Content uploaded by Frans Harren
Author content
All content in this area was uploaded by Frans Harren
Content may be subject to copyright.
TB, DC, PM/227563, 9/11/2006
INSTITUTE OF PHYSICS PUBLISHING PHYSIOLOGICAL MEASUREMENT
Physiol. Meas. 27 (2006) 1–12 UNCORRECTED PROOF
The suitability of Tedlar bags for breath sampling in
medical diagnostic research
Marco M L Steeghs, Simona M Cristescu and Frans J M Harren
Life Science Trace Gas Facility, Molecular and Laser Physics, Institute for Molecules and
Materials, Radboud University, Nijmegen, The Netherlands
E-mail: m.steeghs@science.ru.nl
Received 28 June 2006, accepted for publication 31 October 2006
Published DD MMM 2006
Online at stacks.iop.org/PM/27/1
Abstract
Tedlar bags are tested for their suitability for breath sampling for medical
diagnostic purposes. Proton-transfer reaction-mass spectrometry was used to
monitor the changes in composition of various mixtures contained in custom-
made black-layered Tedlar bags. Characteristic ions at m/z 88 and 95 amu
reflect considerable pollution from the bag material. The pollutant found on
m/z 88 amu is most probably N,N-dimethylacetamide, a latent solvent used
in the production of Tedlar film. Gas composition losses during filling were
found to range from 5 to 47%, depending on the compound. Once stored, the
half-lives of methanol, acetaldehyde, acetone, isoprene, benzene, toluene and
styrene were estimated between 5 and 13 days. Losses from breath samples
(52 h after filling) were found to be less than 10%. No observable decrease
was found for ethylene over 3 days, using laser-based photoacoustic detection.
For the use of Tedlar bags, a standardized protocol is advised, where the time
point of analysis is fixed for all samples and should be kept as close as possible
to the time of sampling.
Keywords: PTR-MS, breath analysis, Tedlar bags, VOCs
1. Introduction
The study of biological processes via the emission of volatile organic compounds (VOCs)
is a widely used and accepted method in Life Sciences, including different fields of biology
(Boschetti et al 1999, Phillips et al 1999,Harrenet al 1999, Woltering et al 1989, Steeghs
et al 2004, Holzinger et al 2000), environmental monitoring (Hewitt et al 2003, Warneke
et al 1996) and medical diagnostics (Phillips 1992, Berkelmans et al 2003, Jordan et al 1995,
Karl et al 2001, Lechner et al 2005, Dannecker et al 1981, Steeghs et al 2006). Gas analysis
is also becoming an important tool in control of industrial processes such as combustion and
0967-3334/06/000001+12$30.00 © 2006 IOP Publishing Ltd Printed in the UK 1
2 M M L Steeghs et al
plasma diagnostics, investigation of engines or automobile exhaust measurements or headspace
analysis during food production (Werle 2003, Lindinger 2005, Jobson 2005). Several
publications have shown the diagnostic potential of relatively new, advanced techniques.
These include gas chromatography–mass spectrometry (GC–MS) (Phillips and Greenberg
1991), selected ion flow tube-mass spectrometry (SIFT-MS) (Wang et al 2004,Davies
et al 2001) and proton transfer reaction-mass spectrometry (PTR-MS) (Lindinger et al 1998a,
Prazeller et al 1998, Critchley et al 2004). The latter two techniques are very promising in
breath analysis and fast diagnostics, since they combine on-line measurement capacity with
non-invasive, sensitive and multi-component detection of VOCs.
It is not always possible to bring the gas emission source of interest and the gas detection
system together. In such cases, the gas sample under analysis needs to be stored in a container
and transported to the laboratory. For this, several methods are used, including canisters,
cold trapping, adsorbing agents and Teflon and Tedlar bags. These containers should be
easy to operate, easy to handle and be able to store a sample for a prolonged period of time.
Costs, re-usability, durability, size and versatility are other important issues when choosing a
sample holder. Adsorbing agents like Tenax are compound specific, whereas canisters have
the disadvantage that they need to be evacuated before sampling, making them inapplicable to
breath measurement. Teflon bags are relatively expensive and fragile, making them difficult
to be re-used. Here, as an alternative to these widely used methods we have investigated the
use of Tedlar bags for the sampling of breath from patients and healthy volunteers.
Tedlar bags have been used previously by Nielsen and co-workers (Nielsen 2002), who
have measured nine volatile sulphur compounds from biogas-production plants and reported
no detectable losses over a period of 20 h. Only water was found to permeate through the
walls until the relative humidity inside reached ambient air values. Sulyok et al (Sulyok
2001) observed a 90% recovery rate of sulphuric compounds after 1 week of containment
for clear layered Tedlar bags, but found significantly lower recovery rates for black-layered
bags. They assumed the losses to be due to uncontrolled adsorption to the carbon-filled black
layer. Lau and co-workers (Lau 1989) reported half-lives of six sulphur gases ranging from
0.02 days to 100 days. For acetone, a significant decrease after the first 6 h was found, using
standard Tedlar bags (Deng 2004). The same result has been obtained for isoprene levels
(Hyspler 2000). N-hexene, p-xylene and 1-propanol were stable for at least 30 min (Myung
1997).
Groves et al (Groves and Zellers 1996) studied the significance of water content in the
sampled air for the stability of methanol, acetone, 2-butanone, m-xylene, trichloroethane and
perchloroethylene. They spiked the bags already containing the sampled air with water to
supersaturation to invoke condensation in the bag. Significant differences between wet and
dry samples were only found for methanol, acetone and 2-butanone. However, the relative
humidity needed to induce this difference was about three times higher than the humidity
inside the bag expected from breath (Groves and Zellers 1996). Only methanol was slightly
affected at breath humidity levels.
These studies mostly used transparent Tedlar bags and focused on isolated compounds
or a complex of several sulphur compounds. It is not clear how this can be translated to
breath samples and to black-layered Tedlar bags. To our knowledge, the storage capabilities
of these bags for most breath VOCs have not been tested. Here, we test the changes in gas
composition of breath and controlled mixtures during storage in black-layered Tedlar bags.
PTR-MS, of which the potential for breath measurements was shown before (Jordan et al 1995,
Karl et al 2001, Lechner et al 2005), was used for the analysis of more complex mixtures,
including breath, while photoacoustic spectroscopy was used to monitor the behaviour of
ethylene (Harren and Reuss 1997, Harren et al 1999). The contents of bags with several
The suitability of Tedlar bags for breath sampling in medical diagnostic research 3
Figure 1. Schematic view of the PTR-MS instrument. (1) is the hollow cathode ion source,
(2) the drift tube, (3) the transition chamber, (4) the detection chamber containing the quadrupole
and the secondary electron multiplier (5).
different mixtures were analysed and compared at several instances between 4 and 250 h after
filling.
2. Materials and methods
2.1. Instrumentation
The proton transfer reaction–mass spectrometer (PTR-MS) used for this work is analogous to
that described in Lindinger et al (1998b) (figure 1). A detailed description of the home-built
system can be found in Boamfa et al (2004) and Steeghs et al (2006). The working principles
of PTR-MS have been given in detail elsewhere (Lindinger et al 1998b, Boamfa et al 2004,
De Gouw et al 2003). Therefore, only a brief description is given here.
The instrument consists of four parts: an ion source where H
3
O
+
ions are produced, a drift
tube section, a separately pumped transition chamber and an ion detection section containing
a quadrupole mass spectrometer and a secondary electron multiplier. In the drift tube, the
trace gases from the sample gas are ionized by proton-transfer reactions with H
3
O
+
ions:
H
3
O
+
+R
k
−→ RH
+
+H
2
O(1)
where k is the reaction rate constant, usually close to or equal to the collision rate constant.
This reaction only takes place when the proton affinity (PA) of the trace compound R is higher
than that of water (166.5 kcal mol
−1
= 7.16 eV/molecule). A major advantage of using H
3
O
+
as the reagent ion is that the PA of water is higher than the PA of the normal constituents
of air (cf N
2
,O
2
,CO
2
,CH
4
, CO, NO, Ar) and that most of the typical organic compounds
are ionized by the proton-transfer (PT) reaction, since their PA are in the range between
7 and 9 eV. The reaction rate can be measured or calculated and is known for many of the PT
reactions of interest (http://Webbook.Nist.Gov/Chemistry/, Lindinger et al (1998b)). Since
the excess energy of the reaction is low, it results in only one or two characteristic ions per
neutral molecule. For some molecular groups dissociation can occur to form one or two
fragments of significant intensity (e.g. alcohols can split off a water molecule, which results
in a fragment ion at molecular mass minus 17). Due to this soft-ionization the matrix of
signals is less complicated than with other mass spectrometry techniques. One disadvantage
of the technique is that it does not provide positive identification of the compounds monitored
(Steeghs et al 2004).
4 M M L Steeghs et al
Ethylene, a well-known indicator for lipid peroxidation (Kneepkens et al 1994)was
measured using laser-based photoacoustic detection, as described in Harren and co-workers
(Harren et al 1999, Harren and Reuss 1997). Briefly, the infrared laser is tuned to a wavelength
where the gas under investigation has a high absorption coefficient. By modulating the intensity
of the laser light, a periodic heating of the sample gas is caused. This results in pressure
variations inside the sample cell of the same frequency as the laser intensity modulation. By
carefully choosing the modulation period, an acoustic wave is created inside the closed sample
cell, which is detected with a microphone. The amplitude of the acoustic wave is directly
proportional to the laser intensity and to the concentration of the absorbing compound. By
monitoring the laser power and the intensity of the sound wave, the concentration of the
compound under investigation can be determined.
For both systems, calibration of concentrations was performed using certified mixtures.
For the PTR-MS system we used range of dilutions of a mixture containing 600 ppbv of
methanol, 800 ppbv of acetaldehyde, 900 ppbv of acetone and 1000 ppbv of isoprene, benzene,
toluene and styrene in nitrogen. For the photoacoustic system the same procedure was used
with a mixture containing 1 ppmv of ethylene in air.
2.2. Tedlar bags
We used 1.0 l black-layered Tedlar bags (SKC Limited, UK), especially designed for single-
breath sampling experiments. The inner bag is made of a clear thin film of Tedlar and is
covered by an outer layer of black Tedlar. This outer layer is blackened by adding carbon to
the film to avoid UV-induced breakdown of any of the compounds possibly contained in the
bag. Every bag has two Teflon hose/valve septum fittings, which simplifies the automated
analysis and cleaning of the bags before re-use.
2.3. Experiments
All bags were flushed with synthetic air (mixture of purified nitrogen and 20 ± 1% purified
oxygen; <3 ppm H
2
O, <0.5 ppm C
n
H
m
; Air Liquide BV, Eindhoven, The Netherlands) with a
flow of ∼25 l h
−1
for more than 2 h before use. Tedlar bags were filled with various mixtures
of air using mass flow controllers (Brooks Instruments, 850S) to compare the composition of
the contents at several points in time within a period of 10 days. Since the drift tube of the
PTR-MS is operated at a pressure of around 2.25 mbar, no flow controllers or pumps were
needed to measure contents of the bags. The sampling lines to and from the bags as well
as the in-stream PTR-MS inlet were made of Teflon (PTFE and PFA, Polyfluor Plastics, The
Netherlands) to minimize losses due to sticking of compounds to the walls. The calibrated
mixture through the flow controller is kept constant at 0.2 l h
−1
at all times, so equilibrium
situation is guaranteed and no losses occur here either.
We investigated whether any artefacts (increase of concentrations of known or unknown
compounds) could be observed and whether there were losses to or through the bag wall. The
influence of humidity on the stability of the contents was tested by varying the humidity of
the sampled mixtures. Some bags were heated (T ∼ 60
◦
C) to see if the temperature had an
effect on the concentrations found from the bag contents.
2.3.1. Pollution of contents. A total of eight bags were filled with pure nitrogen (4.5 purity
level) or synthetic air. These gas mixtures contained no detectable concentrations of
hydrocarbons (at the (sub-) ppbv level) before filling. The contents of these bags were
monitored to check for increase in concentrations from polluting compounds, either by
The suitability of Tedlar bags for breath sampling in medical diagnostic research 5
permeation through the walls of the bags, or by production from the bag material. These
and all other measurements were performed at laboratory temperature, which is controlled at
22
◦
C, unless stated otherwise. Bags were always stored at lab temperature.
2.3.2. Loss of contents. The stability of a controlled mixing ratio of hydrocarbons was
checked by filling 12 bags with a certified mixture (600 ppbv of methanol, 800 ppbv of
acetaldehyde, 900 ppbv of acetone, 1000 ppbv of isoprene, benzene, toluene and styrene,
Scott Specialty Gases, USA), diluted to a mixing ratio of 1 in 5 with nitrogen.
Also, a total of 24 bags were filled with the breath of 9 healthy volunteers. The contents of
10 of these bags were monitored 6 times in a period of 250 h; the other 14 bags were analysed
5 times within 72 h. It should be noted that condensation of water contained in the exhaled
breath occurs in the sampling device before the bag, so no water droplets are formed inside
the bag.
To test the suitability of these Tedlar bags for ethylene, 3 bags were filled with room
air, containing low amounts of ethylene (below 1 ppbv), another 3 bags were filled with
breath and 3 bags were filled with varying higher amounts of ethylene (65–100 ppbv). The
concentrations in the bags at 1 h after filling were compared to the concentrations at 72 h after
filling.
Possible losses could occur due to permeation through the walls or due to adhesion to
the walls. To test whether compounds permeate through the wall, three bags were filled with
a calibrated mixture (containing methanol at 600 ppbv, acetaldehyde at 800 ppbv, acetone at
900 ppbv and isoprene, benzene, toluene and styrene at 1000 ppbv). These bags were placed
in a larger glass cuvette. The glass cuvette was flushed with nitrogen (4.5 purity level) gas
for 10 min via an inlet and outlet before sealing, to make sure that no detectable hydrocarbon
levels surrounded the bags at the moment the experiment started. After 2, 3 and 7 days, the
contents of the glass cuvette were measured by PTR-MS and compared to nitrogen values and
lab air values. After 7 days the contents of the bags were also analysed.
2.3.3. Relative humidity. Relative humidity was increased to 80% for 6 of the 12 bags by
mixing the calibrated mixture with humidified (100% RH) nitrogen gas. The latter was done
by bubbling the nitrogen gas through a water reservoir, resulting in a RH of 80% for the
calibrated mixture. These experiments were performed to see if the water vapour in the bag
affects the lifetime of compounds inside. The relative humidity is monitored by PTR-MS via
the amount of cluster formation at mass 37 amu, where the water clusters H
3
O
+
· H
2
O can be
observed.
All bags were measured several times over a prolonged period of time. Due to the limited
volume of the bag (1 litre), we could only sample 5 or 6 times. Each time the bag contents
were analysed, 8 mass scans from 20 amu to 150 amu were made (1 amu s
−1
scan rate)
and averaged. The bags were stored and measured at laboratory temperature, which is set to
22
◦
C.
3. Results and discussion
3.1. Pollution of contents
Several bags were filled with clean air or pure nitrogen. These bags displayed significant
amounts of pollution. Figure 2 (left panel) shows pollution reflected by characteristic ions
at masses 88 and 95 amu where significant increases were found. Every single bag used in
this study showed an increase of these two masses. The compound found at mass 88 is most
6 M M L Steeghs et al
Figure 2. Pollution of synthetic air and nitrogen gas (4.5 purity). Left panel: increase in masses
88 and 95 due to pollution from the Tedlar material (background values are zero). The large error
bars are caused by large differences in amounts of pollution coming from the respective bags.
Right panel: increase in masses 33, 45 and 59 (corresponding to expected compounds methanol,
acetaldehyde and acetone, respectively) concentrations (background values subtracted). The large
errors are caused by the low concentrations of the compounds measured and the variation between
different bags.
probably N,N-dimethylacetamide (C
4
H
9
NO, MW = 87 amu, PA 217.0 kcal mol
−1
) and was
found to originate from the Tedlar film (Chase 2001, Du Pont 1964). The Tedlar bag is made
out of polyvinyl fluoride film, which is produced from polyvinyl fluoride powder, which is
dissolved in a ‘latent solvent’ at temperatures above 100
◦
C, so the particles are allowed to
coalesce into a film before the solvent evaporates (Chase 2001). Several candidates for this
latent solvent are listed, among which are N,N-dimethylacetamide and related compounds
(Du Pont 1964). After the drying process the total amount of latent solvent left in the material
is assured to be less than 0.5% (Du Pont 1964). The concentrations of N,N-dimethylacetamide
ranged from 40 to 185 ppbv (this is a lower limit, calculated using k = 2 × 10
−9
cm
2
s,
assuming no fragmentation) at mass 88 after 4 h and increased in time. Clean air and nitrogen
contained no measurable quantities of N,N-dimethylacetamide (<0.5 ppbv). The manufacturer
of the bags assured that besides N,N-dimethylacetamide, no other solvent or chemical was
used to make the bags. Our results, however, show that a second pollutant is present at mass
95. Independent GC-MS analysis has identified the compound at mass 88 amu to be indeed
N,N-dimethylacetamide and the compound at mass 95 amu to be phenol (Amann, private
communication, Di Francesco, private communication). Heating of the bags increased the
See endnote 1
signals on both mass 88 and 95 amu significantly (∼170% and 90% respectively, results not
shown).
Some other masses also show a slight increase in ion intensity. This much lower degree
of pollution was only observed for those bags containing synthetic air or nitrogen (figure 2,
right panel). Mass 59 was found to be increased to about 5 ppbv after 250 h. It should be
noted that these values are still less than 10% of the values expected in human breath and
the influence will therefore be within the system’s uncertainty of the determination of breath
concentrations.
The suitability of Tedlar bags for breath sampling in medical diagnostic research 7
Figure 3. Relative decreases in signals for seven different gases and relative humidity as a function
of time. Values are normalized to the signals found in the sample flow immediately before filling.
Humidity is monitored by the number of water clusters H
3
O
+
· H
2
O at mass 37 amu. All values
are normalized to values at t = 0h.
3.2. Loss of contents
Losses of compounds contained in a Tedlar bag were monitored by sampling controlled
amounts of calibrated mixture and breath. Figure 3 shows the evolution of the concentrations
of the various compounds from the calibrated mixture. The values at 0 h are the mixed flows
used to fill the bags.
There is a considerable loss ranging from 5% up to 47% at 4 h after filling. We consider
these losses to occur during filling, presumably due to sticking to the septum in the inlet valves,
since over time we find only a gradual decrease with respect to those values measured after
4 h. From the seven compounds, styrene is the only compound found to decrease for more
than 15% (30 ± 7%) between 4 h and 52 h. From the decline of their concentration between
4 h and 72 h after gas sampling, a half-life for the respective compounds can be calculated.
These half-lives are 8.1 ± 0.6 days for methanol, 6.5 ± 0.3 days for acetaldehyde, 8.4 ±
0.6 days for acetone, 13 ± 0.8 days for isoprene, 8.2 ± 0.8 days for benzene, 8.4 ± 0.7 days
for toluene and 5.9 ± 0.3 days for styrene.
Other authors (cf Sulyok (2001)) have found lower recovery rates in black-layered Tedlar
bags, as compared with clear Tedlar bags. In this study we aimed at characterizing black-
layered bags and no attempt was made to compare both black-layered and clear Tedlar bags.
Low ethylene levels in the bags displayed no measurable decrease over 3 days. These
bags include the ones filled with breath, where ethylene levels are usually around 1 ppbv.
Higher ethylene concentrations in the bags decreased with less than 5% after 3 days.
Water vapour content decreases drastically by 80% within about 24 h (figure 3) and is
then only slowly reduced to about 12.5% of its initial value after 250 h. Heating the bag does
not regain this loss of water. This establishes the findings of Groves and co-workers (Groves
and Zellers 1996), who suggest water to permeate through the walls of the bag.
In bags filled with breath, the decrease of sampled gas concentrations in time was found
to be somewhat lower (table 1). From breath (tentatively identified) methanol, acetaldehyde,
8 M M L Steeghs et al
Table 1 . Relative average ion intensity values found at several masses from the breath of ten volunteers at different time points. The data at 2 h indicate average (±sd) count rates, where
the values at later time points are expressed in percentages of the values at 2 h. P-value indicates the significance of a change in ion intensity between 2 h and 240 h. The spread in the
2 h values indicates the inter-personal variation in the concentration of the compounds.
Mass (compound
a
) 2 h signal intensity P value of
(amu) (ncps ± sd) 24h(%± %) 72h(%± %) 144 h (% ± %) 240 h (% ± %) change
b
31 14 ± 10 2.6 × 10
2
± 1.3 × 10
2
2.2 × 10
2
± 1.8 × 10
2
2.3 × 10
2
± 1.6 × 10
2
1.8 × 10
2
± 1.4 × 10
2
0.144
33 (methanol) 2.7 × 10
3
± 1.2 × 10
3
1.1 × 10
2
± 0.2 × 10
2
0.9 × 10
2
± 0.4 × 10
2
0.7 × 10
2
± 0.1 × 10
2
0.4 × 10
2
± 0.1 × 10
2
0.001
b
41 87 ± 50 3.7 × 10
2
± 3 × 10
2
3.6 × 10
2
± 2.8 × 10
2
4 × 10
2
± 2.4 × 10
2
4.6 × 10
2
± 3.4 × 10
2
0.001
b
43 2.8 × 10
2
± 1.1 × 10
2
1.2 × 10
2
± 0.3 × 10
2
1.7 × 10
2
± 1 × 10
2
1.9 × 10
2
± 0.7 × 10
2
1.8 × 10
2
± 0.8 × 10
2
0.019
b
45 (acetaldehyde) 3.4 × 10
2
± 2 × 10
2
1.2 × 10
2
± 0.3 × 10
2
1.7 ×10
2
± 1.4 ×10
2
1.5 ×10
2
± 0.5 ×10
2
1 ×10
2
± 0.4 ×10
2
0.726
47 (ethanol) 1.6 × 10
2
± 0.6 × 10
2
1.2 × 10
2
± 0.4 × 10
2
0.9 ×10
2
± 0.4 ×10
2
1 ×10
2
± 0.5 ×10
2
0.8 ×10
2
± 0.6 ×10
2
0.072
b
59 (acetone) 3.7 × 10
3
± 1.5 × 10
3
1.3 × 10
2
± 0.2 × 10
2
1 ×10
2
± 0.4 ×10
2
1.4 ± 0.9 ×10
2
1 ×10
2
± 0.5 ×10
2
0.851
61 1.7 × 10
2
± 0.6 × 10
2
0.9 × 10
2
± 0.4 × 10
2
1.2 × 10
2
± 0.6 × 10
2
1.1 × 10
2
± 0.6 × 10
2
1.3 × 10
2
± 0.5 × 10
2
0.450
63 45 ± 23 4.2 × 10
2
± 3.7 × 10
2
3.8 × 10
2
± 3.1 × 10
2
3.6 × 10
2
± 2.4 × 10
2
2.6 × 10
2
± 1.4 × 10
2
0.0018
69 (isoprene) 4.6 × 10
2
± 1.1 × 10
2
1.2 × 10
2
± 0.3 × 10
2
1.2 × 10
2
± 0.4 × 10
2
1.3 × 10
2
± 0.4 × 10
2
0.9 × 10
2
± 0.3 × 10
2
0.380
71 60 ± 61 1 × 10
2
± 0.9 × 10
2
1.2 ± 0.7 × 10
2
1.8 × 10
2
± 1.6 × 10
2
2 × 10
2
± 16 × 10
2
0.925
88 (N,N-dimethylacetamide) 6.8 × 10
2
± 2 × 10
2
0.8 × 10
2
± 0.4 × 10
2
2.5 × 10
2
± 1.6 × 10
2
1.1 × 10
2
± 0.6 × 10
2
2.9 × 10
2
± 2 × 10
2
0.007
b
93 22 ± 14 0.6 × 10
2
± 0.8 × 10
2
0.6 × 10
2
± 0.9 × 10
2
1.2 × 10
2
± 1 × 10
2
1.5 × 10
2
± 1.3 × 10
2
0.410
95 2.5 × 10
2
± 0.8 × 10
2
1.3 × 10
2
± 0.5 × 10
2
2.9 × 10
2
± 1.6 × 10
2
2.1 × 10
2
± 0.7 × 10
2
3.3 × 10
2
± 1.7 × 10
2
0.000
b
a
Identification is tentative.
b
P value of student’s t-test for equality of means.
∗
P values smaller than 0.1 are considered significant.
The suitability of Tedlar bags for breath sampling in medical diagnostic research 9
Figure 4. Concentrations of six different compounds in the glass cuvette enclosing three Tedlar
bags filled with calibrated mixture. As the concentrations of compounds in the bags decrease (see
figure 3), a corresponding increase is found in the surrounding airspace in the cuvette. Temperature
was the same for all time points (22
◦
C).
acetone and isoprene can be compared with the certified mixture measurements. Of these
four compounds, only methanol is found to decrease significantly. Besides methanol, only
ethanol is found to also decrease significantly. On the other hand masses 41, 88, 95 increase
significantly. Deviations from the average values are increasing over time.
Heating of the bags resulted in an increase of several compounds, although the effect for
most compounds was in the range of 2–10% of the value for the unheated bag. This indicates
that only a small proportion of the volatile contents is adsorbed to the wall and that most of
it probably diffused through. Masses 88 and 95, which are due to the bag material itself,
displayed an increase of about 170% and 88% with respect to the values of the unheated bag.
A separate experiment was performed to test the diffusion through the bag wall.
Figure 4 shows the concentrations of six compounds in the glass cuvette enclosing three Tedlar
bags on four different moments. Methanol, acetaldehyde and acetone clearly increase over
time, whereas benzene and toluene show no significant increase. The gain in concentration
in the cuvette corresponds well with the decreases found from the bags for acetaldehyde and
acetone (within measurement uncertainty). For methanol, only 20% of the losses from the
bag content are found back in the cuvette. As can be seen from the trends in figure 4 the loss
rate decreases when the difference in concentration between the inside and outside of the bags
decreases, which can be expected from a diffusion phenomenon.
3.3. Relative humidity
The influence of relative humidity on the concentrations in a bag was tested by increasing
the relative humidity to at least 80% (100% for bags containing air or nitrogen, 80% for bags
containing calibrated mixture). There was no observable effect of water vapour on the stability
of the bag contents. No significant differences could be found between ‘dry’ (RH = ‘0’%)
and ‘wet’ air (RH > 80%) (results not shown).
10 M M L Steeghs et al
4. Conclusion
We have tested the suitability of black-layered Tedlar bags for breath sampling. The contents
of the bags were found to be polluted, by two compounds giving rise to characteristic ions at
masses 88 and 95 amu, identified by three independent groups as N,N-dimethylacetamide and
phenol, respectively (Amann, private communication; Di Francesco, private communication).
It should be kept in mind that these products are present in significant quantities when analysing
the volatile content of the bags. The ion intensities and the variation therein essentially make
them ‘blind spots’ in PTR-MS analysis of the bag contents, since the amount of pollution
varies greatly from one bag to another.
Compounds from the standard gas mixture are found to display considerable and
reproducible losses between 5 and 50% during filling. The losses possibly occur due to
sticking effects on the septum in the fitting. After this drastic decrease during filling, all
compounds except styrene show only a slow decrease in concentration, ranging from ∼10%
within 2 days for isoprene to ∼25% for styrene, which shows by far the fastest decrease of all
compounds tested (see figure 3). The compounds in the test mixture represent most chemical
groups of volatile organic compounds usually found in breath (Amann 2005). An important
other chemical group, sulphur compounds, has been investigated in more detail elsewhere
(Nielsen 2002). The pure compounds tested here all display a half-life of ∼6 days or longer.
Immediate losses during filling can also be expected for breath VOCs, although these
have not been tested here. Most of the losses during storage from compounds in breath
were limited to less than 10% within 52 h. These observed decreases from VOCs in breath
samples (compared are methanol, acetaldehyde, acetone and isoprene) are lower than the
decreases measured for the compounds in the standard gas mixtures. This might be due
to the relatively high concentrations (200 ppbv or higher) from the pure compounds in
comparison to most values commonly found in breath. Diffusion, which we showed to
be a loss mechanism, depends on the concentration gradient, decreasing the losses with
decreasing differences in concentration between the inside of the bag and the environment
around it.
Relative humidity showed not to have a significant effect on the lifetime of compounds
in the bags. Groves and co-workers (Groves and Zellers 1996) did find a significant influence
for methanol and acetone, but only at supersaturated humidity levels.
Even though two compounds are demonstrated to pollute the contents of the bags, these
will not interfere with the analysis of other compounds contained in them. After immediate
losses during filling, only slow and reproducible (comparable for all bags measured in two
series of measurements) decreases in concentrations are observed. The variations in these
losses are generally smaller than the inter-personal differences for most compounds present
in breath (for these variations, see, for instance, Turner et al (2006a, 2006b, 2006c, 2006d)).
To increase reproducibility, we advise a fixed point in time after sampling is chosen for the
analysis of the contents. We conclude that, as long as the bags are characterized for the
compounds measured in such a study and a fixed time point after collection is chosen, black-
layered Tedlar bags are suitable to be used for sampling breath in clinical studies, where it is
not possible to analyse the breath sample immediately.
Acknowledgments
The authors would like to thank Professor Norman Ratcliffe from the University of West
England, Professor Anton Amann, University of Innsbruck, Professor Fabio di Francesco,
The suitability of Tedlar bags for breath sampling in medical diagnostic research 11
University of Pisa who kindly and independently confirmed us from their GC-MS data, the
identification of both N,N-dimethylacetamide and phenol.
References
Amann A 2005 Breath analysis for medical diagnosis and therapeutic monitoring Spectrosc. Eur. 17 18
Berkelmans H W A, Moeskops B W M, Bominaar J, Scheepers P T J and Harren F J M 2003 Pharmacokinetics of
ethylene in man by on-line laser photoacoustic detection Toxicol. Appl. Pharmacol. 190 206–13
Boamfa E, Steeghs M, Cristescu S and Harren F 2004 Trace gas detection from fermentation processes in apples:
an intercomparison study between proton-transfer-reaction mass spectrometry and laser photoacoustics Int. J.
Mass Spectrom. 239 193–201
Boschetti A et al 1999 PTR-MS real time monitoring of the emission of volatile organic compounds during postharvest
aging of berryfruit Postharvest Biol. Technol. 17 143–51
Chase R E 2001 Properties and manufacture of Tedlar Polyvinyl fluoride film ERC Technical Report VERL, Ford
Research Laboratory, May 1st, 2001
Critchley A, Elliott T, Harrison G, Mayhew C, Thompson J and Worthington T 2004 The proton transfer reaction
mass spectrometer and its use in medical science: applications to drug assays and the monitoring of bacteria
Int. J. Mass Spectrom. 239 235–41
Dannecker J, Shaskan E and Phillips M 1981 A new highly sensitive assay for breath acetaldehyde detection of
endogenous levels in humans Anal. Biochem. 114 1–7
Davies S, Spanel P and Smith D 2001 A new ‘online’ method to measure increased exhaled isoprene in end-stage
renal failure Nephrol. Dial. Transpl. 16 836–9
de Gouw J, Warneke C, Karl T, Eerdekens G, van der Veen C and Fall R 2003 Sensitivity and specificity of atmospheric
trace gas detection by proton-transfer-reaction mass spectrometry Int. J. Mass Spectrom. 223–224 382
Deng 2004 Determination of acetone in human breath by gas chromatography-mass spectrometry and solid-phase
See endnote 2
microextraction with on-fiber derivatization J. Chromatogr. B 810 269–75
Co du Pont E I 1964 Process of preparing oriented, organic thermoplastic polymeric film Patent nr. 3.138.470, United
States Patent Office
Groves W and Zellers E 1996 Investigation of organic vapor losses to condensed water vapor in Tedlar bags used for
exhaled-breath sampling Am. Ind. Hyg. Assoc. J. 57 257–63
Harren F J M, Berkelmans R, Kuiper K, Hekkert S T L, Scheepers P, Dekhuijzen R, Hollander P and Parker D H
See endnote 3
1999 On-line laser photoacoustic detection of ethene in exhaled air as biomarker of ultraviolet radiation damage
of the human skin Appl. Phys. Lett. 74
Harren F J M and Reuss J 1997 Photoacoustic Spectroscopy vol 19, ed G L Trigg (New York: Wiley) (Encycl. Appl.
Phys.) pp 413–35
Hewitt C, Hayward S and Tani A 2003 The application of proton transfer reaction-mass spectrometry (PTR-MS) to
the monitoring and analysis of volatile organic compounds in the atmosphere J. Environ. Monit. 5 1–7
Holzinger R, Sandoval-Soto L, Rottenberger S, Crutzen P and Kesselmeier J 2000 Emissions of volatile organic
compounds from Quercus ilex L. measured by proton transfer reaction mass spectrometry under different
environmental conditions J. Geophys. Res.-Atmos. 105 20573–9
http://webbook.nist.gov/chemistry/
Hyspler 2000 Determination of isoprene in human expired breath using solid-phase microextraction and gas
chromatography-mass spectrometry J. Chromatogr. B 739 183–90
Jobson B 2005 On-line analysis of organic compounds in diesel exhaust using a proton transfer reaction mass
spectrometer (PTR-MS) Int. J. Mass Spectrom. 245 77–88
Jordan A, Hansel A, Holzinger R and Lindinger W 1995 Acetonitrile and benzene in the breath of smokers and
nonsmokers investigated by proton-transfer reaction mass spectrometry (PTR-MS) Int. J. Mass Spectrom. Ion
Process. 148 L1–3
Karl T, Prazeller P, Mayr D, Jordan A, Rieder J, Fall R and Lindinger W 2001 Human breath isoprene and its relation
to blood cholesterol levels: new measurements and modeling J. Appl. Phys. 91 762–70
Kneepkens C M F, Lepage G and Roy C C 1994 The potential of the hydrocarbon breath test as a measure of lipid
peroxidation Free Radic. Biol. Med. 17 127–60
Lau Y K 1989 Measurement of sulphur gases in ambient air Environ. Monit. Assess. 13 69–74
Lechner M, Colvin H P, Ginzel C, Lirk P, Rieder J and Tilg H 2005 Headspace screening of fluid obtained from the
gut during colonoscopy and breath analysis by proton transfer reaction-mass spectrometry: a novel approach in
the diagnosis of gastro-intestinal diseases Int. J. Mass Spectrom. 243 151–4
Lindinger C 2005 Unambiguous identification of volatile organic compounds by proton-transfer reaction mass
spectrometry coupled with GC/MS Anal. Chem. 77 4117–24
12 M M L Steeghs et al
Lindinger W, Hansel A and Jordan A 1998a Proton-transfer-reaction mass spectrometry (PTR-MS): on-line
monitoring of volatile organic compounds at pptv levels Chem. Soc. Rev. 27 347–54
Lindinger W, Hansel A and Jordan A 1998b On-line monitoring of volatile organic compounds at pptv levels
by means of proton-transfer-reaction mass spectrometry (PTR-MS)—medical applications, food control and
environmental research Int. J. Mass Spectrom. 173 191–241
Myung 1997 Gas chromatographic-mass spectrometric analysis of mercaptan odorants in liquefied petroleum gas and
liquefied natural gas J. Chromatogr. A 791 367–70
Nielsen 2002 Quantification of volatile sulfur compounds in complex gaseous matrices by solid-phase microextraction
J. Chromatogr. A 963 57–64
Phillips M 1992 Breath tests in medicine Sci. Am. 267 74–9
Phillips M et al 1999 Breath markers of heart transplant rejection Circulation 100 525–5
Phillips M and Greenberg J 1991 Method for the collection and analysis of volatile compounds in the breath
J. Chromatogr. Biomed. Appl. 564 242–9
Prazeller P, Karl T, Jordan A, Holzinger R, Hansel A and Lindinger W 1998 Quantification of passive smoking using
See endnote 4
proton-transfer reaction mass spectrometry Int. J. Mass Spectrom. 178
Steeghs M, Bais H, de Gouw J, Goldan P, Kuster W, Northway M, Fall R and Vivanco J 2004 Proton-transfer-reaction
mass spectrometry as a new tool for real time analysis of root-secreted volatile organic compounds in arabidopsis
Plant Phys. 135 47–58
Steeghs M M L, Moeskops B W M, van Swam K, Cristescu S M, Scheepers P T J and Harren F J M 2006 On-line
monitoring of UV-induced lipid peroxidation products from human skin in vivo using Proton-Transfer Reaction
Mass Spectrometry Int. J. Mass Spectrom. 253 58–64
Sulyok 2001 Investigation of the storage stability of selected volatile sulfur compounds in different sampling containers
J. Chromatogr. A 917 367–74
Turner C, Spanel P and Smith D 2006a A longitudinal study of breath isoprene in healthy volunteers using selected
ion flow tube mass spectrometry (SIFT-MS) Phys. Meas. 27 13–22
Turner C, Spanel P and Smith D 2006b A longitudinal study of ammonia, acetone and propanol in the exhaled breath
of 30 subjects using selected ion flow tube mass spectrometry, SIFT-MS Phys. Meas. 27 321–37
Turner C, Spanel P and Smith D 2006c A longitudinal study of methanol in the exhaled breath of 30 healthy volunteers
using selected ion flow tube mass spectrometry, SIFT-MS Phys. Meas. 27 637–48
Turner C, Spanel P and Smith D 2006d A longitudinal study of ethanol and acetaldehyde in the exhaled breath of
30 subjects using selected ion flow tube mass spectrometry, SIFT-MS Rapid Commun. Mass Spectrom. 20 61–8
Wang T, Spanel P and Smith D 2004 A selected ion flow tube study of the reactions of H
3
O
+
,NO
+
and O-2(+center
dot) with some phenols, phenyl alcohols and cyclic carbonyl compounds in support of SIFT-MS and PTR-MS
Int. J. Mass Spectrom. 239 139–46
Warneke C, Kuczynski J, Hansel A, Jordan A, Vogel W and Lindinger W 1996 Proton transfer reaction mass
spectrometry (PTR-MS): propanol in human breath Int. J. Mass Spectrom. Ion Process. 154 61–70
Werle P 2003 Diode-Laser Sensors for in-situ Gas Analysis ed P Hering, J P Lay and S Stry (Berlin: Springer)
Woltering E, Harren F and Bicanic D D 1989 Laser photo-acoustics, a novel method for ethylene determination in
plant physiology studies Acta Horticultura 201–8
Endnotes
(1) Author: Please include references ‘Amann, private communication’ and ‘Di Francesco,
private communication’ in the references list.
(2) Author: Please provide complete names of authors in references Deng (2004), Hyspler
(2000), Myung (1997) and Sulyok (2001).
(3) Author: Please provide page number in reference Harren et al (1999).
(4) Author: Please provide page number in reference Prazeller et al (1998).
(5) Author: Please check whether the sense of the sentence ‘The data at 2 h indicate
average .....’ retains your intended sense in caption of Table 1.
(6) Author: Please check the value ‘1.8 × 10
2
± 0.8 × 10
2
in the table as set.
Reference linking to the original articles
References with a volume and page number in blue have a clickable link to the original
article created from data deposited by its publisher at CrossRef. Any anomalously unlinked
references should be checked for accuracy. Pale purple is used for links to e-prints at arXiv.