption device;
a source of sample gas and a valve for directing the sample gas onto the sorbent material and a device for selectively opening and closing the valve; a source of carrier gas connected to the sorbent material to provide the carrier gas; a pumping device for selectively reducing the pressure within the confined space containing the sorbent material; a detection device for detecting the trace components in the carrier gas; a confined passageway leading to the detection device so that when the sorbent material desorbs the trace components received from the sample gas, the trace components are transported by the carrier gas to a pressure greater than the pressure of the sample gas; is conveyed to the detection device at substantially reduced pressure such that the ratio of the density of the trace component to the mass of the carrier gas is substantially lower than the ratio of the density of the trace component to a similar mass of the sample gas. Includes, which has been increased. 20. The apparatus of claim 19, wherein the detection device includes a mass spectrometer. 21. The apparatus of claim 20, wherein the mass spectrometer includes an electron impact ionization ion source. 22. The apparatus of claim 20, wherein the mass spectrometer includes a chemical ionization ion source. 23. The apparatus of claim 22, wherein said carrier gas is a reagent gas that reacts with said trace component. 24. Claim 20, wherein the mass spectrometer includes a gas-emitting ion source operating at a pressure below atmospheric pressure.
The device described in. 25. A portion of the carrier gas entering the gas emitting ion source is removed therefrom by a vacuum pump, and the remainder is removed through an aperture to the high vacuum of the mass spectrometer. 25. The device according to item 24.