Vapor Generator for the Calibration and Test of Explosive

2008 IEEE International Conference on Technologies for Homeland Security
Vapor Generator for the Calibration and Test of Explosive Detectors
Bogdan V. Antohe, Donald J. Hayes, David W. Taylor, David B. Wallace, Michael E. Grove and Mark Christison
MicroFab Technologies, Inc., 1104 Summit Ave. 110, Plano, Texas 75074 (972) 578-8076
[email protected], [email protected], [email protected], [email protected],
[email protected]
Abstract— Thousands of explosive vapor trace detectors
deployed in the field require frequent verification and
calibration to guarantee their accurate operation. In this
paper we describe the use of precision micro-dispensing
technology in the development of vapor generation
instruments for the calibration and testing of explosive
detectors. Several digitally controlled systems ranging from
bench-top/research instrument to a hand held unit using inkjet technology have been prototyped and tested. The use of
digitally controlled ink-jet dispensing to precisely eject
minute amounts of dilute explosive solutions and convert
them into vapor has been demonstrated. An electrical pulse
applied to a piezoelectric micro-dispenser causes a drop of
fluid to be ejected through a precise orifice. These droplets
land on a heater that converts them to vapor. The amount of
explosive delivered to the detectors can be controlled by the
number of drops (dose mode – specified number of drops is
generated) and the frequency of the droplet generation
(continuous mode – droplets are generated continuously at
fixed frequency). An additional control of the amount of
explosive is provided by the concentration of the explosive
solution that is dispensed facilitating the generation of
infinitely small amounts of explosive vapors. The prototypes
built have been evaluated with commercially available
detectors in both dose and continuous modes. Ultimately,
the ability to further miniaturize the vapor generators will
lead to units that are embedded into next generation
detectors for real-time verification and calibration.
1. INTRODUCTION
Vapor Trace Detectors
The need to detect very low levels of illicit substances
(chemical and biological agents and explosives) has become,
after September 11, 2001, a priority for the federal, state and
local government agencies. Systems capable of detecting
minute amounts of the above materials are required in the
airports, border crossings and high security areas.
Explosives represent one important class of illicit substances
with the military explosives (e.g. TNT, RDX, PETN, HMX)
being an important subclass that is currently targeted by the
various trace detection methods. Trace detection – detection
of very small amounts of the explosives – identifies people
or things that have come in contact with explosives. The
trace detection methods have been implemented in a variety
of instruments ranging from hand held and portable to
bench-top or portals.[1,2]
The detection methods can be classified in: separation
methods (gas chromatography - GC, high performance
liquid chromatography – HPLC, capillary electrophoresis CE), ion detection methods (mass spectroscopy – MS, ion
mobility spectrometry – IMS), vibrational spectrometric
methods (infrared absorption, Raman scattering, etc.),
UV/visible methods (fluorescing polymers, colorimetric
reactions), immunochemical sensors, electrochemical
sensors. [1,3,4]
The Need for Vapor Calibrator
The vapor generator provides the reliable standards that are
needed to maintain the accuracy and sensitivity of the vapor
trace detectors that are used in the field. To make sure that
the sensitivity of a system is still acceptable, periodic
evaluation and, if necessary, recalibration are required. By
creating explosive vapors of known concentration, the vapor
generator will provide the means to verify the detection limit
of the systems in the field and their recalibration. IMS is one
of the most popular technologies employed in the vapor
trace detection, but it is sensitive to variations in pressure
due to weather or altitude. The vapor generator can be used
to recalibrate IMS systems at various operating /
environment conditions.
The continuous research and development for the
improvement of the detection limit requires a vapor source
of very low concentration. It is desired that such a vapor
source is portable, because a large number of the vapor trace
detectors deployed in the field are fixed. Existing
technologies [5,6] are not very precise and cannot be easily
miniaturized. By creating a vapor containing a precise
amount of explosives, the vapor generator provides the
means to quantify the improvements in the development of
the explosive detectors. Moreover, the capability to
dynamically adjust the vapor concentration by several
mechanisms (amount of fluid used for evaporation, temporal
distribution of the evaporated material), will allow the
verification of the full range of the detectors and allow
calibration over its entirety.
The wide variety of methods and their implementation in the
vapor trace detectors make their comparison difficult. Moore
[4] states that “there is an urgent need for standard testing
protocols to enable fair and complete comparisons of
available trace and stand off detection methods”. An
accurate explosive vapor generator can provide the means to
calibrate the instruments deployed in the field and, at the
same time, can be a tool to compare the sensitivity of
various instruments.
2. INK-JET
Background
In the case of the drop on demand (DOD) ink-jet printing
systems, a volumetric change in the fluid is induced either
by the displacement of a piezoelectric material that is
coupled to the fluid [7], or by the formation of a vapor
bubble in the ink, caused by heating a resistive element.[8]
This volumetric change causes pressure/velocity transients
to occur in the fluid and these are directed to produce a drop
that issues from an orifice.[9,10] Demand mode ink-jet
printing systems produce a droplet only when desired.
Figure 1 illustrates a piezoelectric based DOD system, in
which a droplet is generated only when a voltage pulse is
applied to the piezoelectric actuator. In traditional printing
systems the drops are targeted to specific locations on the
substrate/paper. For the vapor calibrator discussed in here,
the droplets land on the heating element.
Piezo-Transducer
Orifice
are approximately equal to the orifice diameter of the droplet
generator.
Figure 2 – Sequence of generation of 50µm droplets
(ethylene glycol) at 4kHz
The Advantages of Ink-Jet Technology
Some of the general benefits of the ink-jet dispensing are
transferred to this application. They are:
• High precision: Ink-jetting produces highly repeatable
drops. The very sharp edges in the photo in Figure 2
reflect the high repeatability of the dispensing.
• Continuous variation: The very small size of the
individual drops (20-200picoliters) produces, from the
perspective of this application, almost continuous
variation of the total (accumulated) amount.
• Dynamic range: The output range of a vapor generator
based on ink-jet microdispensers extends from almost
zero (equivalent of several drops) to several thousands
of parts per trillion. The low end resolution can be
further increased by using more dilute solutions
containing the substances of interest.
• Data driven: The piezoelectric dispensers are
electrically driven and they can be controlled from data
files. This makes the technology easily adaptable for
automatic testing.
3. VAPOR CALIBRATORS
Vapor Generator for Olfaction Threshold
The initial instrument developed for detector evaluation was
designed for the determination of the threshold (sensitivity)
of the human nose which can be used as an early symptom
1.0
Substrate
0.9
0.8
Substrate
Motion
Signal [V]
Driver
Character
Data
0.7
Fluid at
Ambient
Pressure
Data Pulse Train
Figure 1 – Schematic of a drop on demand ink-jet
printer in which each pulse produces a drop
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.0
5.0
10.0
15.0
20.0
Time from increase [s]
The process of droplet formation and ejection from the
orifice of a DOD device is indicated in the time-sequenced
photos of Figure 2, taken under stroboscopic illumination at
the driving frequency of the dispensing device. The droplets
Figure 3 – Photoionization detector response for
various number of drops; bottom curve - 6 drops and
top curve - 618 drops.
for the onset of neurodegenerative diseases like Alzheimer’s
and Parkinson’s.[11,12] For this application droplets of
lemon extract and phenethyl alcohol are generated and
deposited onto a heater where they are evaporated. The
vapors are presented to the patient nose. A modified
staircase testing procedure was implemented to determine an
olfaction threshold for each subject. Experiments with this
olfactometer provided information on the output as a
function of number of drops (Figure 3).
The dispense head sub-system (Figure 5) consists of the
MicroJet devices, heater, reservoirs, temperature controls
and interconnects.
Bench-top Prototypes
Second Prototype—A second prototype of a bench-top
vapor generator (Figure 6) was reduced to a single dispenser
per head, but incorporated other functions: ability to control
four separate dispenser heads, cooled reservoirs for the
explosive solution, software controlled heater temperature.
First Prototype—A first prototype was built for NIST and it
incorporated six different MicroJet dispenser, each with a
separate reservoir. This system permits the use of multiple
fluids and thus provides the capability of testing interactions
between the different substances.
Fluid Drops
This system was used in conjunction with various detectors
in continuous operation. The parameters of the studies
consisted in the type of explosive used, the flow rate of the
carrier gas (dry nitrogen) and the frequency of generation of
the droplets. [13]
Heated Wall
Sensor
Under
Test
Jet
Air Flow
Vaporizer
(Heater)
Figure 4 – System schematic and key components
Figure 6 – Multi channel bench-top vapor generator
The principle of operation is similar to the olfactometer and
is depicted in Figure 4. The droplets landing on the heater
(at set temperature) are evaporated and the vapors are
carried towards the detector / sensor to be tested. A
particularity of this system consists in the fact that the walls
of the outlet tube are heated to prevent the deposition
(condensation) of the explosive vapors on the wall. Because
of the heating a cooling coil (Figure 5) was used to maintain
a lower temperature in the region upstream of the vaporizer.
Additional functions were incorporated in the software to
control all four dispenser head, to record environmental
conditions and the tests that were run. Current work includes
the control of the temperature profile in time for the
evaporator/heaters. This function might be desired in the
case of detectors that are sensitive to the solvent that is used
to dissolve the explosives. In this case, the heater could be
set initially at a relatively low temperature that ensures the
evaporation of the solvent. Once the solvent is driven away,
the detector can be exposed to the air stream coming out of
the vapor generator while the temperature is increased to
values that drive off the explosive from the heater. A final
heating at high temperature ensures that all the residuals on
the heater are burned off.
Portable System
Reservoirs
Wall Heater
Cooling Coil
Vaporizer
Figure 5 – Dispense head sub-system
A prototype of a portable vapor generator was designed and
built (Figure 7). This hand held system has a similar
principle of operation as the bench-top systems. Most of the
differences are in the actual construction, especially in the
control part. The control of the vapor generator was done
using a two channel MP3 player. One channel provides a
frequency signal that is used to adjust the temperature of the
heater (the higher the frequency the higher the heater
temperature). The second channel provides the trigger signal
to generate the drops. The two outputs can be synchronized
to correlate the drop generation to the temperature of the
heater.
Thermal Analysis—Thermal analysis was also run to
determine if the heater power is sufficient to achieve
temperatures in the 200ºC range (possibly required for the
residual burn off). Continuous power on to the heater
produced temperature exceeding the desired values (Figure
9).
Output—The final part of the simulations incorporated the
analysis of the concentration of the explosive vapors at the
outlet as a function of time (Figure 10). In these simulations
we have assumed a triangular release of the explosive on the
heater.
Figure 7 – Prototype of a handheld vapor
calibrator
2.5E-06
point 1
point 4
point 7
point 10
Flow Analysis—This vapor generator does not incorporate a
heater for the outlet area. Even though the output tube is
fairly short, we have evaluated the trajectory of the particles
leaving the heater by numerical simulations in Fluent. The
flow pattern has indicated that the markers released from the
heater surface would be concentrated to the axial region of
the output (Figure 8).
Explosive mass fraction [-]
2.0E-06
point 2
point 5
point 8
outlet average
point 3
point 6
point 9
1
2
3
1.5E-06
10
9
8
4
5
6
1.0E-06
7
5.0E-07
0.0E+00
0
0.25
0.5
0.75
1
Time [s]
Figure 10 – Time dependent explosive mass fraction at
various points in the outlet plane
4. TESTING RESULTS
Figure 8 – Output section of the portable vapor
generator. The colored lines represent the
trajectory of the markers leaving the heater
Figure 9 – Temperature distribution on the solid
surfaces presented as isothermal contours
The systems are capable of operating in two modes: a
continuous mode in which the droplets are generated
continuously at a selectable fixed frequency and a dose
mode which consists of the generation of a specified number
of drops at a selected frequency. From a practical
perspective, the portable vapor generator might only need
the dose mode operation.
NIST has used the first bench-top system fabricated by
MicroFab to evaluate the potential range provided by a
vapor generator employing ink-jet microdispensers for
several explosives (RDX, TNT and PETN) and has shown
that the concentration can be varied almost continuously
from 0 to hundredths of parts per trillion (v/v) when
operating the dispenser in continuous mode.[13,14]
Experiments have also shown the ability to step up the
output by stepping up the drop generation frequency.
Tests were also made with the second bench-top vapor
generator in both continuous mode and dose mode. Figure
11 shows a sample result. In the case depicted in the figure a
solution obtained by diluting a TNT standard using ethanol
to an effective concentration of 1µg/mL was dispensed in
dose mode using one to five drops. The equivalent amount
of explosive varied from 62 to 310 femtograms.
18
Detector response
9
16
Number of drops
8
14
7
12
6
10
5
8
4
6
3
4
2
2
1
Number of drops
Detector response
The system can be further miniaturized and implemented in
a module. Such a modular component can be incorporated in
the vapor trace detectors for automatic testing and
calibration.
10
20
0
0
0
50
100
150
200
250
300
350
Amount of TNT dispensed [femtograms]
Figure 11 – Detector response at various amounts of
dispensed explosive solution
5. CONCLUSIONS AND FUTURE DIRECTIONS
Feasibility of the use of the ink-jet based explosive vapor
generator to calibrate the explosive trace detectors was
demonstrated with the two bench-top prototypes. The
prototype of the portable vapor generator has shown the
ability to miniaturize the system for a future hand–held
(field) calibrator. All built prototypes have shown that they
are capable to generate femtogram level of explosives. The
range of explosive vapors produced by the prototypes satisfy
the current detectors on the market and can be easily adapted
for the use of any forthcoming vapor detectors.
Work will continue with the design and fabrication of a
research test bed that incorporates the explosive vapor
generation using ink-jet. This system will be evaluated in
terms of the explosive output by Gas Chromatography Mass
Spectroscopy (GCMS) and with commercial explosive
vapor detectors.
A hand held (portable) vapor generator will be designed and
built using selected and optimized functions that are
transferred from the research test bed. Concept of a hand
held vapor generator is shown in Figure 12.
ACKNOWLEDGEMENT
This work was sponsored in part by NSF Grant OII0611204. The initial olfaction work was funded by NIH
Grant MH/DC56335. The authors would like to thank the
Surface and Microanalysis Science Division at NIST for
helping define the first bench-top vapor generator and
providing feedback. Thanks are also given to ICX Nomadics
and GE Security for supporting our efforts.
REFERENCES
[1] L. Theisen, D. W. Hannum, D. W. Murray, and J. E.
Parmeter, “Survey of Commercially Available Explosives
Detection Technologies and Equipment 2004,” Report for
NIJ
grant
96-MU-MU-K011,
2004.
http://www.ncjrs.org/pdffiles1/nij/grants/208861.pdf
[2] S. F. Hallowell, “Screening people for illicit substances:
a survey of current portal technology,” Talanta 54, 447-458,
2001.
[3] D. S. Moore, “Recent Advances in Trace Explosives
Detection,” Instrumentation Sensing and Imaging 8, 9-38,
2004.
[4] D. S. Moore, “Instrumentation for trace detection of high
explosives,” Rev. Sci. Instrm. 75, 2499-2512, 2004.
[5] G. A. Eiceman, “Quantitative calibration of the vapor
levels of TNT, RDX, and PETN using a diffusion generator
with gravimetry and ion mobility spectrometry,” Talanta 45,
57-74, 1997.
[6] J. P. Davies, L. G. Blackwood, S. G. David, L. D.
Goodrich, and R. A. Larson, “Design and calibration of
pulsed vapor generators for 2,4,6-Trinitrotoluene, cyclo1,3,5-trimethylene-2,4,6-trinitramine, and Pentaerythritol
Tetranitrate,” Anal. Chem. 65, 3004-3009, 1993.
[7] D. B. Wallace, "A Method of Characteristics Model of a
Drop-On-Demand Ink-Jet Device Using an Integral Method
Drop Formation Model," ASME publication 89-WA/FE-4,
December 1989.
[8] J. S. Aden, J. H. Bohorquez, D. M. Collins, M. D.
Crook, A. Garcia, and U. E. Hess, “The Third Generation
HP Thermal InkJet Printhead,” Hewlett-Packard Journal 45,
February 1994.
Figure 12 – Concept of a hand held vapor
calibrator.
[9] D. B. Bogy and F. E. Talke, "Experimental and
Theoretical Study of Wave Propagation Phenomena in
Drop-On-Demand Ink Jet Devices," IBM Journ. Res.
Develop. 29, 314-321, 1984.
[10] J. F. Dijksman, "Hydrodynamics of Small Tubular
Pumps," Journ. Fluid Mech. 139, 173-191, 1984.
[11] D. B. Wallace, D. Taylor, B. V. Antohe, I. Achiriloaie,
N. Comparini, R. M. Stewart and M. K. Sanghera,
"Determination of the olfactory threshold using a
piezoelectric microdispenser for neurodegenerative disease
diagnostics," Meas. Sci. Technol. 17, 3102-3109, 2006.
[12] D. J. Hayes, D. B. Wallace, D. Taylor, B. V. Antohe, I.
Achiriloaie and N. Comparini, “A Digital Olfactometer for
Smell Threshold Measurements for Neurodegenerative
Disease Diagnostics,” Proceed. IEEE Eng. and Med. in Bio.
Workshop, UT Dallas, November 11-12, 2007.
[13] R. M. Verkouteren, G. Gillen, and D. W. Taylor,
“Piezoelectric trace vapor calibrator,” Rev. Sci. Instrum. 77,
085104 (2006)
[14] R. M. Verkouteren, G. Gillen and D. Taylor, Vaporjet
calibrator for IMS-based trace explosive and CWA
detectors. ISIMS 2005, Chateau Maffliers, Paris, France,
2005.
http://www.cstl.nist.gov/div837/Division/outputs/Explosives
/ISIMS_2005_Paris.pdf