Design and construct of a new detector for gas chromatography based on continuous negative corona discharge.

Abstract

In this work, a new detector was designed and constructed based on negative corona discharge. This detector can be used separately or as a detector in gas chromatography. The detector and chromatographic variables including cell temperature, gas flow rates, voltage between the two electrodes, and column temperature were optimized. Chloroform was used as a test compound to evaluate the performance of the detector. The detection limit of chloroform was obtained 0.78 ng∕ml and its dynamic range was over the range of 2-840 ng∕ml. The relative standard detection was about 6% for the limit of quantification. This detector is able to be used as an alternative for analysis of compounds containing electronegative elements.

Full text

Page 1

Design and construct of a new detector for gas chromatography
based on continuous negative corona discharge
M. Sharifian Ghahfarokhi and T. Khayamian

Citation: Rev. Sci. Instrum. 82, 055114 (2011); doi: 10.1063/1.3589263
View online: http://dx.doi.org/10.1063/1.3589263
View Table of Contents: http://rsi.aip.org/resource/1/RSINAK/v82/i5
Published by the American Institute of Physics.

Related Articles
Electrical and spectroscopic analysis in nanostructured SnO2: “Long-term” resistance drift is due to in-diffusion
J. Appl. Phys. 110, 093711 (2011)
Irradiance-based emissivity correction in infrared thermography for electronic applications
Rev. Sci. Instrum. 82, 114901 (2011)
Symmetry breaking effect of dc bias current on magnetoimpedance in microwire with helical anisotropy:
Application to magnetic sensors
J. Appl. Phys. 110, 086105 (2011)
A carbon nanotube immunosensor for Salmonella
AIP Advances 1, 042127 (2011)
Enhanced H2 sensitivity at room temperature of ZnO nanowires functionalized by Pd nanoparticles
J. Appl. Phys. 110, 084312 (2011)

Additional information on Rev. Sci. Instrum.
Journal Homepage: http://rsi.aip.org
Journal Information: http://rsi.aip.org/about/about_the_journal
Top downloads: http://rsi.aip.org/features/most_downloaded
Information for Authors: http://rsi.aip.org/authors
Downloaded 19 Nov 2011 to 212.50.248.4. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions

Page 2

REVIEW OF SCIENTIFIC INSTRUMENTS 82, 055114 (2011)
Design and construct of a new detector for gas chromatography based on
continuous negative corona discharge
M. Sharifian Ghahfarokhi and T. Khayamiana)
Department of Chemistry, Isfahan University of Technology, Isfahan 84156-83111, Iran
(Received 8 November 2010; accepted 17 April 2011; published online 25 May 2011)
In this work, a new detector was designed and constructed based on negative corona discharge. This
detector can be used separately or as a detector in gas chromatography. The detector and chromato-
graphic variables including cell temperature, gas flow rates, voltage between the two electrodes, and
column temperature were optimized. Chloroform was used as a test compound to evaluate the per-
formance of the detector. The detection limit of chloroform was obtained 0.78 ng/ml and its dynamic
rangewasovertherangeof2–840ng/ml.Therelativestandarddetectionwasabout6%forthelimitof
quantification. This detector is able to be used as an alternative for analysis of compounds containing
electronegative elements. © 2011 American Institute of Physics. [doi:10.1063/1.3589263]
I. INTRODUCTION
Electron capture detector (ECD) is a well-known selec-
tive and sensitive detector coupled to gas chromatography
(GC) for the analysis of compounds associated with the elec-
tronegative elements such as polyhalogenated organic com-
pounds, pesticides, herbicides, etc. The key limitation for
utilizing this detector is attributed to the radioactive63Ni
electron source in its body. Researchers have tried to re-
place63Ni with other electron sources such as photoioniza-
tion by using hydrogen discharge lamp,1photoionization by
ultraviolet light,2,3alkali cation emitter,4,5thermo emitter,6,7
and an activated photocathode.8Pulsed corona discharge in
a constant potential or current operation mode has also been
introduced.9These detectors were named as the pulsed dis-
charge electron capture detectors. In this work, a continuous
corona discharge in a negative mode was used as an electron
source in ECD and its performance was evaluated.
Corona discharge is an atmospheric pressure chemical
ionization that could be considered as one of the best and eas-
iest ways to obtain glow-like discharge in gaseous media at
atmospheric pressure. It is one of the varieties of electronic
discharges, which may develop in a strongly non-uniform
field where the radius of curvature of at least one electrode
is much smaller than the interelectrode distance. Corona dis-
charge can be generated in positive or negative modes. In the
negative mode, the needle electrode is connected to the neg-
ative terminal of the power supply. Furthermore, the corona
discharge operates in the pulse or continuous modes. A con-
tinuous corona discharge is more stable than the other one,
because it has enough time to reach self-sustained condi-
tions while a pulse corona discharge is limited by the pulse
width. Different types of electrode geometry might be used
for making a corona discharge such as a point-to-plane ge-
ometry, a point-to-ring geometry, a wire-in-cylinder geome-
try, a multipoint-to-plate geometry,10–12and a point-to-point
geometry.13,14A point-to-plane geometry produces a highly
non-uniform field at the point electrode and a uniform field
a)Author to whom correspondence should be addressed. Electronic mail:
taghi@cc.iut.ac.ir. Telephone: +98 311 391 2351. Fax: +98 311 391 2350.
at the plane electrode which is suitable for making a corona
discharge.
In this work, a continuous negative corona discharge was
introduced as a new GC detector for analysis of organic com-
pounds containing electronegative elements. The analysis of
chloroform as a test compound was examined and the results
demonstrated the applicability of the detector.
II. THEORY OF CORONA DISCHARGE
A. The mechanism of corona discharge
The mechanism of corona discharge is dependent on
its polarity (positive or negative corona) and it is well de-
scribed in books.15,16The mechanism of negative corona dis-
charge with a point-to-plane geometry in the pure nitrogen
is described as follows: at a sufficient voltage, electrons start
to leave the negative electrode point. These electrons are
then multiplied due to electron impact ionization of the gas
molecules and produce more positive ions. The positive ions
hit the negative tip and knock out more electrons. The cur-
rent for the negative corona discharge in pure nitrogen, or any
non-electron attaching gas such as helium or argon is much
higher than that of the positive corona. In the negative corona,
the positive ions are accelerated in the high field region to-
ward the needle and then hit the needle tip with a high energy.
This action causes to generate a large number of electrons.
The current in negative corona discharge in pure nitrogen was
measured and it was about 30–60 mA whereas the total cur-
rent of the63Ni (∼12 mCi), used in ion mobility spectrometry
as an ionization source, is about 1 nA. Therefore, it can be
concluded that the electron density in the negative corona dis-
charge is about 1 × 106times higher than that in the63Ni
source.17In such a high electron density, electron capture ef-
ficiency is very high, thus, the linear range and the sensitiv-
ity for detecting electronegative species is expected to be en-
hanced considerably.17–21
In the negative corona discharge, produced electrons
may participate in three types of reactions: recombination
reaction, dissociative electron attachment, and associative
0034-6748/2011/82(5)/055114/5/$30.00 © 2011 American Institute of Physics
82, 055114-1
Downloaded 19 Nov 2011 to 212.50.248.4. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions

Page 3

055114-2M. Sharifian Ghahfarokhi and T. KhayamianRev. Sci. Instrum. 82, 055114 (2011)
electron attachment.8,21In addition to these reactions, the
corona electrical current can be decreased by two other fac-
tors; First, excitation of molecules instead of ionization and
formation of radicals such as N, O, OH, H, HO2, and O3.
When oxygen as the impurity is introduced into the cell, some
other ions would be produced during corona discharge such
as O2−, O2 (H2O)−, and NO2−. In addition, with present
of moisture in nitrogen gas, some clusters of O2(H2O)n−,
OH(H2O)n−, NOx(H2O)n−would be produced.21,22Cluster
ions produced in corona discharge make intensive chemical
background and as a result signal to noise ratio decreases.23
A variety of methods might be used to avoid producing of
these ions in corona discharge including using high tempera-
ture in chamber to reduce cluster ions,24using dry and pure
gas,25and increasing electrode distance.13To identify pro-
duced ions in corona discharge, it is suggested to use mass
spectrometry.26
B. Townsend curves
Current-to-voltage curve in corona discharge was de-
scribed by Townsend equation in 1914,27
IαV(V − V0).
(1)
In this equation, I is the corona current, V is the corona
voltage, and V0is the corona threshold voltage. Based on this
equation, the curve I/V based on V is linear in corona dis-
charge region. A critical value of threshold voltage is depen-
dent on the polarity, fluid pressure, temperature, corona gas
flow rate, geometry of electrodes, and other factors such as
the presence of impurities.8,17,28,29
III. EXPERIMENTAL SECTION
A. Apparatus and chemicals
A gas chromatograph system COMPACT II (Ireland) was
coupled to the continuous negative corona discharge as the
detector. A packed column of 10% SE-30 on Chromosorb
W AW, 60/80 was used. Column and injection port tempera-
tures were 100◦C and 160◦C, respectively. Column gas flow
rate was 50 ml/min. A gas cylinder delivered two streams of
gas for generation of the corona discharge and as the carrier
gas for the GC column. The system also had two flow me-
ters to adjust the flow rates of these gases. Chloroform and
Needle electrode
Target electrode
Gas Exit
Gas Inter
Teflon O-ring
Mini-ruler
45 mm
25 mm
FIG. 1. (Color online) Image of the corona cell.
toluene were purchased from Merk. Teflon (Germany) and
Pyrex (Germany) were obtained from home shops.
B. Corona cell
The image of the corona discharge cell is shown in
Fig. 1. The corona cell was constructed by Pyrex, because
it was more convenient to observe the inside of the cell and
the needle tip. A stainless steel sharp needle was used as the
needle electrode, and a thin gold plate was used as a counter
electrode. Two electrodes are removable; therefore it can be
possible to adjust the electrode distances. In order to avoid air
leaking and impurities into the corona cell, the two ends of the
glass tube were sealed by Teflon O-rings. The corona cell was
placed in a homemade heater to heat the detector. It would be
possibletoreplacestainlesssteelneedle byplatinumneedleto
reduce corrosion of the needle tip in the continuous negative
corona discharge.
C. Electrical circuit
A high voltage direct current (dc) power supply was em-
ployed to generate the corona discharge, shown in Fig. 2. To
make a negative corona discharge, the needle was connected
to the negative pole of the power supply and the counter elec-
trode was connected to the positive pole. A high resistance
of 12 M? was used in the circuit to reduce the current and
prevent damaging from it. A low resistance (33 k?) was used
in the negative pole of the circuit to measure the corona dis-
charge current. An amplifier with adjustable gain was used to
amplify the signal. The amplified signal then transferred to
an analog-to-digital (A/D) converter to monitor signal on the
computer. All of the electronic elements (dc power supply,
R=33 KΩ
R=12 MΩ
H. V. Power Supply
A/D Converter
Amplifier
Needle electrode Target electrode


FIG. 2. (Color online) Electrical circuit used to record the corona signal.
Downloaded 19 Nov 2011 to 212.50.248.4. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions

Page 4

055114-3M. Sharifian Ghahfarokhi and T. KhayamianRev. Sci. Instrum. 82, 055114 (2011)
amplifier, and A/D converter) have been designed and con-
structed in the University of Technology.
IV. RESULTS AND DISCUSSIONS
A. The effect of temperature
The effect of temperature on negative corona discharge
current was investigated using the circuit shown in Fig. 2.
The corona gas is passed toward a flow meter and then en-
ters the corona chamber by passing around the needle. There
is an exit for the gases at the end of the cell. To make an inert
medium in the discharge region, all the connecting materials
were made by Teflon and the cell was constructed by Pyrex.
The distance between the two electrodes, d, could be pre-
cisely optimized by using a mini-ruler, signed on the outside
of the cell body, shown in Fig. 1. The positions of the elec-
trodes were fixed in their locations using two Teflon O-rings.
Figure 3(a) shows the effect of temperature on the negative
corona discharge current in pure nitrogen. The results show
that in a constant corona voltage, gas flow rate, and electrode
distance, the corona current increases by increasing the cell
temperature. Besides, the corona threshold voltage, V0 de-
creases by increasing the cell temperature. The obtained ex-
perimental results were anticipated based on the mechanism,
described in Sec. II B.
B. The effect of electrode distances
In order to investigate the electrode distance on the nega-
tive corona discharge current, the distance between the two
electrodes was changed in a constant nitrogen flow rate,
50 ml/min, and temperature, 70◦C. The results are shown in
Fig. 3(b) and it shows that the current increases by decreasing
the electrode distance. The corona threshold voltage, V0, also
decreases by decreasing the electrode distance.
C. The effect of gas flow rate
In a constant electrode distance and cell temperature,
d = 1 cm, and T = 25◦C, the effect of gas flow rate on the
negative corona discharge current is investigated and the re-
sults are shown in Fig. 3. The results show that V0shifts to
the lower voltages by increasing the gas flow rate.
D. Corona discharge cell as the detector
Figure 4 shows the corona discharge detector coupled to
the gas chromatograph. When a compound associated with an
electronegative element was injected into the injection port
of the GC, it was observed that the electrical current of the
corona discharge decreased. The detector was tested by in-
jecting solutions of chloroform in toluene. The variables for
creating the corona discharge including electrode distance,
cell temperature, and chromatographic variables such as the
column temperature and gas flow rate were optimized. The
optimized conditions are listed in Table I. The peak area
of the chloroform was related to its concentration in quan-
titative analysis. The results of injection of 1μl solution of
R2 = 0.9941
R2 = 0.987
R2 = 0.9907
R2 = 0.996
R2 = 0.993
0
0.5
1
1.5
2
2.5
3
Voltage (kV)
I / V (microA/kV )
T = 25 oc
T = 70 oc
T = 100 oc
T = 130 oc
T = 160 oc
R2 = 0.987
R2 = 0.989
R2 = 0.999
R2 = 0.9993
0
2
4
6
8
10
12
3
Voltage (kV)
I / V (microA/kV )
d = 10 mm
d = 8 mm
d = 6 mm
d = 4 mm
R2 = 0.9907
R2 = 0.9941
R2 = 0.9954
R2 = 0.9941
R2 = 0.9984
R2 = 0.9986
0
0.5
1
1.5
2
2.5
3
5
Voltage (kV)
I / V (microA/kV)
f = 25 (mL/min)
f = 50 (mL/min)
f = 100 (mL/min)
f = 150 (mL/min)
f = 190 (mL/min)
f = 230 (mL/min)
(a)
(b)
(c)
10 9876
5
10 9876
5
10
9
8
76
4
FIG. 3. (Color online) Townsend curves; (a) The effect of temperature on the
negative corona discharge current in pure nitrogen; d, electrode distance is
1 cm; f, nitrogen flow rate is 50 ml/min; (b) The effect of electrode distance
on the negative corona discharge current in pure nitrogen; T, the temperature
is 70◦C; f, the flow rate is 50 ml/min; (c) The effect of gas flow rate on
thenegativecoronadischargecurrentinpurenitrogen;d,electrodedistanceis
1 cm; T, temperature is 25◦C.
Downloaded 19 Nov 2011 to 212.50.248.4. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions

Page 5

055114-4M. Sharifian Ghahfarokhi and T. KhayamianRev. Sci. Instrum. 82, 055114 (2011)
Flowmeter
A/D Converter
Amplifier
GC
R=33 KΩ
R=12 MΩ
H. V. Power Supply
Pure N2
Heater
Exit
FIG. 4. (Color online) Schematic diagram of the corona discharge detector
coupled to the gas chromatograph.
chloroform in toluene with two different chloroform concen-
trations are shown in Fig. 5. The chloroform peaks were orig-
inated from chloroform solutions with 78.1 and 840 ng/ml
concentrations, respectively. In Fig. 5(a), The first peak orig-
inated from the highest chloroform concentration in the dy-
namic range of the calibration curve. The second peak origi-
nated from the solvent, toluene (1 μl), with the concentration
of more than 1 × 106times larger than the highest concen-
tration of chloroform. This peak might be originated from the
huge vapor of the solvent around the needle tip and conse-
quently affect the corona discharge.
TABLE I. The optimized conditions of the GC and the continuous neg-
ative corona discharge cell for injection of 1 μl solution of chloroform in
toluene to gas chromatography.
GC column temperature100◦C
GC injection port temperature
GC column gas flow rate
Corona cell temperature
Corona electrode distance
Corona gas flow rate
Corona discharge electrical voltage
160◦C
50 ml/min
70◦C
10 mm
50 ml/min
8000 V
E. Calibration curve and analytical parameters
Figure 6 shows the calibration plot of chloroform in
toluene solution. The detection limit was obtained 0.78 ng/ml
and the linear dynamic range was in the range of 2–840 ng/ml
with a limit of quantification of 2 ng/ml. The coefficient of
determination (R2) of the plot was 0.9923. The relative stan-
dard detection was about 6% for the limit of quantification.
The results show the capability of the detector for determi-
nation of a very low concentration of chloroform. In order to
compare the performance of this detector with the ECD detec-
tor, the results indicate that the dynamic range in this detector
is wider than that in the ECD, however the detection limit is
higher.30,31Since, the electron density in continuous negative
corona discharge is much higher than that in63Ni, therefore
we could expect to obtain a lower detection limit and wider
dynamic range for the proposed method. The dynamic range
is wider, because in the continuous corona discharge, enough
(a)
(b)
FIG. 5. Signal obtained from injecting 1 μl solution of chloroform in toluene with different chloroform concentrations to the gas chromatograph.(a) Chloroform
concentration, 78.1 ng/ml; gain, 10;(b) chloroform concentration, 840 ng/ml; gain, 2.
Downloaded 19 Nov 2011 to 212.50.248.4. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions

Page 6

055114-5M. Sharifian Ghahfarokhi and T. KhayamianRev. Sci. Instrum. 82, 055114 (2011)
CHCL3 (ng/mL)
Peak Area (V.Sec)
FIG. 6. (Color online) The calibration curve of chloroform.
electrons are available for analytes to capture electron, but
electron generation by63Ni is restricted. The reason for the
higherdetection limitisprobablyrelatedtonecessityformore
optimization of the variables. For example, by reducing the
dimensions of the detector or volume of it, the peak broad-
ening reduces and therefore, the performance of the detector
improves. Furthermore, it would be possible to use hardware
or software method to reduce noise and to enhance the signal
to noise ratio. Finally, it is necessary to emphasize that the
main advantage of the proposed detector relative to the ECD
is attributed to its safety.
V. CONCLUSION
A new simple detector based on continuous negative
corona discharge was designed and constructed. This detector
has ability to be used as a separate unit for analysis of any
compound containing an electronegative element or it can
be used as a detector in GC. Furthermore, the detector has
potential to be constructed potable for using in field. The
performance of the detector was evaluated by the determi-
nation of the analytical parameters of chloroform as the test
compound. The results demonstrated the applicability of
the detector for analysis of organic compounds containing
electronegative elements and its ability to be used as an
alternative detector. In addition to its excellent performance,
its safety adds another advantage relative to the detectors and
used radioactive materials.
1C. S. Leasure, M. E. Fleischer, G. K. Anderson, and G. A. Eiceman, Anal.
Chem. 58, 2142 (1986).
2W. E. Wentworth, J. Chromatogr. A 112, 229 (1975).
3M. A. Baim, R. L. Eatherton, and H. H. Hill, Jr., Anal. Chem. 55, 1761
(1983).
4K. N. Vora, D. N. Cambell, R. C. Davis, G. E. Spangler, and
J.A.Reateni, Europeanpatent
1994).
5J. R. Maldonado and S. T. Coyle, U.S. patent 7,015,467 (21 March
2006).
6V. L. Budovich, A. A. Mokhailov, and G. Arnold, U.S. patent 6,023,169 (8
February 2000).
7J. Adler, G. Arnold, and H. R. Doring, in Proceedings of the Third In-
ternational Workshop on Ion Mobility Spectrometry Dresden, Germany,
1997.
8E. M. Veldhuizen and W. R. Rutgers, in Proceedings of the Frontiers in
Low Temperature Plasma Diagnosis IV (Rolduc, The Netherlands, 2001),
p. 40.
9H. Cai, S. D. Stearns, and W. E. Wentworth, Anal. Chem., 70, 3770
(1998).
10P. Stanislav, in Proceedings of the 27th EPS Conference on Controlled
Fusion and Plasma Physics Budapest [Fusion Plasma Phys. 24B, 472
(2000)].
11C. S. Ren, T. C. Ma, D. Z. Wang, J. L. Zang, and Y. N. Wang, J. Electrostat.
64, 23 (2006).
12A. Jaworek and A. Krupa, J. Electrostat. 38, 187 (1996).
13R. Turner, S. J. Tailor, A. Clark, and P. D. Arnold, European patent appli-
cation WO 9728444 (1997).
14J. Adler, G. Arnold, H. R. Doring, V. Starrock, and E. Wulfing, in Pro-
ceedings of the 6th International Workshop on Ion Mobility Spectrometry,
Bastei, Germany, 1997, p. 109.
15J. M. Meek and J. D. Craggs, Electrical Breakdown of Gases (Wiley,
Chichester, 1978).
16Y. P. Raizer, Gas Discharge Physics (Springer, Berlin, 1991).
17M. Tabrizchi, T. Khayamian, and N. Taj, Rev. Sci. Instrum. 71, 2321
(2000).
18G. L. Weissler, Phys. Rev. 63, 96 (1943).
19S. C. Haydon and O. M. Wiliams, J. Phys. B 6, 227 (1973).
20H. Koge, Ph.D. dissertation, Physica Universitatis Tartuensis (Tartu,
Estonia, 1992).
21S. Sakata and T. Okada, J. Aerosol Sci. 25, 879 (1994).
22H. C. Mishra, Fundamental Inorganic Chemistry (Motilal Banarsidass,
Delhi, 1981).
23E. C. Huang, T. Wachs, J. J. Conboy, and J. D. Henion, Anal. Chem. 62,
342 (1990).
24W. E. Wentworth, C. F. Batten, and D. E. Desai, J. Chromatogr. 390, 249
(1987).
25M. W. Siegel and W. L. Fite, J. Phys. Chem. 80, 2871 (1976).
26C. J. Proctor and J. F. J. Todd, Org. Mass Spectrom. 18, 509 (1983).
27J. S. Townsend and P. J. Edmunds, Philos. Mag. 28, 789 (1914).
28P. P. M. Blom, Ph.D. dissertation (Eindhoven University of Technology,
1997).
29N. Y. Bahaeva and G. V. Naidis, in Modeling of Streamer Propagation in:
Electrical Discharges for Environmental Purposes, edited by E. M. van
Veldhuizen (Nova Science Publisher, New York, 2000), p. 21.
30D. A. Skoog, F. J. Holler, and T. A. Nieman, Principles of Instru-
mental Analysis, 5th ed. (Saunders College Publishing, Philadelphia,
1997).
31M. Schellin and M. Popp, J. Chromatogr. A 1103, 211 (2006).
EP0198154 B1(30 November
Downloaded 19 Nov 2011 to 212.50.248.4. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions