Note: Descriptions are shown in the official language in which they were submitted.
2~661
Process and device for detecting measured substances in an
ambient subs~ancer es~ecially_for detectin~ gaseous warfare
a~ents in ambient air
The invention concerns a process for detecting measured
substances in an ambient substance, especially for detecting
gaseous warfare agents in ambient air, in which light
reaction ions are first generated from a reaction substance
and are added to the mixture of measured substanae and
ambient sub~tance, in such a way that reaction ions in a
spatially inhomogeneous distribution in a measurement
chamber attach themselves to heavy molecules of the measured
substance to ~orm molecular ions; and that in the me
asurement chamber, an electric field that varies over time,
.. ~
- 2 - 20~5 6G ~
alternates about a zero line, and has a predefined basic
frequency an~ amplitude, is generated, the molecular ion
c:urrent resulting from the electric field is measured, and
the measured signal is analyzed.
The invention further concerns a device for detecting
measured substances in an ambient substance, especially ~or
detecting gaseous warfare agents in ambient air, with a
measurement chamber, with first means for introducing the
mixture of measured substance and ambient substance into the
measurement chamber; with second means for generating light
reaction ions from a reaction substance and for adding the
reaction ions to the mixture; with third means for attaching
the reaction ions to heavy molecules of the measured
substance in the measurement chamber; with fourth means for
generating in the measurement chamber an electric field
which alternates a bout a zero line and has a predefined
basic frequency and amplitude, with the fourth means being
configured as an electrode pair in the measurement chamber
and as a voltage source connected to the electrode pair; and
with fifth means for detecting and analyzing the molecular
ion current caused by the electric field in the chamber.
A process and a device of the aforesaid type are known from
the article entitled "Detection of Chemical Warfare Agents
by Means of an Ionization Chamber Operated in AC Mode," by
S. Milinkovic et al., in "Proc. 3rd Int. Symp. Protection
Against Chemical Warfare Agents", Umea, Sweden, June 11-16,
1989.
In many areas of environmental protection and military
engineering, it is desirable to be able to detect certain
measured substances, preferably hazardous measured
2Q~66~
substances, in an environment consisting of air, water, or
the like. If extremely dangerous measured substances, for
example gaseous warfare agents, are to be detected in this
context, not only are high detection sensitivity and
reliability desirable, but stringent demands are also made
in terms of response speed. This is especially true when it
is not known if the measured substances in question are
present in the relevant environment at all, since constant
monitoring for the occurrence of these measured substances
then takes place, and an alarm must be triggered as quickly
as possible if the occurrence of these measured substances
is observed.
~n the area of military engineering, this applies especially
to the area of toxic gases, which are known in a variety of
compositions and effects.
For example, toxic gases are essentially differentiated into
vesicants on the one hand, and nerve poisons on the other.
While vesicants act on the human skin, where they cause
erosion and burning, nerve poisons act on the human nervous
system, leading to respiratory paralysis and the like.
The best known of the vesicants are the mustard gases, in
this case especially the 2,2'-dichlorodiethyl sulfide known
as sulfur mustard, with the structural formula:
Cl -CH2 -CH2 -S-CH2 -CH2 -Cl
and the tris(2-chloroethyl) amine with the structural
formula:
CH2-CH2-Cl
~ N ~
1 CH2 CH2 CH2-CH2-Cl
206a661
or:
Cl3
Cl-cH2-cH2-N-cH2 -CH2-Cl
Among the nerve poisons, the most important are the
dimethylphosphoramidocyanidic acid ethyl ester known as
Tabun, with a molecular weight of 162 amu and the following
structural formula:
(cH3)2N-~-cN
1-C2H5
and the 0-isopropyl or -pinacolyl ester of methyl-
phosphonofluoridic acid known as Sarin or Soman, with
molecular weights between 140 and 182 amU and the following
structural formula:
CH o
13 ll
R-C-0-P-F R = -CH R = -C (CH3)3
H CH3
Last to be mentioned of the gaseous warfare agents are the
phosphorylthiocholines known ~enerally as V agents, with
molecular weights of 267 amU and the following structural
formula:
o
R0 - ~ - SCH2CH2NR2
' ' ,
~' .
~ 5 - 206~6~1
especially "VX", with an even higher molecular weight and
the following structural formula:
2H5 o ~ - SCH2CH2N (iC3H7)2
CH3
A variety of processes and devices for detecting these
chemical warfare agents are already known. Some of these
known processes and devices are based on the detection of
mobile ions.
For example, the article by Milinkovic mentioned earlier
describes a coaxial ionization chamber in which, in a
cylindrical housing, a radial inlet tube and a radial outlet
tube opposite it are provided for a gas mixture. For this
purpose, gas is drawn in from the environment and sucked in
through the inlet tube into the measurement chamber, being
then discharged from it through the outlet tube once the
measurement has been made.
The known measurement chamber uses a cylindrical electrode
arrangement with an axial inner electrode and an outer
electrode arranged on the cylindrical enveloping surface,
with the gas flow under investigation flowing through the
gap between the two electrodes. A radiation source that also
influences tbe volume through which the gas flows is located
at the axial end of the measurement chamber in a radial
plane.
~.
.
2~$~6 1
- 6 -
In the known device, ambient air is then pumped through the
measurement chamber at a constant flow rate, and ionized by
the radioactive source. The result of this is to form
"reaction ions" from the ambient air t2, N2) in combination
with atmospheric moisture (H20). Specifically, this occurs
as follows:
First the ambient air (2' N2) is ionized under the action
of B radiation, forming 2+ and N2+ ions and free electrons
e .
If we now consider the N2+ ions, for example, these first
bind to a nitrogen molecule 2N2 to form N4+ and N2. The N4+
in turn reacts with atmospheric moisture (H20) to form 2N2
and H20+. The H20+ in turn reacts with the H20 in the
atmospheric moisture to form H30+ and OH, which corresponds
to a reaction between (H20)H+ and H20 to form the positive
reaction ions (H20)8H+. This positive reaction ion has a
weight of 145 amU and a mobility of 2 cm2/Vs.
Similarly, it can be shown for the negative reaction ions
that the 2 and the free electrons e ultimately react with
atmospheric H20 to form (H20)602, the negative reaction ion.
~his has a weight of about 140 amU and a mobility also on
the order of 2 cm2/Vs.
The positive and negative reaction ions explained above now
react with the molecules of any gaseous warfare agents that
may be present. I~ the molecules of the nerve poisons
(organophosphorus compounds) are referred to as MA and the
molecules of the vesicants, especially sul~ur mustard, as
MB, the following reactions can then be written:
(H20)8H + MA ~ MA (H2~)6H + 2H20
MA(H20)6H + MA ~ MA2H + 6H20
- 7 ~
with the the formation of MA3H+ also possihle at higher
concentrations of organophosphorus compounds.
With the vesicants, the corresponding reaction is as
follows:
(H2o)6o2 + MB ~~~~ ~ (H2)402 + 2H20
The quasi-molecular ions or product ions formed in this
manner have, in the case of the organophosphorus compounds,
a mobility on the order of 1.5 to 0.5 cm2~Vs at a molecular
weight in the range between 250 and 700 amu, while for the
vesicants (sulfur mustard), the ion mobility i5
approximately 1.5 cm2/Vs with a molecular weight of 250 amu.
The above explanation shows the expected result, namely that
the reaction ions are considerably lighter in weight and
faster in terms of mobility than the product ions or quasi-
molecular ions, with considerably higher molecular weights
along with lower mobility.
For the sake of simplicity, the abbreviations "~+" and "2"
for the reaction ions, and the abbreviations "MA+" and "MB-"
for the quasi-molecular ions, will be used in the
description which follows.
~f the gas flowing through the measurement chamber now also
contains measured substances, especially gaseous warfare
agents, that have molecules with a considerably higher
molecular weight, the aforementioned reaction ions will then
attach themselves to the molecules of the measured
substances.
- 8 - 2Q5~661
The type of attachment - i.e. whether the positively charged
protons or the negatively charged oxygen ions attach to the
molecules of the measured substances - depends on the nature
of the measured substances.
In the case of the gaseous ~arfare agents explained above,
conditions are such that with the mustard gases, the
negatively charged reaction ions attach to the mustard
molecules, while with the nerve poisons (Tabun, Sarin,
Soman, V agents, and especially VX), a proton transfer
occurs from the positively charged reaction ions to the
warfare agent molecules. In the case of the latter reaction,
it is also possible that in each case two molecules of the
measured substance will bind a proton in pairs.
It is especially important in the present connection that
with both types of gaseous warfare agents, namely with the
vesicants (mustards) on the one hand and the nerve poisons
(Tabun, Sarin, Soman, VX agents) on the other, differently
charged molecular ions, namely negatively charged quasi-
molecular ions for the vesicants and positively charged
quasi-molecular ions for the nerve poisons, are formed in
the respective cases.
It is also important, in the case of the known device, that
the radioactive source is arranged so that it produces a
spatially inhomogeneous distribution of reaction ions and
thus also of guasi-molecular ions in the measurement volume
of the measurement chamber. The inhomogeneity o~ the spatial
distribution is configured so that on average, the distance
~rom the quasi-molecular ions to one electrode is greater
than the distance to the other electrode.
9 2 Q ~
In the known device, an alternating current is then applied
'to the electrode arrangement. The resulting alternating
electric field, i.e. a field oscillating about a zero line,
then exerts a force on the quasi-molecular ions in the
direction of one electrode during one half-wave, and in the
direction of the other electrode during the other half-wave.
However, since the average path lengths for the quasi-
molecular ions to the two electrodes are, as mentioned, of
different lengths, a greater ion current flows during one
half-wave than during the other half-wave, since with an
electric field of a suitably high frequency, the half-period
is not sufficiently long to bring all the relatively heavy
quasi-molecular ions to the respective target electrode.
The result, with the known arrangement, is therefore a
direct-current component of the ion current. The sign of
this direct-current fraction depends on which type of
measured substance is present in the ambient air. If it is
one of the aforementioned vesicants with negatively charged
quasi-molecular ions, the siqn of the direct-current
component is positive, while with the nerve poisons it is
negative.
With the known device, the result is therefore a positive or
a negative output signal depending on the type of gaseous
warfare agent detected; and it is known that this output
signal can be further optimized in terms of amplitude by
setting the frequency of the electric field apprapriately.
The known device has the disadvantage, however, that it
consistently produces measurement errors or even fails if
what is present in the environment is not exclusively
measured substances of a sin~le type, but rather mixtures of
measured substances of both types, for example vesicants and
nerve poisons simultaneously.
10 -- 2 0 ~ ~) 6 6 1
On the other hand, it is of considerable advantage to be
able to have available processes and devices with which,
even in situations in which vesicants and nerve poisons are
being used simultaneously, both the one and the other
substance can be detected quickly and reliably, without
having the signals from one substances compensate for the
signals of the other substance.
A further problem that arises in the detection of measured
substances of the type discussed here is that detection of
these measured substances must be differentiated from
interfering substances and other interfering factors that
occur in a real-world environment.
If we once again consider the area of military engineering
in this context, it may for example be necessary in a
conflict situation to be able to detect gaseous warfare
agents selectively, even when the ambient air is at the same
time permeated by hydrocarbon vapors, for example leaking
gasoline and diesel fuel, or by smoke or other organic
compounds. In addition, detection must also be possible if
fluctuations in pressure, temperature, or atmospheric
humidity occur during the measurement. The interfering
factors just mentioned can occur, for example, if the
process is ~erformed on board a vehicle or if the device is
located on board a vehicle. If the vehicle is, for example,
a helicopter, a measurement device will also be exposed to
numerous interfering effects which must suppressed as much
as poss ible when detecting these highly sensitive measured
substances.
It is also a known process to use "tandem measurement
chambers" to suppress the influence of interfering
- 11 - 20~66~
substances. Such tandem measurement chambers have an
lmaltered measurement chamber, for example of the type
mentioned initially, and also have a second reference
chamber in the gas inlet of which is arranged a filter that
Eilters out the substances being measured. When measuring
gaseous warfare agents, such a filter can be, for example,
the usual filter for a gas mask.
When such a tandem measurement chamber is used in a
measurement situation in which, ~or example, gasoline vapors
are acting as an interEering factor, these gasoline vapors
are not retained by the reference chamber filter and
therefore permeate the measurement chamber and reference
chamber simultaneously. Consequently, the readings from the
two chambers change simultaneously, so that this type of
reading can be eliminated as an interfering variable. On the
other hand, if a gaseous warfare agent encounters the tandem
measurement chamber, it permeates only the measurement
chamber and not the reference chamber, since it is blocked
off from the latter by the filter, so that a true reading
occurs only in the measurement chamber.
In an alternative configuration, a filter can also be placed
in front of a measurement chamber and then removed from it,
so that one measurement chamber serves in succession as
measurement chamber and reference chamber.
The known arrangements just described have the disadvantage,
however, of being encumbered with a long time constant. For
example, in reality it is impossible to obtain filters that
ideally retain the measured substance and immediately allow
free passage of all other substances. In practice,
conditions are in fact such that a filter also presents at
least some resistance to interfering substances, for example
- 12 - 2~ 61
gasoline vapors, with the consequence that the reference
chamber is slow to fill up with the interfering substance,
while the latter has already penetrated into the measurement
chamber without hindrance. To eliminate incorrect
measurements, it is therefore always necessary to wait a
certain amount of time for equilibrium to occur between
measurement chamber and reference chamber. The times
required for this in practice are on the order of minutes;
but this period is too long for the detection of gaseous
warfare agents. What are instead desired are reaction times
on the order of, for example, 10 seconds, so that an alarm
can be triggered before the suddenly occurring gaseous
warfare agents have caused any damage.
The periodical "Isotopenpraxis," Volume 26, No. 4, pages
176-180 (1990), discloses a mathematical method for using a
Fortran program to find an iterative solution to
differential equation systems that are applicable to the
kinetics of various physical processes. These include, for
example, an ionization gas detector, in which ions present
in a measurement chamber are influenced by a pulsed electric
field.
The article "New Ion Mobility Techniques" by C. ~lanchard,
published in "Proceedings of the 1987 U.S. Army Scientific
Conference on Chemical Defense Research," Aberdeen Proving
Ground, MD, CRDBC-Sp-88013, 1971 (April lg88) discloses an
ion mobility spectrometer (IMS) in which non-linear electric
fields are applied to tbe mobile ions. These non-linear
electric fields are partly spatially and partly temporally
non-linear, and are designed to raise the sensitivity of the
spectrometer.
- 13 - ~ 05S~1
European Patent EP-A-O 253 155 discloses a field-regulating
arrangement for an ion mobility spectrometer (IMS) in which
a temporally non-linear electric field is also generated.
However, the emphasis with this IMS, as with the IMS explained
above, is on spatial separation of regions with differing field
strength profiles. The possibility of separate detection of
two measured substances with differently charged quasi-molecular
ions is neither discussed nor possible in this connection.
The object on which the invention is based is therefore that
of developing a process and a device of the aforementioned type
in such a way that even with mixtures of different measured
substances, especially with mixtures of vesicants and nerve
poisons, rapid and differentiated measurement, along with any
required alarm, is possible.
According to the invention, this object is achieved, in accord-
ance with a process of the aforementioned type, by the fact
that in order to differentiate a first measured substance from
a second measured substance present simultaneously in the ambient
substance,
- First, for a defined m~asurement chamber, the function
of the quasi-molecular ion current vs. basic frequency,
vs. amplitude, and vs. an asymmetry of the electric field
stren~th vs. time function with respect to the zero line
are determined, each time separately for the first and
for the second measured substanae;
- Then a first measurement of the mixture is performed with
a first asymmetry;
- 14 - 20 6~
-- Then a second measurement of the mixture is performed
with a second asymmetry; and
~- The measured signals of the two measurements are
logically correlated.
According to the invention, the object is further achieved,
in a device of the aforementioned type, by the fact that the
voltage source comprises a changeover switch by means of
which two voltage profiles with different asymmetries with
respect to the zero line can be switched alternatively and
successively to the electrode pair; and that the fifth means
comprise logical circuit means for comparing the measured
values of the quasi-molecular ion current.
The object underlying the invention is completely achieved
in this manner.
Specifically, because two measurements are performed in
succession with different measurement parameters, it is
possible to influence the measured substances in different
ways during the two measurements, so that with a suitabls
measurement chamber, the results are two readings from which
the presence of one of the two, or both, measured substances
can be determined.
According to this invention, therefore, there is no longer
any risk that compensation in readings will occur due to
differently polarized quasi-molecular ions.
Moreover, the defined measurement conditions, i.e. the
predefined settings for basic frequency, amplitude, and
asymmetry, ensurs that selectivity is high, thus
simultaneously allowing the suppression of interference
, ~ ,
- 15 - 20S~fil
E;ignals that might result from interfering substances in the
environment.
In preferred variants of the process according to the
invention, water is used in a known manner as the reaction
substance from which the reaction ions are generated by
radioactive irradiation. Preferably a ~ radiator in the
measurement chamber is used for this purpose, with its
inside dimensions being made considerably greater than the
half-life range of the B radiator so as to produce the
desired inhomogeneous distribution.
It is moreover preferred if a preferably periodic voltage is
used to generate the electric field. It is further
preferable if the voltage is given a profile over time
having first and second pulses alternating from one to
another above and below the zero line.
~he first pulses are, in this context, preferably followed
immediately in time by the second pulses.
In this connection, a particularly good effect is achieved
if the first pulses are made to have the same pulse area as
the second pulses.
Preferably the pulses are rectangular in shape, with the
first pulses being higher and narrower than the second
pulses; also preferably, the first pulses are made to be two
to five times as high as the second pulses.
~t is especially advantageous i~ the first and the second
asymmetry are produced by the fact that for the first
measurement the pùlses are given a first polarity, and for
,~
- 16 -
l:he second measurement are given the opposite polarity. This
can easily be achieved by simply changing the sign of the
voltage.
A basic frequency in the range between 100 and 500 Hz has
proved particularly successful for implementing the process
according to the invention in the detection of gaseous
warfare agents, with the volume of the measurement chamber
being set in the range between 0.5 and 5 cm3. The voltage
was given an effective value in the ran~e between 50 and
200 V.
In a further group of exemplary embodiments of the process
according to the invention, in a further, i.e. third,
process step, a constant electric field two to twenty times
as strong as the effective value of the alternating electric
field is generated in the measurement chamber; the resulting
quasi-molecular ion current is measured; and the measured
signal obtained thereby is compared with the measured
signals of the first and second measurements.
This feature has the advantage that this third comparative
measurement sucks up all the free charge carriers in the
measurement chamber due to the very high electrical field
strength, thus ma~ing available a signal that reflects the
total number of free charge carriers present in the
measurement chamber. Since this total number contributes to
the measured signal for both the first and the second
measurement, in which neither the concentration of the
measured substances nor the concentration o~ any interfering
substances that may be present are, however, known, it is
possible to determine these un~nown values on the basis of
the results of the three measurements.
- 17 -
This procedure has the additional advantage that interfering
environmental influences such as fluctuations in pressure,
temperature, and humidity are also eliminated, since if the
aforesaid three measurements succeed one another immediately
in time, these interfering effects contribute equally to all
three measurements, and are therefore eliminated from the
calculation when calculating the variables of interest, for
example the concentration of the measured substance.
In preferred developments of the device according to the
inventionl the measurement chamber is cylindrical in shape.
This feature has the advantage of resulting in a particularly
simple confi~uration that is both manageable in terms of
calculations and results in reproducible conditions. This is
important because in accordance with the process according
to the invention described above, a calibration of the
measurement chamber is performed first, for which it is
advantageous to use mathematical models that in turn require
calculable geometric conditions in the measurement chamber.
In a preferred development of the device according to the
invention, the measurement chamber is lined with a
radioactive material. This feature has the advantage, known
in the art, that if a suitable radioactive m~terial is
selected, for example Ni63 with an activity of lO mCi, a
sufficiently inhomogeneous distribution of reaction ions and
thus of quasi-molecular ions occurs in a cylindrical
measurement volume lO mm in diameter and between 10 and 30
mm long.
It is ~urthermore preferable if the walls of the measurement
chamber are made of a plastic, since this makes them
especially easy to produce.
2 ~
It is furthermore preferable if the electrode pair consists
of an axial rod electrode and an electrically conductive
lining of the cylindrical enveloping surface of the
measurement chamber.
This feature has the advantage of resulting in an electrode
arrangement that is easy to manage in terms of calculation.
The feature has the additional advantage that the
radioactive lining of the measurement chamber can
simultaneously be used as the outer electrode.
Lastly, it is preferred if the rod electrode has a
stainless-steel surface. This has the advantage of making
available an element for the rod electrode that is
electrically conductive but not chemically active.
Altogether, the invention thus has the advantage of allowing
very rapid detection of measured substances; for example,
the aforesaid vesicants or nerve poisons can be detected
with a measurement time of about 10 seconds, within a
typical concentration range of 10 to 100 ppb.
~he possibility of calibrating the measurement chamber on
the basis of mathematical models also has the advantage of
allowing optimization for certain measured substances
essentially without practical tests, eliminating the need
for hazardous and therefore complex laboratory experiments
with measurement chambers of different designs.
The invention can therefore advantageously be used in the
entire area of environmental engineering and military
engineering. It is described hereinafter, however, with
reference to the example of detecting gaseous warfare
agents, although this application in no way limits the
invention.
- 19 - 2~ fil
It als~ goes without saying that the features mentioned
]previously and those yet to be explained below can be used
not only in the particular combinations indicated, but also
in other combinations or in isolation, without leaving the
context of the present invention.
Exemplary embodiments of the invention are depicted in the
drawings and will be explained in more detail in the
description below. In the drawings,
ig, 1 is a perspective view, partly cut away, through a
measurement chamber as used in the context of the
present invention;
ig. 2 is a highly schematic block circuit diagram
explaining one exemplary embodiment of a device
according to the invention;
ig. 3 is a calibration diagram explaining one exemplary
embodiment of the process according to the
invention.
In Figure 1, 10 designates a cylindrical measurement chamber
that is essen~ially de~ined by a cylindrical casing 11, an
upper end wall 12, and a lower end wall 13. In this respect
the measurement chamber 10 is preferably made of a plastic,
for example polyethylene or polypropylene, and can be
produced as an injection-molded housing.
Loaated in the upper end wall 12 i8 an inlet tube 14 through
whi~h, as indicated by an arrow 15, a medium, for example
ambient air, can be drawn in.
~ .
. ~
.
2 0 ~
- 20 -
Iocated correspondingly in the lower end wall 13 is an
c,utlet tube 16 which is connected to a conduit 17. The
conduit 17 leads to a suction pump 18. The end walls 12, 13
are made of insulating material, preferably polyethylene or
E\lYpropylene.
Arranged along one axis 20 of the measurement chamber 10 is
a rod electrode 21 that extends through the upper end wall
12 and is preferably spaced a certain distance away from the
lower end wall 13. The rod electrode 21 preferably consists,
at least on its surface, of a chemically inactive material,
for example stainless steel. The rod electrode 21 is
connected to a first lead 22 that will be further explained
later on.
Located on the inside of the cylindrical casing 11 is a
lining 23 made of a radioactive material. The lining 23
preferably consists of a radioactive nickel isotope,
specifically Ni63. Since the lining 23 is electrically
conductive, it can simultaneously serve as the outer
electrode, i.e. as the electrode that cylindrically
surrounds the rod electrode 21 which acts as the inner
electrode. Por this purpose, the lining 23 is connected to a
second lead 24 that is guided through the cylindrical casing
11 and whose function will be further explained in detail
later on.
In one practical exemplary embodiment of the measurement
chamber 10, the latter has an inside diameter of 10 mm and
an axial length of 10 to 30 mm~ For measurement~ of gaseous
substances/ the suation pump 18 is set for a gas flow rate
of 10 to 30 liters/hour. The gas then flows through the
inlet tube 14 in~o the measurement chamber 10, as indicated
- 21 - ~3~
by arrows 26, and is extracted from the measurement chamber
:L0 through the outlet tube 16, as indicated by arrows 27.
With the geometry of the measurement chamber 10 as described
above, the lining 23 has, for example, an activity of 10
mCi.
A radially oriented electric field, indicated in Figure 1 by
arrows labeled E, forms when a voltage is applied via the
leads 22, 24 to the electrodes 21, 23.
Figure 2 once again shows the measurement chamber 10 on the
left side, in a highly schematic sectioned view.
Figure 2 also indicates the B radiation emerging from the
radioactive lining 23 at 30. Indicated at 31 with dot-dash
lines is a radial plot of the intensity of the B radiation
30, showing that the B radiation 30 has already decayed
sharply after a few millimeters, and in any event over a
distance that is considerably less than the radius of the
measurement chamber 10.
As a result, an inhomogeneous distribution of radiation
intensity is created in a radial plane of the measurement
chamber 10, since the intensity rapidly decreases from the
inner periphery of the measurement chamber 10, i.e. from the
radioactive lining 23, inwards towards the axis 20.
If a gas mixture, indicated in Figure 2 as 32, is then drawn
into the measurement chamber 10 through the inlet tube 14,
the following occurs (if the gas mixture 32 is of suitable
composition):
- 22 - 20SS~fi~
I.et it be assumed that the gas mixture 32 consists
e!ssentially of ambient air which nevertheless contains
certain fractions of two measured substances A and B. These
measured substances A and B may be, for example, gaseous
warfare agents, and in particular, measured substance A may
be a vesicant and measured substance B may be a nerve
poison.
The moisture (water) present in the ambient air is now
exposed to the B radiation 30 in the measurement chamber 10.
The ~ radiation 30 has an ionizing effect, resulting in the
formation, from the atmospheric moisture which is used as
the reaction substance, of reaction ions, namely negatively
charged oxygen ions ~2-) on the one hand and positively
charged protons (H+) on the other hand, as depicted
schematically in Figure 2.
Let is further be assumed that the molecules MA of the first
measured substance (gaseous vesicants) have an affinity for
the negatively charged oxygen ions (2-)' whereas the
molecules MB of the second measured substance (gaseous nerve
poisons) have an affinity for the positively charged
reaction ions (H+). The corresponding reaction ions
therefore attach themselves to the respective molecules,
resulting in electrically charged quasi-molecular ions MA-
and MB+.
It must be noted in this connection that the number of
reaction ions generated by ionization is very much greater
than the number of measured substance ions present in the
ambient substance, so that even after attachment to the
reaction ions, there is still a great excess of unattached
reaction ions.
- 23 - ~a~
Also worth noting is that the weight of the raaction ions is
considerably less, specifically about two to five times
less, than the molecular weight of the measured substance
ions.
once ionization and attachment have occurred, the
measurement chamber 10 thus contains high-mobility reaction
ions 2- and H+, and low-mobility quasi-molecular ions MA-
and MB+. The concentrations of these various charge carriers
are, however, initially unknown.
A three-channel voltage source 40 is then used to generate
the electric field E inside the measurement chamber 10. The
voltage source 40 is connected, in series with a measurement
resistor 41 and a third conductor 42, to the leads 22 and 24
mentioned earlier, resulting altogether in a closed circuit.
Specifically, when the voltage source 40 applies a voltage
to the electrodes 21, 23, the charge carriers located in the
measurement chamber are thereby set in motion, and an ion
current - detectable at the measurement resi stor 41 as a
voltage drop - flows.
A control unit 50, that is connected via a measurement
conductor ~1 to the measurement resistor 41 and via three
control conductors 52, 53, 54 to the three channels of the
voltage source 40, is provided in order to control the
measurement process.
The first channel 60 of the voltage source 40 is a first
pulse generator, the second channel 61 is a second pulse
generator, and the third channel 62 is a constant voltage
generator.
- 24 - 206.jfi6
A selector switch labeled 63 in the control unit 50 can now
]be used to switch the first, second, or third channel 60,
61, or 62 alternatively into the circuit.
It is self-evident in this connection that the circuit
elements already described, as well as the circuit elements
yet to be explained below, can in each case be configured as
separate electrical components (hardware), or as software in
the context of a control program.
Lastly, the control unit 50 is also connected to a
characteristic curve memory 64, which will be explained in
further detail below.
Time-varying voltage profiles, preferably having the shape
of periodic pulse sequences, are generated in the channels
60 and 61 of the voltage source 40 that are configured as
pulse generators. For example, as indicated in Figure 2, a
pulse sequence can alternate about a zero line 70, with a
first pulse 71 being a positive, narrow, high pulse, while a
second pulse 72 is a negative, broad, flat pulse. Preferably
the total pulse areas of the pulses 71 and ~2 are equal in
magnitude.
A similar pulse sequence 71', 72' with reference to the zero
line 70, in which the shape and polarity of the pulses are
simply transposed with respect to the pulses 71, 72, is
generated in the second channel 61.
In the pulses 71, 72 and 71', 72', the ratio between height
and width is preferably between two and five.
Lastly, a constant voltage 73, the magnitude of which is
several ti~as greater than the effective value of the pulse
- 25 - s
sequences 71, 72 or 71', 72', is generated in the third
channel 62. Preferably the amplitude of the constant voltage
73 is two to twenty times the effective value of the pulse
sequences, so when the constant voltage 73 is applied to the
electrodes 21, 23, an electric field strength of, for
example, lOoO V/cm is generated in the measurement chamber
10 .
A first measurement output 80 for measurements with the
first channel 60 and a second measurement output 81 for
results of measurements with the second channel 61 are
provided on the control unit 50. Since each measurement
output 80, 81 has two connections, a total of four
connections 82 to 85 are available. Two of these connections
82, 84 are inverted with inverters 86, 87. The four output
conductors that result are directed, in the manner
illustrated in Figure 2, to four AND gates 90 to ~3, with
indicators ~4 to 97 connected downstream from each of them.
The indicator 94 indicates the simultaneous presence of
nerve poisons (+) and vesicants ~-); the second indicator 95
indicates that only vesicants (-) were detected; the third
indicator 96 indicates the presence of nerve poisons (+)
only; while the fourth indicator signals that neither
vesicants nor nerve poisons were detected~
To explain the operation of the arrangement according to
Figure 2, first of all the family of characteristic curves
in Figure 3 will be described:
Figure 3 is a plot of calibration curves that were
determined using mathematical models for a defined geometry
of a measure~ent chamber 10 under defined boundary
conditions. In the characteristic curves in Figure 3, the
- 26 - ~ S;~
~uasi-molecular ion current~I is plotted on the ordinate,
while the effective value of the voltage applied to the
e~lectrodes 21, 23 is plotted on the abscissa.
A first characteristic curve in Figure 3 is labeled B. Curve
B applies to a predetermined existing concentration of
measured substance B tnerve poisons). Curve B indicates the
quasi-molecular ion current of quasi-molecular ions MB+ when
the first channel 60 of the voltage source 40 was switched
to the measurement chamber 10, i.e. when a voltage profile
with the pulses 71, 72 was in effect. It is evident that the
molecular ions M8+ f the first measured substance B
initially result, at low effective voltages U, in only a
small molecular ion current ~ I, which then rises to a
maximum, drops to zero as the effective voltage increases
further, and then decreases to a negative limit value.
Simi~arly, curve B' in Figure 3 shows the change in
molecular ion current under the same conditions, the only
difference being that this time the second channel 61 of the
voltage source 40 was switched on. Under these conditions,
the curve for the molecular ion current ~I as a function of
effective voltage U is different: specifically, the
molecular ion current ol remains constant, with a weakly
developed maximum.
Lastlyr A' depicts a third curve which plots the correlation
between the molecular ion current ~I for the molecular ions
MA- of the first measured substance A (vesicants) as a
function of the ef~ective voltage U. The shape of curve A'
is similar to that of curve B, although of opposite sign. It
is interesting that curves B and A' intersect approximately
at the zero point 100.
- 27 - 2 0 ~ ~ ~ fi 1
Curves B, B', and A~ - which, as mentioned, were determined
by theoretical or practical preliminary tests - are stored
in the characteristic curve memory 64 in the device in
Pigure 2, and influence the operation of the control unit
50.
At this point it is worth mentioning again that the diagram
in Figure 3 refers only to a specific configuration of the
measurement chamber 10, namely to a predefined geometry and
to predefined measurement parameters, for example including
a predefined frequency for the pulses 71, 72 and 71', 72'.
It goes without saying in this connection that a large
number o~ characteristic curve families like those in Figure
3 can be determined for various measurement chambers or
measurement conditions, and stored in each case in the
characteristic curve memory 64 for various practical
measurements.
We will now assume that all the measurement parameters that,
for example, served as the basis for the diagram in Figure
3, have been applied in the arrangement in Figure 2.
The effective voltage of the pulses 71, 72 and 71', 72' can
now be freely selected within the range of variation of the
diagram in Figure 3. It is advantageous if a measurement is
performed at the zero point 100, i.e. if, in the case of the
example depicted in Figure 3, the pulses 71, 72 or 71', 72'
are set so that the effective voltage is on the order of 110
V.
If the Eirst channel 60 of the voltage source 40 is then
switched into the circuit of ths measurement chamber 10 by
actuating the selector switch 63, any molecular ion current
~I will then generate a voltage at the measurement resistor
41 which will be transferred via tne measurement conductor
51 to the control unit 50.
- 28 - 2~
If no measured signal is observed during this test, this
means that no MA- molecular ions are present (since at this
operating point they would have resulted in a molecular ion
current); but the result does not also mean that no MB+
molecular ions are present, since the intersection of curve
B with the zero point 100 indicates that even if ~ +
molecular ions were present, no quasi-molecular ion current
~ I would be generated.
If, however, a measured signal occurs during the
measurement, this means that MA- quasi-molecular ions must
be present, but once again no statement can be made
concerning the presence of ~ + quasi-molecular ions, since
at the ~ero point 100, a quasi-molecular ion current can
never result from MB- ~uasi-molecular ions.
With this intermediate result in hand, the system is
switched to the second channel 61, and a measurement with
the pulses 71', 72' is immediately performed. The
measurement point is once again the zero point 100.
I~ once again no measured signal occurs with this second
measurement, this means (when considered in isolation) that
no ~ + quasi-molecular ions can be present, since, on the
evidence of curve B' in Figure 3, they would necessarily
have produced a molecular ion current at an effective
voltage of 110 V.
On the other hand, the absence of the measured signal in
this isolated case does not also mean that no MA- molecular
ions are present, since here again it is fundamentally
impossible for any measured signal from MA- quasi-molecular
ions to be generated at the zero point 100.
2 ~
Xf, on the other hand, the second measurement yields a
positive measured signal, this means that MB+ quasi-
molecular ions at least are present, while on the other hand
it is once again impossible to make any statement concerning
~IA- quasi-molecular ions.
These four possible results must then be fed out to the
outputs 82 to 85 of the control unit 50. In the case of the
first measurement with the first channel 60, the zero signal
(no measured current) is present at the output 82, while a
positive measured signal ~measured current present) would
exist at output 83. The same applies for outputs 84 and 85
for measurements with the second channel 61.
The four possible output signals from the two ~easurements
are then correlated by means of the logical elements 86, 87,
and 90 to 93 so as to display the four possible results.
For example, if neither MA- quasi-molecular ions nor MB+
quasi-molecular ions are present, no measured signal will
occur on either channel 60 or 61 during the two measurements.
In the depiction in Figure 2, this zero signal is present at
the outputs 82 and 84, is reshaped in the inverters 86, 87
into positive signals, and in that form passes to the two
inputs of the AND gate 93, thus illuminating the indicator
97 which indicates that neither of the two measured
substances A or B was present in the gas mixture 32.
If, on the other hand, both measured substances A and B were
present in the gas mixture 32, a positive measured signal
will then be generated in both measurements, and will thus
~e applied as positive logical signals to the outputs 83 and
85. The AND gate 90 will conse~uently become conductive and
actuate the associated indicator 94. The same applies for
219 ~ ~ ~ 6 1
the two additional indicators for isolated occurrence of the
first measured substance A (indicator 95) or measured
substance B (indicator 96).
It goes without saying in this context that the logic
clepicted in Figure 2 was made relatively complex in
configuration solely for purposes of clarity, and that in an
actual application, other logical elements and connections
can also be used.
It also goes without saying that the measurements do not
necessarily need to be performed at the zero point 100.
Other measurement points can also be selected, provided no
undesired compensation occurs because calibration curve
values have the same magnitude but different signs. One such
"forbidden" operating point is labeled 101 in Figure 3.
In a third step, or in a step preceding the first step, the
third channel 62 of the voltage source 40 can now be
switched in, thus applying the very high constant voltage 73
to the electrodes 21, 23.
The result of applying the very high constant voltage 73
with a very high static electric field strength of, for
example, 1000 V/cm, is that all the charge carriers located
in the measurement chamber 10, i.e. the mobile reaction ions
2- and H+ as well as the relatively slow-moving quasi-
molecular ions MA- and MB+, migrate to the respective
electrode, thus producing a total ion current that
corresponds to the total number of free charge carriers in
the measurement chamber 10.
This measurement result, in conne~tion with the two previous
measurements using channels 60 and 61, now makes it possible
2 ~
- 31 -
first to undertake quantitative measurements and thus
determine concentrations of the measured substances A and B,
while on the other hand concentrations of interfering
substances can also be determined and thus subtracted;
lastly, the effect of external interfering influences, for
example due to fluctuations in temperature, pressure, and
humidity, also becomes evident.
The last statement is true because the three measurements
with channels 60, 61, and 62 are performed in immediate
succession, for example at intervals of 100 ms, so that all
three measurements are performed under the same conditions
(which may be affected by interference), and these
interfering variables become evident by comparing the
measurements.
With regard to determining the concentration of the measured
substances A and B and any interfering substances that may
be present, it must be remembered that in the first two
measurements with channels 60 and 61, the result in each
case is a function of the (known) attachment probability,
the (unknown) concentration of the molecules of measured
substances A and B, the (unknown~ concentration of the
molecules of the interfering substance, and the (also
unknown) concentration of the reaction ions.
But if the total number of charge carriers was determined by
means of the third (or prior) measurement, and if on the
other hand the ratio between the concentrations of the
molecules of the measured substances is known beaause a
ratio has been defined between the first and the second
measurement, the remainlng unknown variables can thus be
determined by calculation or by suitable calculation
circuitry.
- 32 - 2 ~ r~ ~ ~
A calculation of this kind can, for example, be performed as
follows:
First the equations for product ion formation are
considered, as follows:
(H20)nH ~ (A) (A)H + nH20
(H20)mO2 + (A) ~ (~)2 + mH20
in which H20 is the contribution of water clusters to the
reaction ions due to atmospheric humidity of the ambient
air; H+ and 2- designate the reaction ions; and A and B in
turn indicate the measured substance to which the positive
reaction ions H or negative reaction ions 2- are
respectively attached. In the example explained above, A
thus again designates the group of nerve poisons, and B the
group of vesicants.
X and Y denote the probabilities of formation applicable to
the ionization of measured substances A and B respectively.
The formation probabilities X, Y could be defined by ion
mobility spectroscopy (IMS) if ideal conditions are present,
although this requires expensive and complex apparatus.
If we now consider the three different measurements of the
types described in detail above, namely the zero-point
measurement at high voltage as well as the two pulse
measurements, the formation probabilities X and Y are
reduced by different amounts in each of the three
measurements. This is indicated hereinafter by the indices
Xl, X2, X3 and Yl, Y2~ and Y3. The respective reduction
factors Xl/X, X2~X, X3/X and Yl/Y, Y2/Y, 3/
' :
,, , .
2 ~
defined, according to the invention, by ion mobillty
spectroscopy in which the ion source is also the measurement
cell of the device.
Eurthermore, a constant K is used hereinafter to denote the
number of unattached residual ions, with K being dependent
on the pressure, temperature, drift velocity, half-life, and
the like, in other words the general measurement conditions.
If we then initially consider, as the first measurement, the
zero-point measurement at high voltage, we find the
following for the resulting ion current Il:
I1 = K + XlA + YlB
If we then consider that in this case Xl and Y1 are
approximately zero, since the reaction ions have a very high
velocity compared to the quasi-molecular ions, we get
approximately the following result for the ion current in
this measurement:
Il K
If we then consider the second case - ionization of the
first measured substance A - we get for the resulting ion
current I2:
I2 = K + X2~ + Y2B
~Q6:~6~l
- 34 -
w~ith the additional fact that in this case X2 is much
greater than Y2.
For the third measurement instance - ionization of measured
substance B - we get the following for the resulting ion
current I3:
I3 = K + X3A + Y3B
with X3 being much smaller than Y3 in this case.
From these three measurement results, we can thus define the
concentration of measured substance A as
2 1 Y2A)X2 = (I3 - I1 -Y3B)/X
and the concentration of measured substance B as
X2 X2
B = (-(I3 - I1) ~ (I2 ~ I1))/(X Y3 Y2) -
,: :
