http://www.estcal.com/TechPapers/Food/Ethanol.doc
Detection of Ethanol in Water and Air
Using the zNose
Edward J. Staples, Electronic Sensor Technology
Electronic Noses
Conventional electronic noses (eNoses) produce a recognizable response pattern
using an array of dissimilar but not specific chemical sensors. Electronic noses have interested developers of neural networks and artificial intelligence algorithms for some
time, yet physical sensors have limited performance because of overlapping responses
and physical instability. ENoses cannot separate or quantify the chemistry of aromas.
A new type of electronic nose, called the zNose, is based upon ultra-fast gas
chromatography, simulates an almost unlimited number of specific virtual chemical sensors, and can produce high-resolution two-dimensional olfactory images based upon
aroma chemistry. The zNose is able to perform analytical measurements of volatile
organic vapors and odors in near real time with part-per-trillion sensitivity. Separation
and quantification of the individual chemicals within an odor is performed in seconds.
Using a patented solid-state mass-sensitive detector, picogram sensitivity, universal nonpolar selectivity, and electronically variable sensitivity is achieved. An integrated vapor
preconcentrator coupled with the electronically variable detector, allow the instrument to
measure vapor concentrations spanning 6+ orders of magnitude. In this paper a portable
zNose, shown in Figure 1, is shown to be a useful tool for quantifying the concentration
of ethanol in air and water samples.
Figure 1- Portable zNose technology incorporated into a handheld instrument
1
http://www.estcal.com/TechPapers/Food/Ethanol.doc
How the zNose™ Quantifies the Chemistry of Aromas
A simplified diagram of the zNose™ system shown in Figure 2 consists of two sections. One section uses helium gas, a capillary tube (GC column) and a solid-state detector. The other section consists of a heated inlet and pump, which samples ambient air.
Linking the two sections is a “loop” trap, which acts as a preconcentrator when placed in
the air section (sample position) and as an injector when placed in the helium section (inject position). Operation is a two step process. Ambient air (aroma) is first sampled
and organic vapors collected (preconcentrated) on the trap. After sampling the
trap is switched into the helium section
where the collected organic compounds are
injected into the helium gas. The organic
compounds pass through a capillary column
with different velocities and thus individual
chemicals exit the column at characteristic
times. As they exit the column they are detected and quantified by a solid state detector.
An internal high-speed gate array miFigure 2- Simplified diagram of the zNose™
croprocessor controls the taking of sensor
showing an air section on the right and a hedata which is transferred to a user interface
lium section on the left. A loop trap preconor computer using an RS-232 or USB concentrates organics from ambient air in the
nection. Aroma chemistry, shown in Figsample position and injects them into the heure 3, can be displayed as a sensor spectrum
lium section when in the inject position.
or a polar olfactory image of odor intensity
vs retention time. Calibration is accomplished using a single n-alkane vapor standard. A
library of retention times of known chemicals indexed to the n-alkane response (Kovats
indices) allows for machine independent measurement and compound identification.
Figure 3- Sensor response to n-alkane vapor standard, here C6-C14, can be
displayed as sensor output vs time or its polar equivalent olfactory image.
2
http://www.estcal.com/TechPapers/Food/Ethanol.doc
Chemical Analysis (Chromatography)
The time derivative of the sensor spectrum (Figure 3) yields the
spectrum of column flux, commonly
referred to as a chromatogram. The
chromatogram response (Figure 4) of
n-alkane vapors (C6 to C14) provides
a set of reference retention times.
Graphically defined regions, shown
as red bands, provide a method dependent reference time base against
which subsequent chemical responses
can be compared and indexed. As an
example, a response midway between
C10 and C11 would have a retention
time index of 1050.
Figure 4 - Chromatogram of n-alkane vapors C6 to C14).
Properties of Ethanol
Physical properties of ethanol are shown in Figure 5. Although it is relatively volatile it also is very soluble in water which accounts for the low Henry’s constant e.g. at
room temperature, only 0.0206% partitions into the air from water as headspace vapor.
Figure 5-Physical properties of Ethanol.
3
http://www.estcal.com/TechPapers/Food/Ethanol.doc
Ethanol Standards
A stock solution of pure
(neat) ethanol can be used to
produce vapor and water
standards. A vapor standard
is produced by injecting a
measured amount of pure ethanol into a tedlar bag filled
with 1 liter of air. Water
standards using headspace
measurements can be created
by injecting pure ethanol into
20 milliliters of water in a
sealed 40-milliliter vial.
Figure 6- Vapor and water ethanol calibration standards using tedlar bags and 40-milliliter septa-sealed vials.
System Settings for Ethanol
Due to its low molecular
weight and high volatility, ethanol is
at the lower limit of detectable compounds which can be analyzed by the
zNose. Sensitivity is determined by
the loop trap’s ability to concentrate
ethanol from air samples and the ability of the SAW sensor to condense
and detect it as it elutes from the GC
column. To optimize the trapping
efficiency of the loop trap system
temperatures are necessarily low.
Maximum SAW sensitivity is
achieved by using a crystal temperature of 10oC or 0oC. A valve temperature of 70oC and an inlet temperature of 50oC are recommended. Because the loop trap contains only 1
mg of Tenax®, the breakthrough volume is relatively low and sampling Figure 7- Detected counts vs sample time for two different
times greater than 5 seconds are inefsystem temperatures.
fective. The effects of system temperatures and sample time are shown
in Figure 5.
4
http://www.estcal.com/TechPapers/Food/Ethanol.doc
Use of Pre-filter with the zNose®
The zNose® is a very sensitive gas chromatograph designed to detect vapors covering a range of approximately C4 to C24 n-alkanes. Particles, liquids, and high molecular
weight compounds with high boiling points if ingested can contaminate and in some cases damage the instrument. To prevent this from occurring care should be taken to insure
that direct injections of liquid standards do not contain high molecular weight compounds
and are less than 1 microliter. Similarly when sampling air in dusty environments a prefilter should be used to prevent particles from entering the inlet of the instrument.
Sampling headspace vapors above liquid samples should be done with great care so
as to avoid sucking liquids or droplets into the instrument. Vapors from water samples
heated to approximately 40oC normally do not cause any significant problem other than
perhaps a higher detector temperature and a 1-2 second wait state before firing the trap.
However, heating of water samples above this can produce very high concentrations of
condensable water vapor leading to the formation of droplets in the relatively cool sampling needle and loop trap. For example, water at 60oC produces a vapor concentration
of 130 micrograms/milliliter and preconcentration of even 1 milliliter of vapor will form
droplets within the sampling needle and loop trap. When the sample is transferred to the
helium section and the trap rapidly heated to 250oC during injection, the water vaporizes
and expands more than 1000X. This in turn creates a very high-pressure burst of super
heated steam when passes through the column and strikes the surface of the SAW detector. Often this destroys the electrodes on the surface of the crystal and the detector must
be replaced.
To prevent high concentrations of water vapor and particles from entering the instrument an inert PTFE pre-filter, shown in Figure 8, is recommended. The filter shown
comes in a plastic housing with luer fittings and is easily attached to the inlet of the instrument with a simple luer adaptor. Because the filter housing is plastic inlet temperatures should always be less than 70oC.
Figure 8- Pre-filters using inert PTFE.
5
http://www.estcal.com/TechPapers/Food/Ethanol.doc
Detecting Ethanol in Ambient Air
To calibrate the instrument for detecting ethanol in air, standard vapors are first
created by injecting known amounts of ethanol into a known volume of air in a tedlar
bag. For example injecting 1 L of pure ethanol into
1000 mL of air will produce 415.3-ppmv of ethanol vapor (789 nanogram/milliliter). Using a side-ported
sampling needle and a PTFE pre-filter attached to the
inlet of the zNose (Figure 9), standard vapors are then
sampled to determine linear response factors, method
detection limits (mdl), and calibrate the instrument. Calibration settings, which are method dependent, allow the
instrument to display air concentrations as ppmv values.
As an example, the ethanol response using a 415 ppmv
Figure 9- Ethanol Vapor Standard
vapor standard is compared to that of n-alkane vapors
(C6-C10) using a 1ps3a1b method in Figure 10. The system software correctly displays
the Kovats index of ethanol as 552 and concentration as 415 ppm
The zNose® system response is linear over a wide range of vapor concentrations.
This is clearly shown by the N-point responses obtained from a series of 1-liter tedlar
bags injected with increasing amounts of ethanol as shown in Figure 11.
Figure 10- Calibrated system response to 415 ppm vapor standard. Alkane response C6-c10 is shown overlaid in Red.
Figure 11-N-point calibration results using 10o detector, 70o valve, 50o inlet, and 1ps3a1b method.
6
http://www.estcal.com/TechPapers/Food/Ethanol.doc
Detector Sensitivity vs Temperature
System sensitivity is controlled by sample preconcentration time and detector temperature. Detector sensitivity is an exponential function of detector temperature as shown
in Figure 12 and maximum sensitivity is achieved with a minimum detector temperature
of 0oC. At very high water vapor concentrations in the sampled air the detector may experience overloading at the lowest temperature settings.
Figure 12- Sensitivity is an exponential function of detector temperature.
Ethanol Minimum Detection Limit
A measurement detection limit is defined by replicate measurements at the lowest
quantifiable level. Statistically this is defined for 7 replicate measurements as 3.14 x
standard deviation. Using a 1ps2a1b method, 70oC valve, 50oC inlet, and 10oC detector
temperatures, 7-replicate measurements at 100 and 13-ppm ethanol concentrations yield
MDLs of 14.1 and 5.2 ppm respectively as shown in Figure 13. .
Figure 13- Method Detection Limits for ethanol.
7
http://www.estcal.com/TechPapers/Food/Ethanol.doc
Detection of Ethanol in Water
Headspace vapors from water
standards (20 mL water in a 40 mL vial) were sampled directly using a sideported sample needle and PTFE filter
attached to the inlet of the zNose as
shown in Figure 14. The concentration
of ethanol in headspace vapors from
water at room temperature is very low
because ethanol is very soluble in water. Heating the water sample using a
two-zone (top and bottom) vial heater
can greatly increase headspace concentration. For example a 300-ppm water
standard at 40oC had a response of 284 Figure 14- Direct headspace sampling of heated water
counts or 104 ppmv. Increasing the water temperature to 65oC increased the response to
957 (353 ppmv), an increase of more than 300%.
Replicate direct sample measurements (offset in x and y-directions) are shown in
Figure 15 for a 100-ppm water standard at 65oC. The high degree of reproducibility at
this level is shown in the inset showing a standard deviation of approximately 6% for 15
replicate measurements.
Figure 15- Replicate runs on 100 ppm water sample at 65oC.
8
http://www.estcal.com/TechPapers/Food/Ethanol.doc
Ethanol headspace measurements were linear over a wide range of water concentrations as evidenced by N-point calibration curves shown in Figure 16. In this case a series
of water standards at 50, 100, 150, 200, 250, and 300 ppm were measured in triplicate
with 10oC and 20oC detector settings, a 1ps3a1b method, 70oC valve, and 50oC inlet.
Each data point is the average of three triplicate measurements. The averaged data points
and linear response factor (counts/ppm) for each case is shown as insets in the figure.
Figure 16- N-point calibration curves for water standards at 65oC.
9
http://www.estcal.com/TechPapers/Food/Ethanol.doc
Summary
Detecting ethanol in air and water is fast and easy using the zNose electronic
nose or portable gas chromatograph. Even though ethanol is at the lower limit of detectable compounds, concentrations well into the low part-per-million range can be
quantified with good precision and accuracy. Because ethanol is very soluble in water,
headspace measurements are best performed with water samples elevated to at least 40oC.
Use of a PTFE inlet filter is recommended to prevent water droplets from forming,
entering the instrument, and possibly damaging the sensitive vapor detector. A summary
chart of ethanol MDL amounts is shown in Table I.
Table I - Ethanol Method Detection Limits
Method
MDL
Direct Sampling in Air
<5 ppmV
Water headspace Sampling
<5 ppm
For good sensitivity to very volatile organic compounds (VVOC), a valve
temperature of 70oC and an inlet temperature of 50oC works well. Although the column
can be operated isothermally, better chromatography is achieved with at least a
1oC/second temperature ramping of the column. Sample times longer than 5 seconds are
ineffective due to ethanol breakthrough in the tenax filled loop trap. Detector sensitivity
is an exponential function of temperature and a minimum detector temperature setting of
0oC results in maximum sensitivity. A linear response is achieved from 0 to over 3000
ppmv of ethanol. Ambient air measurements take less than 15 seconds and can be
repeated at one minute intervals.
10