since their very similar chemical and physical
properties cause them to coelute with the com-
pound to be quantified. The isotope dilution
assay is the most accurate method currently
available for the quantification of labile
odorants and those in low concentrations. After
the first application of IDA for flavor research
(Schieberle and Grosch, 1987) the method was
systematically developed for >60 potent
odorants (Schieberle, 1995a). The accuracy of
IDAs and quantification with unlabeled inter-
nal standards was recently discussed (Preinin-
ger, 1998). Table G1.3.1 lists important aroma
compounds as well as their labeled analogs and
gives an extensive overview of the literature
using this technique in flavor research.
One of the main advantages of IDA is that
quantitative isolation of odorants from the sam-
ple is not required, provided that the internal
standard is homogeneously distributed
throughout the sample. Solid samples should
be finely ground under liquid nitrogen in order
to facilitate even penetration of the labeled
standard throughout the sample. The time nec-
essary for the standard to be evenly mixed with
the analyte can be checked (Milo and Blank,
1998). Due to the high chemical and physical
similarities of isotopomers, losses during the
work up procedure are ideally compensated.
Another strength of the method is the high
selectivity and sensitivity of GC-MS, particu-
larly in selected ion monitoring mode. Com-
pounds that are poorly resolved by GC may not
interfere in GC-MS due to different fragmen-
tation patterns and, therefore, less sample
clean-up is needed (and under CI less separa-
tion is needed).
The main hurdle to the application of IDA
for quantitative measurements of aroma com-
pounds continues to be the limited commercial
availability of the labeled internal standards.
Critical Parameters and
Troubleshooting
Identification and quantification of
volatiles (GC-FID)
The most critical parameter for the Basic
Protocol is the yield of analyte at the detector.
Simulations and models are the best methods
to identify the extent of the problem and to
determine correction factors.
Quantification of aroma compounds by
isotope dilution assays
The labeled standards used need to be iso-
topically stable. Isotopic stability may become
an issue if the compound is labeled with deu-
terium in an enolizable position as is the case
for the
-position of carbonyl functions. The
deuterium may then be exchanged with protons
from the sample during the workup and there-
fore falsify the result. In order to rule out such
D/H exchanges, the standard may be tested
under the conditions of the isolation of volatiles
from the sample or, better, be replaced by a
standard labeled with
13
C.
Furthermore the labeling of the compound
should increase its molecular weight by at least
2 units, preferably 3 units, in order to minimize
interferences with the natural isotope distribu-
tion of the analyte.
For noncommercially available standards,
the chemical purity plays an important role. The
price for a custom synthesis may be lower if
only 70% chemical purity is negotiated. As long
as the contaminants do not interfere with other
compounds to be quantified in the sample and
will not convert into the final labeled product
during the workup procedure, this can be toler-
ated.
Although IDA is particularly useful for the
quantitation of labile compounds, one needs to
pay attention to possible degradation products.
The possible conversion of (Z)-3-alkenals to
the (E)-2 alkenals will bias results if these
compounds are to be quantified simultane-
ously. Another example are thiols and their
dimers that should be quantified in separate
samples using the respective labeled standards.
The EI mode can be used in IDA, provided
that characteristic ions of high intensity are
available for quantification. In some cases,
however, it is necessary to improve the effi-
ciency of IDA by changing and optimizing the
ionization technique as recently demonstrated
for epoxy-alkenals (Blank et al., 1999).
Anticipated Results
Due to the bias that can occur in the Basic
Protocol, results can be as low as 1% of the
correct value for the concentration of a volatile
in a sample. Most results are <80% of the
correct value. This bias is usually caused by
differences in the recovery between the internal
standard and the analytes. Choosing an internal
standard that is similar to the analyte can reduce
this bias but it cannot be completely eliminated
unless a separate standard is used for each
analyte (see Alternate Protocol). Nevertheless,
the Basic Protocol is often sufficiently accurate
because the olfactory system, unlike the taste
system, is compressive and insensitive to small
Current Protocols in Food Analytical Chemistry
G1.3.9
Smell Chemicals