Identification of surrogate compounds for the emission of PCDD/F (I-TEQ value) and evaluation of their on-line realtime detectability in flue gases of waste incineration plants by REMPI-TOFMS mass spectrometry

Article abstract of DOI:10.1016/S0045-6535(00)00223-X

Correlations between products of incomplete combustion (PIC), e.g., chloroaromatic compounds, can be used to characterise the emissions from combustion processes, like municipal or hazardous waste incineration. A possible application of such relationships may be the on-line real-time monitoring of a characteristic surrogate, e.g., with Resonance-Enhanced Multiphoton Ionization-Time-of-Flight Mass Spectrometry (REMPI-TOFMS). In this paper, we report the relationships of homologues and individual congeners of chlorinated benzenes (PCBz), dibenzo-p-dioxins (PCDD), dibenzofurans (PCDF) and phenols (PCPh) to the International Toxicity Equivalent (I-TEQ) of the PCDD/F (I-TEQ value) in the flue gas and stack gas of a 22 MW hazardous waste incinerator (HWI). As the REMPI detection sensitivity is decreasing with the increase of the degree of chlorination, this study focuses on the lower chlorinated species of the compounds mentioned above. Lower chlorinated species, e.g., chlorobenzene (MCBz), 1,4-dichlorobenzene, 2,4,6-trichlorodibenzofuran or 2,4-dichlorophenol, were identified as I-TEQ surrogates in the flue gas. In contrast to the higher chlorinated phenols, the lower chlorinated phenols (degree of chlorination <4) were not reliable as surrogates in the stack gas. The identified surrogates are evaluated in terms of their detectability by REMPI-TOFMS laser mass spectrometry. The outcome is that MCBz is the best suited surrogate for (indirect) on-line measuring of the I-TEQ value in the flue gas by REMPI-TOFMS. The correlation coefficient r of the MCBz concentration to the I-TEQ in the flue gas was 0.85.

Full text of DOI:10.1016/S0045-6535(00)00223-X

Chemosphere 42 (2001) 507±518
Identi®cation of surrogate compounds for the emission of
PCDD/F (I-TEQ value) and evaluation of their on-line real-
time detectability in ¯ue gases of waste incineration plants by
REMPI±TOFMS mass spectrometry
a
b,
b
a,b
M. Blumenstock , R. Zimmermann *, K.-W. Schramm , A. Kettrup
a

Lehrstuhl fur Okologische Chemie und Umweltanalytik, Technische Universitat Munchen, D-85354 Freising, Germany
b

GSF-Forschungszentrum fur Umwelt und Gesundheit GmbH, Institut fur Okologische Chemie, D-85764 Oberschleiûheim, Germany
Abstract
Correlations between products of incomplete combustion (PIC), e.g., chloroaromatic compounds, can be used to
characterise the emissions from combustion processes, like municipal or hazardous waste incineration. A possible
application of such relationships may be the on-line real-time monitoring of a characteristic surrogate, e.g., with
Resonance-Enhanced Multiphoton Ionization)Time-of-Flight Mass Spectrometry (REMPI±TOFMS). In this paper,
we report the relationships of homologues and individual congeners of chlorinated benzenes (PCBz), dibenzo-p-dioxins
(PCDD), dibenzofurans (PCDF) and phenols (PCPh) to the International Toxicity Equivalent (I-TEQ) of the PCDD/F
(I-TEQ value) in the ¯ue gas and stack gas of a 22 MW hazardous waste incinerator (HWI). As the REMPI detection
sensitivity is decreasing with the increase of the degree of chlorination, this study focuses on the lower chlorinated
species of the compounds mentioned above. Lower chlorinated species, e.g., chlorobenzene (MCBz), 1,4-dichloro-
benzene, 2,4,6-trichlorodibenzofuran or 2,4-dichlorophenol, were identi®ed as I-TEQ surrogates in the ¯ue gas. In
contrast to the higher chlorinated phenols, the lower chlorinated phenols (degree of chlorination <4) were not reliable
as surrogates in the stack gas. The identi®ed surrogates are evaluated in terms of their detectability by REMPI±TOFMS
laser mass spectrometry. The outcome is that MCBz is the best suited surrogate for (indirect) on-line measuring of the I-
TEQ value in the ¯ue gas by REMPI±TOFMS. The correlation coecient r of the MCBz concentration to the I-TEQ in
the ¯ue gas was 0.85. Ó 2001 Elsevier Science Ltd. All rights reserved.
Keywords: Surrogates; Polychlorinated dibenzo-p-dioxins/furans; Chlorinated benzenes; Chlorinated phenols; Laser mass spectrom-
etry; On-line monitoring; PCDD/F, I-TEQ; REMPI±TOFMS
1. Introduction
gates for the emissions of dibenzo-p-dioxins (PCDD)


and dibenzofurans (PCDF) (Oberg and Bergstrom,
1985; Oehme et al., 1987). Recently the correlation
strength of compounds like hexa- and pentachlorinated
benzene or pentachlorinated phenol was reinvestigated
(Kaune et al., 1994, 1998). Moreover, lower chlorinated
congeners of PCDD/F or chlorinated benzenes (PCBz)
were used to predict the emitted toxicity of PCDD/F
(I-TEQ) with simple or multivariate regression analysis
Some chlorinated aromatic compounds, especially
chlorinated phenols and benzenes are known as surro-
* Corresponding author: Tel.: +49-089-3187-4544; fax: +49-
089-3187-3371.
E-mail address: ralf.zimmermann@gsf.de (R. Zimmer-

(Blumenstock et al., 1999a,b; Gullett and Wikstrom,
1999).
mann).
0045-6535/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 5 - 6 5 3 5 ( 0 0 ) 0 0 2 2 3 - X

508
M. Blumenstock et al. / Chemosphere 42 (2001) 507±518
Fig. 1. Schematic view of the HWI plant (thermal power: 22 MW).
However, the PCBz, phenols (PCPh) and low chlo-
particulate matter from the ¯ue gas it is passed
through a condensate ¯ask and a cartridge ®lled
with 30 g quartz glass wool. (Blumenstock et al.,
1999b). To avoid memory e€ects in the sampling unit
all glass parts were changed after each sampling pe-
riod. In order to avoid sampling losses the glass inlets
of the cooled ¯ue gas probe were ¯ushed with toluene
used for the extraction of the corresponding XAD-2
resin.
rinated PCDD/F (one- to threefold chlorinated) are also
present in the ¯ue gas during waste incineration pro-
cesses, partly in even higher concentrations than the
respective higher chlorinated congeners. Thus they may
be valuable target compounds for on-line measurements
with Resonance-Enhanced Multiphoton Ioniza-
tion)Time-of-Flight Mass Spectrometry (REMPI±
TOFMS) (Blumenstock et al., 1998a, 1999b; Dellinger
and Hall, 1986; Zimmermann et al., 1996a).
PCBzs and PCPhs were sampled from a small bybass
at 70°C before the gas passes through the condensate
¯ask of the sampling unit. For trapping of both com-
pound classes, Carbotrap was used as adsorbent. Sam-
pling time for the volatile compounds (VOCs) were 15±
45 min and for the SOCs between 90 and 360 min. The
sampling times were much more shorter than the rec-
ommended 6 h sampling period in the VDI guideline,
because we want to get time resolved information about
the formation of chloroaromatics in the ¯ue gas. The
clean-up procedure of SOCs and the VOCs has been
described elsewhere (Schramm et al., 1995; Blumenstock
et al., 1999a,b).
First ®eld applications of this two-dimensional laser
mass spectrometry technique for real-time on-line moni-
toring of trace compounds in di€erent industrial com-
bustion and pyrolysis processes were performed in the last
three years (Zimmermann et al., 1997, 1998, 1999b).
In this comprehensive study, the International Tox-
icity Equivalent (I-TEQ) surrogates within di€erent
substance classes (chlorinated aromatics) are identi®ed
and compared. The di€erent chlorinated aromatics in-
clude PCBz, PCDD, PCDF and PCPh. The feasibility of
on-line REMPI±TOFMS detection of the identi®ed
I-TEQ surrogates is evaluated.
For determination of the PCBz a thermodesorption
unit (Perkin Elmer ATD 400) coupled to a HRGC/
LRMS was used and the samples were injected without
any previous cleanup. The PCPhs containing tubes were
extracted and the solvent samples derivitised with acetic
anhydride for analysis with HRGC/MS.
2. Methods and materials
In Fig. 1, a schematic view of the hazardous waste
incineration (HWI) plant investigated is given, including
the conditions at the sampling points (SP). The SP1 is
installed in the ¯ue gas and SP2 in the stack gas. Sam-
pling was performed simultaneously at the two SPs with
a conventional isokinetic sampling system. Furthermore
at the SP1 a REMPI±TOFMS device was installed.
The sampling was done isokinetically according to
the German guideline VDI DIN 3499 (VDI 3499, 1990)
with a completely automated operating system described
elsewhere (Schramm et al.,1998).
All compounds classes were detected with HRGC/
MS or HRGC/HRMS in the selected ion mode (SIM)
using isotope dilution with at least one 13C-labelled
isomer for each degree of chlorination within each
substance class for quanti®cation. Details on the
analytical procedure have been published elsewhere
(Blumenstock et al. 1999a).
All concentrations are normalised to dry air, 11% O2,
1013 mbar and 273 K. The concentrations are given on a
MASS/VOLUME base as used for determination of the
I-TEQ for legal purposes in Germany. The data was
For trapping semivolatile compounds (SOCs) car-
tridges containing XAD-2 were used. To separate

M. Blumenstock et al. / Chemosphere 42 (2001) 507±518
509
transformed to a log scale to obtain normal distribution
for statistical regression analysis.
± mostly trichlorinated ± dibenzo-p-dioxins (PCDD) are
surrogates at this HWI plant.
On the other hand the lower chlorinated PCDF (one-
and twofold chlorinated) did not show high correlations
with the I-TEQ. Accordingly this result is also true for
the homologues. An exception within the lower chlori-
nated species is the 1-MCDF, exhibiting a r of 0.69 for I-
TEQ prediction. The 2,4,6-TriCDF is highly correlated
with the I-TEQ value and with the I-TEQ value carried
from the PCDF, denoted as I-TEQPCDF. This isomer
also achieved high coecients of determination in mul-
tiple regression analysis in the work of Gullett and Wi-
3. Results and discussion
3.1. Lower chlorinated PCDD/F as surrogates for the I-
TEQ emission
In Tables 1 and 2, the correlation coecients for
lower chlorinated PCDD/F congeners in the ¯ue gas of
the HWI are shown. Signi®cant values (p-level < 5%) are
printed in bold italics. The I-TEQ values in the ¯ue gas
range from 0.56 to 11.21 ng/m3.

kstrom (1999). The low or sometime even negative
correlation coecients of low chlorinated PCDF with
the I-TEQ value or the PCDD/F homologues shows that
they behave di€erent with respect to high chlorinated
PCDF and PCDD. This e€ect, partly can be explained
by a dramatic increase of one-to trichlorinated PCDF
after disturbed combustion conditions (e.g., CO-peaks)
at the HWI (Blumenstock et al., 1999b; Zimmermann
et al., 1998). The same behaviour was observed at a waste
incineration pilot plant (Blumenstock et al. 1998a,b).
For the PCDD isomers, we observed signi®cant
correlation coecients with the I-TEQ value for one
dichlorinated and six trichlorinated isomers. The corre-
lation coecients were relatively high for the dichlori-
nated isomers, but not signi®cant. The same behaviour
is observed for the homologue sums. Furthermore the
di- and trichlorinated isomers were highly and signi®-
cantly correlated with the sum of the PCDD/F. So some
Table 1
Correlation coecients for lower chlorinated PCDD isomers and homologues with sum parameters in the ¯ue gas (signi®cant values
(p-level < 5%) in bold italics).a
N ˆ 24 Flue gas samples at 350°C
PCDD/F
PCDD
PCDF
I-TEQ
I-TEQPCDD
1-MCDDb
0.56
0.59
0.62
0.65
0.61
0.67
0.65
0.68
0.52
0.68
0.66
0.68
0.75
0.75
0.73
0.75
0.73
0.74
0.73
0.68
0.72
0.66
0.57
0.73
0.74
)0.15
)0.06
0.78
0.16
0.36
0.31
0.56
0.29
0.56
0.34
0.28
0.56
0.41
0.61
0.60
0.62
0.69
0.58
0.56
0.71
0.72
0.62
)0.13
0.48
0.52
0.45
0.62
0.58
0.57
0.49
0.71
0.66
0.61
0.64
0.73
0.69
0.77
0.72
0.75
0.69
0.71
0.67
0.55
0.65
0.57
0.74
0.48
0.33
0.67
0.71
0.23
0.28
0.62
0.40
0.62
0.59
0.65
0.58
0.57
0.62
0.60
0.71
0.66
0.71
0.73
0.74
0.75
0.72
0.72
0.75
0.76
0.72
0.25
0.68
0.70
)0.22
)0.16
0.59
0.01
0.39
0.37
0.44
0.36
0.44
0.43
0.46
0.61
0.59
0.58
0.62
0.59
0.61
0.59
0.60
0.63
0.65
0.71
)0.17
0.43
0.50
2-MCDDc
1,3-DiCDD
2,7/2,3/2,8-DiCDDd
1,4/1,7-DiCDD
1,8-DiCDD
1,6-DiCDD
1,2-DiCDD
1,9-DiCDD
1,3,7-TriCDD
1,3,8-TriCDD
1,3,6-TriCDD
1,2,4-TriCDD
1,3,9/2,3,7-TriCDD
1,4,7-TriCDD
1,2,3-TriCDD
1,7,8-TriCDD
1,2,7-TriCDD
1,2,8-TriCDD
1,4,6-TriCDD
1,2,6-TriCDD
1,2,9-TriCDD
MCDD
DiCDD
TriCDD
a I-TEQ
denotes the contribution of the PCDD to the total I-TEQ.
PCDD
b For REMPI spectroscopic data on 1-MCDD, see Zimmermann (1998).
c For REMPI spectroscopic data on 2-MCDD, see Zimmermann et al. (1996a).
d For REMPI spectroscopic data on 2,8±2,7 and 2,3-DiCDD, see Weickhardt et al. (1993) and Zimmermann et al. (1994).

510
M. Blumenstock et al. / Chemosphere 42 (2001) 507±518
Table 2
Correlation coecients for lower chlorinated PCDF isomers and homologues with sum parametersa
N ˆ 24 Flue gas samples at 350°C
1-MCDF
PCDD/F
PCDF
PCDD
I-TEQ
I-TEQPCDF
)0.06
)0.14
)0.17
)0.09
)0.50
)0.53
)0.37
)0.64
)0.29
)0.61
)0.35
)0.50
)0.44
)0.55
)0.38
)0.47
)0.23
)0.40
)0.65
)0.64
)0.65
)0.62
)0.33
)0.62
)0.48
)0.57
)0.48
)0.48
)0.59
)0.53
)0.02
)0.38
)0.35
)0.51
0.03
0.22
0.25
)0.09
)0.19
)0.15
)0.03
)0.62
)0.49
)0.28
)0.69
)0.17
)0.45
)0.18
)0.53
)0.39
)0.41
)0.22
)0.56
)0.48
)0.44
)0.51
)0.77
)0.53
)0.78
)0.21
)0.62
)0.27
)0.43
)0.66
)0.32
)0.57
)0.68
0.18
0.69
0.34
0.51
0.07
3-MCDF
2-MCDF
0.21
0.27
0.37
0.52
0.10
0.28
4-MCDF
1,3-DiCDF
)0.21
)0.26
)0.25
)0.39
0.09
0.08
)0.04
0.07
)0.11
)0.23
)0.01
)0.20
0.01
1,7-DiCDF
1,4-DiCDF
1,8-DiCDF
1,6-DiCDF
)0.04
0.27
0.08
1,2)DiCDF
2,4-DiCDF
3,7-DiCDF
)0.35
0.03
)0.08
0.03
0.28
0.10
)0.19
)0.57
)0.35
)0.07
)0.19
0.06
)0.21
0.12
2,7-DiCDF
2,3-DiCDF
2,8/3,6-DiCDFb
2,6-DiCDF
0.10
)0.09
0.19
0.04
)0.23
)0.03
)0.15
)0.11
)0.06
)0.30
)0.33
)0.28
)0.28
)0.1
)0.22
)0.03
)0.24
)0.26
)0.08
)0.26
)0.14
0.27
1,9/3,4-DiCDF
4,6-DiCDF
0.11
0.19
)0.03
)0.35
)0.37
)0.45
)0.41
0.01
1,3,7-TriCDF
1,3,8-TriCDF
1,3,6-TriCDF
1,3,4/1,4,8-TriCDF
1,2,4/1,6,7-TriCDF
1,7,8-TriCDF
1,2,3-TriCDF
1,2,7-TriCDF
1,2,9-TriCDF
1,3,9-TriCDF
2,4,7-TriCDF
1,2,6/1,2,8-TriCDF
2,4,8-TriCDF
1,4,9-TriCDF
2,3,7-TriCDF
1,6,8/2,3,4/2,3,8-TriCDF
2,3,6/3,4,7-TriCDF
2,6,7-TriCDF
1,4,6/1,4,7-TriCDF
3,4,6-TriCDF
2,4,6-TriCDF
MCDF
)0.10
)0.15
)0.14
)0.14
0.15
)0.4
)0.08
0.15
)0.22
)0.39
)0.33
)0.13
)0.39
)0.34
0.32
)0.12
)0.15
0.15
)0.13
)0.01
0.49
)0.2
)0.41
)0.14
)0.67
)0.11
)0.26
)0.79
)0.38
0.42
)0.11
0.29
0.02
)0.25
0.10
)0.07
)0.25
)0.12
)0.1
)0.64
)0.28
0.52
)0.16
0.44
0.07
0.34
0.26
)0.37
)0.8
)0.43
0.23
)0.24
)0.01
0.72
)0.33
)0.12
0.51
)0.09
)0.51
)0.47
0.26
)0.20
)0.22
0.09
)0.63
)0.34
0.54
0.09
0.30
)0.12
)0.08
DiCDF
TriCDF
0.09
a I-TEQ
denotes the contribution of the PCDF to the total I-TEQ). Signi®cant values (p-level < 5%) in bold italics.
PCDF
b REMPI spectrum of 2,8-DiCDF available in Zimmermann et al. (1996a).
In summary, the results from the regression analysis
of the lower chlorinated PCDD/F, show that PCDF are
not suitable surrogates for the I-TEQ in the ¯ue gas of
this plant (exception: 2,4,6-TriDF), but a few trichlori-
nated PCDD isomers exhibit high correlation coe-
cients with the I-TEQ or with the I-TEQPCDD:
3.2. Chlorinated phenols as surrogates for the I-TEQ
emission
In Table 3 the correlation coecients for the PCPh
with the I-TEQ and other PCDD/F sum parameters are
shown. The coecients of correlation for the PCPh
isomers to the I-TEQ in Table 3 show that only for the
monochlorinated PCPh no correlation was found. For
all the other homologues at least one isomer is corre-
lated signi®cantly with the I-TEQ. Furthermore in the
case of the PCPhs much better correlations are found for
It should be mentioned that measurements in the
stack gas showed similar correlation coecients for the
lower chlorinated PCDD/F with the I-TEQ (n ˆ 9). For
that reason it is possible to use the same surrogates in
the ¯ue and stack of this HWI plant.

M. Blumenstock et al. / Chemosphere 42 (2001) 507±518
511
Table 3
Correlation coecients for chlorinated PCPh isomers with sum parameters and the I-TEQa
N ˆ 12 Flue gas samples at 350°C
PCDD/F
PCDD
PCDF
I-TEQ
2-MCPhb
0.11
)0.05
0.02
0.48
0.32
0.72
0.48
)0.14
0.46
0.71
)0.02
0.65
0.68
0.56
0.81
0.34
0.41
0.13
0.84
0.62
0.61
0.66
0.45
0.68
0.51
0.18
0.73
0.79
0.78
0.84
)0.08
)0.23
)0.01
0.22
0.19
)0.31
)0.17
0.8
4-MCPhb
2,6-DiCPh
2,4-DiCPh
3,4-DiCPhb
2,4,6-TriCPh
2,3,6-TriCPh
2,3,5-TriCPh
2,4,5-TriCPh
2,3,4-TriCPh
3,4,5-TriCPh
2,3,5,6-TeCPh
2,3,4,6-TeCPh
2,3,4,5-TeCPh
PeCPh
0.15
0.67
0.6
0.82
0.54
0.23
0.48
0.64
)0.25
0.69
0.75
0.63
0.79
0.49
)0.06
0.47
0.69
)0.14
0.47
0.8
0.49
0.68
a Signi®cant values (p-level < 5%) in bold italics.
b REMPI spectrum available in Trembeull et al. (1985).
(a)
(b)
Fig. 2. Isomer patterns of chlorinated phenols in the ¯ue (a) and stack (b) gas of the HWI plant.

512
M. Blumenstock et al. / Chemosphere 42 (2001) 507±518
PCDD than for PCDF. This seems to indicate di€erent
formation pathways for PCDD and PCDF in the com-
bustion process (Blumenstock, 1999). It was found that
PCDD and PCPh are formed in the cooling zone of the
HWI plant between 400°C and 300°C whereas 40±60%
of the ®nal amount of the PCDF and PCBz is already
present at the post combustion chamber outlet in the
temperature range between 700°C and 900°C (Blumen-
similar formation temperature window in combustion
processes for PCDD and PCPh may be an explanation
for their good correlation.
In Fig. 2, the isomer patterns for chlorinated phenols
in the ¯ue and stack gas of the HWI are shown. Fig. 2
shows that the patterns of the lower chlorinated PCPhs
are more stable in the ¯ue gas than in the stack gas. The
stack gas patterns show missing isomers, especially for
the lower chlorinated species. This probably is due to the
in¯uence wet scrubber based ¯ue gas cleaning system
(see Fig. 1) which discriminates the lower chlorinated
phenols to a higher extend than the more unpolar higher

stock, 1999). Wikstrom et al. (1998) reported that the
formations of PCDF and PCBz in a laboratory scale
incinerator in the high temperature region is signi®cantly
higher than for the PCDD and PCPh. The observed
Fig. 3. Regression calculations and respective statistical results for selected chlorinated phenols in the ¯ue and stack gas of the HWI
plant: (a) denotes the slope and (b) the intercept of the regression line.

M. Blumenstock et al. / Chemosphere 42 (2001) 507±518
513
chlorinated isomers. As a consequence, we found only a
few isomers with a high I-TEQ correlation in the stack
gas (see Fig. 3(b)).
the 2,4,6-TriCPh showed only a slight increase. But in
general, the last two compounds mentioned could be
assumed as more or less ``reliable I-TEQ'' surrogates. A
similar behaviour for PCPh was reported from Kaune et
al. (1998), who found for di€erent HWI plants increas-
ing r values from the heat recovery zone (boiler section)
to the stack. This observation may be explained by
formation processes of PCDD/F (e.g., condensation re-
actions of precursors or chlorination/dechlorination
mechanisms of already formed PCDD/F) which take
place in the hot parts (>200°C) of the electrostatic pre-
cipitators of the investigated HWI plants. Summarising
the observed behaviour of the PCPh in the HWI plant, it
is obvious that only higher chlorinated phenols (tri-
chlorophenols or higher) are useful as surrogates for the
I-TEQ emission in the ¯ue and stack gas of the HWI.
Further the ¯ue gas cleaning system of the plant in¯u-
ences the PCPh isomer pattern in the stack gas dra-
matically (see Fig. 2).
Please note that the monochlorophenol concentra-
tions in Fig. 2 represent only a lower limit, due to ana-
lytical problems during the measurement with HRGC/
MS. Although the ratios for the M+ and M + 2 traces
for these compounds on the HRGC/MS were correct,
the peak shape was too broad (analytical peak broad-
ening on the GC column). The peak broadening prob-
ably was caused by an overloading of the column
analyte. We require more analytical data on mono-
chlorinated phenols in order to give a ®nal result on
prediction accuracy. Therefore the r values for mono-
chlorophenol in Table 3 have to be interpreted very
carefully.
Fig. 3 gives the regression lines for correlations of
selected chlorinated phenols with the I-TEQ values. All
depicted correlations and regressions were signi®cant on
an p-level < 5%. We found a correlation coecient (r)
for 2,4-dichlorophenol of 0.80 in the ¯ue gas. For the
higher chlorinated phenols with more than two chlorine
atoms in the molecule, r is just in the same range, in this
case for the 2,4,6-TriCPh and 2,3,4,6-TeCPh, respec-
tively. The regression line for PeCPh is not shown in the
®gure for the ¯ue gas, but PeCPh also exhibits an r of
0.84, which is in good agreement with the results from
Kaune et al. (1994,1998).
3.3. Chlorinated benzenes as surrogates for the I-TEQ
emission
The correlation coecients for chlorinated benzenes
are listed in Table 4. Only four isomers are not signi®-
cantly correlated with the I-TEQ, one di-, two tri- and
®nally one tetrachlorinated isomer. In comparison with
the PCPh the PCBz compound class exhibits the closer
relation to the I-TEQ. This result is probably due to
the similar temperature dependence of PCBz and PCDF
In moving from the ¯ue gas of the HWI plant to the
stack gas, the observed correlation for 2,4-dichlorinated
phenol disappears. Only the most dominant congeners
of the stack gas pattern given in Fig. 2 ®t the signi®cance
criteria of 5% for the correlation calculation in the stack
gas (see Fig. 3(b)). In Fig. 3(b) PCP is included in the
regression analysis and showed a r value of 0.82 which is
similar to the r observed in the ¯ue gas. The correlation
coecient for 2,3,4,6-TeCPh increased to 0.82, whereas

in the formation processes (Wikstrom et al., 1998;
Blumenstock, 1999). Furthermore, Blumenstock et al.
(1999b) showed that more than 80% of the total I-TEQ
value is assigned to the PCDF, which is true also for the
stack gas. These two arguments may explain the high
and reliable correlation between chlorinated benzenes
and the emissions of I-TEQ/F. The obtained r values for
Table 4
Correlation coecients for chlorinated PCBz isomers with sum parameters and the I-TEQa
N ˆ 24 Flue gas samples at 350°C
PCDD/F
PCDD
PCDF
I-TEQ
MCBzb
0.71
0.39
0.58
0.66
0.09
0.31
0.77
0.41
0.86
0.81
0.68
0.67
0.41
0.44
0.68
0.2
0.93
0.35
0.71
0.75
)0.23
0.11
0.73
0.22
0.95
0.8
0.85
0.27
0.65
0.78
)0.18
0.45
0.76
0.34
0.91
0.79
0.74
1,3-DiCBzc
1,4-DiCBzc
1,2-DiCBzc
1,3,5-TriCBzc
1,2,4-TriCBzc
1,2,3-TriCBzc
1,2,4,5/1,2,3,5-TeCBzc
1,2,3,4-TeCBzc
PeCBzd
0.4
0.69
0.45
0.84
0.79
0.76
HxCBz
a Signi®cant values (p-level < 5%) in bold italics.
b REMPI spectrum available in Walter et al. (1991) and Heger et al. (1999a).
c REMPI spectrum available in Trembeull et al. (1985), Rohl®ng (1988) and Weickhardt et al. (1994).
d REMPI spectrum available in Weickhardt et al. (1994) and Zimmermann et al. (1994).
0.71

514
M. Blumenstock et al. / Chemosphere 42 (2001) 507±518
Fig. 4. Regression lines and correlation results for selected
chlorinated benzenes in the ¯ue gas: (a) denotes the slope and
(b) the intercept of the regression line.
Fig. 5. Correlation for relative concentrations of monochlo-
robenzene in the ¯ue gas and the I-TEQ in the stack gas (n ˆ 9;
regression equation: log c (I-TEQ/ ꢀpg=m3† ˆ b ‡ a log c
(MCBz/ (ng/m3)).
penta- and hexachlorinated benzenes were in the same
range as found by Kaune et al. (1994, 1998).
MCBz in the ¯ue gas. As a consequence, rapid changes
in combustion conditions manifested in increasing for-
mation of chlorinated aromatics in the ¯ue gas of the
HWI could be monitored on-line e.g., with REMPI±
TOFMS via the MCBz concentration.
In Fig. 4, the regression lines and correlation results
for two selected chlorinated benzenes in the ¯ue gas are
shown. During disturbances in combustion the emitted
PCDD/F burden increased for a time span of hours and
also the concentrations of the chlorinated benzenes, es-
pecially the lower chlorinated species, rose much faster
and more drastically than the higher chlorinated ones.
These emission memory e€ects most likely are due to
carbon surfaces, formed during malfunction periods in
the plant. The layers of the freshly formed carbonaceous
structures are sources for the formation of chlorinated
aromatics (and PAH). This behaviour will be described
in more detail elsewhere (Zimmermann et al., 1999a).
The high correlation coecient value for MCB in con-
trast to the weak correlation of the low chlorinated
PCDF may be caused by the observed pattern changes
of the chlorinated furans which are associated with
disturbed combustion conditions and emission memory
e€ects (Blumenstock et al., 1998b; Zimmermann et al.,
1998, 1999a).
3.4. Evaluation of the on-line measurement feasibility of
the identi®ed PCDD/F I-TEQ surrogates with REMPI±
TOFMS laser mass spectrometry
The REMPI process is a highly selective and ecient
ionisation method for analytical mass spectrometry.
Basic aspects of this method and analytical applications
are described elsewhere (Boesl, 1991; Boesl et al., 1994;
Heger et al., 1999b). REMPI±TOFMS represents a two-
dimensional selective and sensitive analytical method,
with laser wavelength and mass as analytical parame-
ters. In the last years, REMPI±TOFMS laser mass
spectrometry based on-line monitors have been devel-
oped which can measure traces of aromatic compounds
like benzene, naphthalene, phenol or phenanthrene di-
rectly in ¯ue gases of e.g., waste incinerators in real time
(Heger et al., 1999b; Thanner et al., 1998; Zimmermann
et al., 1997). It was suggested in 1994 to use the REMPI±
TOFMS technique for indirect PCDD/F I-TEQ emis-
sion monitoring via surrogate compounds (Zimmer-
mann et al., 1994).
It should be mentioned that all the isomers of the
PCBz which are good surrogates for the emission of the
I-TEQ in the ¯ue gas show similar r values in the stack
gas (n ˆ 9) (Blumenstock et al., 1999b). If patterns are
more or less stable in the course of the ¯ue gas channel
of the HWI plant for PCDD/F and PCBz it is possible to
correlate PCBz in the ¯ue gas with the I-TEQ of the
PCDD/F in the stack gas, as demonstrated in Fig. 5. In
Fig. 5 the ¯ue gas chlorobenzene concentration is cor-
related with the stack gas emission of the I-TEQ. To
prevent disclosure of the real concentration ranges of the
study relative concentrations are plotted. The high r
value of 0.90 means that it is also possible to get a good
estimate for the I-TEQ stack emission by measuring
In Table 5, all compounds with a signi®cant corre-
lation to the I-TEQ are listed, ordered by the respective
correlation coecient. In order to ®gure out the most
suitable I-TEQ indicator for an on-line REMPI±TOF-
MS, relevant molecular parameters for the REMPI-de-
tection are also included. Detailed discussions of the
fundamental spectroscopic aspects of the REMPI-pro-
cess applied to chlorinated compounds are reported

M. Blumenstock et al. / Chemosphere 42 (2001) 507±518
515
Table 5
List of the identi®ed signi®cant I-TEQ surrogates in the ¯ue gas, ordered according to the correlation factor r (second column)a
I-TEQ
surrogate
r for
I-TEQ
Average
concentra-
tion (pptv)
One-color Cl±Cl in or- Cl±O in
tho-positionc ortho-posi-
tiond
PAH
Degree of
REMPI
REMPIb
interferencee chlorinationf spectrum in
literature
(see results)
1,2,3,4-TeCBz
MCBz
0.91
0.85
0.82
0.8
26.98
292.84
33.42
56.36
12.23
8.95
)
+
)
+
+
+
)
)
)
)
)
)
+
)
)
)
+
)
+
)
+
+
)
+
+
+
)
)
+
)
+
+
)
)
)
)
+
)
)
)
)
)
+
+
)
)
+
+
+
+
+
)
+
)
+
)
)
+
)
)
)
)
)
)
)
)
)
)
)
+
+
)
+
+
+
)
)
)
)
+
+
)
)
)
)
)
)
)
)
)
)
)
+
)
)
2,4,6-TriCPh
2,4-DiCPh
PeCBz
+
+
0.79
0.79
0.78
0.76
0.76
0.75
0.75
0.75
0.74
0.74
0.73
0.72
0.72
0.69
0.69
0.65
0.64
0.62
)
)
PeCPh
1,2-DiCBz
48.43
19.51
0.02
+
+
+
1,2,3-TriCBz
1,2,6-TriCDD
2,3,4,6-TeCPh
1,4,6-TriCDD
1,7,8-TriCDD
HxCBz
)
())
?
4.98
0.08
+
())
())
)
())
())
())
)
0.01
7.21
1,2,3-TriCDD
1,4,7-TriCDD
1,2,9-TriCDD
2,4,6-TriCDF
2,3,5,6-TeCPh
1-MCDF
0.01
0.04
0.03
0.18
6.82
0.12
?
+
?
) (202 amu)
+
1,4-DiCBz
2,3,4-TriCPh
1,3DiCDD
29.43
11.87
0.02
+
+
+
())
) (252 amu)
a Further the averaged concentrations and important parameters for an on-line REMPI detection are given (+: favourable for on-line
detection; ±: unfavourable for on-line detection).
b One-color REMPI favourable for on-line detection (one-color REMPI possible: +; two-color REMPI required: ±).
c Ortho-chlorines reduce REMPI eciency (two Cl substituents in ortho-position: ±, otherwise: +).
d Chlorines in ortho position to oxygen can reduce or quench REMPI eciency (Cl substituent in ortho-position to oxygen: ±;
otherwise: +).
e PAH can cause optical and mass interferences (no mass interference with PAH: +, otherwise: ±) (only for masses 252 and lower).
f Increased chlorine substitution results in a decrease in REMPI eciency (mono- and dichlorinated compounds: +, higher chlorinated
compounds: ±).
elsewhere (Trembeull et al., 1985; Zimmermann, 1998;
Zimmermann et al., 1996b).
(iii) Neighbouring group e€ect due to chlorine in the
ortho position: It has been observed that aromatics with
ortho chlorine substitution are less eciently detected by
REMPI with respect to their non-ortho isomers.
Brie¯y, the following aspects have been included for
evaluation of the REMPI applicability:
(i) The concentration of the indicator compound.
Concentrations below the pptv-range are hardly de-
tectable by REMPI±TOFMS, no matter whether an
e€usive- or a supersonic molecular beam inlet is used
without applying sophisticated (and time-consuming)
sample pre-treatment (enrichment) and/or pre-separa-
tion techniques.
(iv) Neighbouring group e€ect due to chlorine in the
ortho-position to oxygen: In a recent study the REMPI-
ionisation of 1-MCDD and 2-MCDD was investigated.
While 2-MCDD gave a normal REMPI spectrum, for 1-
MCDD only very weak and very broad REMPI signals
were observed, that are not useful for analytical detec-
tion (Zimmermann, 1998b). This ``ortho chlorine±oxy-
gen e€ect'' probably is due to an intermolecular
interaction between the chlorine substituent and the
oxygen (i.e., by dpCl ppO orbital interaction) which
causes either a rapid deactivation of the S1 excitation
(inter system crossing, ISC or internal conversion, IC)
and/or a large geometry shift upon excitation. The e€ect
likely can be generalised for all PCDD/F.
(ii) Applicability of one-colour REMPI±TOFMS.
The applicability of the one-colour REMPI±TOFMS
process depends on the relative position of the ®rst UV-
active singlet state and the ionisation potential of the
compound. If one-colour REMPI can not be applied, a
much more complicated two-colour REMPI-scheme has
to be used. The two-colour technique often is less sen-
sitive due to an enhanced background ionisation level
and is less feasible for ®eld applications.
(v) Mass interferences with polycyclic aromatic hy-
drocarbons (PAH): Many PAH exhibit intense and

516
M. Blumenstock et al. / Chemosphere 42 (2001) 507±518
spectroscopically broad transitions in the wavelength
region about 250±320 nm, which is used for detection of
the surrogates. The co-ionizaton of these PAH thus can
not be suppressed in this wavelength range, either by
REMPI with e€usive inlet or by Jet-REMPI. As the
concentration of PAH may be signi®cantly higher than
those of the target surrogates interferences can be an-
ticipated, like in the case of monochlorophenols and
naphthalene (128 amu), monochlordibenzofurans and
four ring-PAHs like pyrene (202 amu) or dic-
hlorodibenzodioxins and ®ve ring PAHs like benzopy-
renes (252 amu).
and can hardly be suppressed by REMPI in the required
wavelength range.
4. Conclusions
Some lower chlorinated PCDD, PCDF, PCPh and
PCBz were identi®ed as surrogates for the emission of
the toxicity carried by the PCDD/F in the ¯ue gas at
the HWI plant. Lower chlorinated PCDF showed only
in few cases signi®cant r values in the ¯ue gas. In the
case of the lower chlorinated PCDD some two- to
threefold chlorinated isomers exhibited signi®cant
higher r values in the ¯ue gas. To perform an indirect
on-line monitoring of the I-TEQ with these compounds
via REMPI±TOFMS the observed concentration range
in the ¯ue and stack gas is still too low (ppqv). Fur-
thermore, the analytical e€orts for the measurements of
two-to threefold chlorinated PCDD/PCDF (two colour
experiments) are too high and not feasible at a large
scale plant.
(vi) Degree of chlorination: Successive addition of
chlorine substituents causes a successive reduction of the
REMPI-sensitivity due to increasing rapid deactivation
of the UV-active singlet state to the triplett state (in-
tersystem crossing). Thus the surrogate should carry as
little chlorine as possible.
(vii) Availability of REMPI-spectra: If REMPI
spectroscopic data is available, the applicability of a
chosen indicator can be assessed conveniently.
Table 5 shows that the eight best indicators are
chlorinated benzenes or phenols. 1,2,3,4-Tetrachloro-
benzene (1,2,3,4 TeCBz) has been identi®ed as the sur-
rogate with the highest I-TEQ prediction accuracy
(r ˆ 0:91). However the REMPI detection of this com-
pound is rather dicult, requiring a two-colour REMPI
scheme. Furthermore 1,2,3,4 TeCBz shows a weak RE-
MPI eciency and an unusual vibronic structure in the
REMPI-spectrum due to the high number of ortho-
standing chlorine substituents (Zimmermann et al.,
1994). The second best I-TEQ surrogate (r ˆ 0:85) is
monochlorobenzene (MCBz). Within the found set
MCB is by far the optimal I-TEQ surrogate for RE-
MPI±TOFMS on-line monitoring. Recently we suc-
ceeded in performing ®rst real-time on-line REMPI±
TOFMS measurements of monochlorobenzene (MCBz)
for indirect on-line monitoring of the I-TEQ at the
hazardous waste incinerator described in this paper
(Zimmermann et al., 1999b). In contrast to the
straightforward assumption that low chlorinated
PCDD/F should be good indicators for the PCDD/F I-
TEQ value, our results show their prediction accuracy to
be signi®cantly lower than that of several PCBz and
PCPh isomers. The detection of the low chlorinated
PCDD/F in 10 ppqv-concentration range by REMPI±
TOFMS is, furthermore, very dicult and requires at
least technically sophisticated pre-concentration or
sample treatment measures. As the number of isobaric
compounds increases largely with the molecular mass,
the possibility of mass (and optical)-interferences rises
with molecular mass. For example the interferences be-
tween MCDF (202 amu) and DCDD (252 amu) and
PAH (e.g., pyrene, 202 amu and benzopyrenes, 252
amu) are evident. The concentration of the PAH is
several orders of magnitude higher than the surrogates
Chlorinated phenols, especially the lower chlorinated
ones, were not found to be reliable surrogates for the I-
TEQ value in the stack gas of the HWI. These com-
pounds were valid as surrogates in the ¯ue gas only. The
observed behaviour was due to the ¯ue gas cleaning
system of the plant, containing a multiple step wet-
scrubber system which discriminated against the more
polar lower chlorinated phenols. Pentachlorophenol and
2,3,4,6-TeCPh were compounds with a high r value in
the ¯ue and stack gas.
Most promising candidates for an easily accessible
compound group of I-TEQ surrogates at the HWI plant
were the chlorinated benzenes. We found high r values
for all degrees of chlorination, also for chlorobenzene,
which showed in the ¯ue gas an r value of 0.85, one of
the highest in this study. This substance can be easily
monitored with REMPI±TOFMS in the ¯ue gas (con-
centration range of ppt to ppb, Zimmermann et al.,
1999b) and furthermore it was shown that there exists an
correlation between the chlorobenzene concentration in
the ¯ue gas and the emitted I-TEQ in the stack gas of the
HWI.
To generalise the observed results at the HWI for the
chlorinated aromatics additional studies are required at
widely di€erent combustion facilities, e.g., municipal
waste incinerators and smaller household ®replaces.
Acknowledgements
We thank the Deutsche Bundesstiftung Umwelt,

Osnabruck (Germany) for ®nancial support and Pro-
fessor E. Rowher from Pretoria (South Africa) for re-
viewing the manuscript.

M. Blumenstock et al. / Chemosphere 42 (2001) 507±518
517
Kaune, A., Lenoir, D., Nikolai, U., Kettrup, A., 1994.
Estimating concentrations of polychlorinated dibenzo-p-
dioxins and dibenzofurans in the stack gas of a hazardous
waste incinerator from concentrations of chlorinated benz-
enes and biphenyls. Chemosphere 29, 2083±2096.
References
Blumenstock, M., 1999. Thesis, Technische, Universitat


Munchen, to be published.
Blumenstock, M., Zimmermann, R., Schramm, K.-W., Kett-
rup, A., 1999a. In¯uence of combustion conditions on the
PCDD/F-, PCB-, PCBz- and PAH-concentrations in the
post combustion chamber of a waste incineration pilot
plant. Chemosphere, in press.
Oehme, M., Mano, S., Mikalsen, A., 1987. Formation and
Presence of polyhalogenated and polycyclic compounds in
the emissions of small and large scale municipal waste
incinerators. Chemosphere 16, 143±153.


Oberg, T., Bergstrom, J.G.T., 1985. Hexachlorobenzene as an
indicator of dioxin production from combustion. Chemo-
sphere 14, 1081±1086.
Blumenstock, M., Zimmermann, R., Schramm, K.-W., Kaune,
A., Nikolai, U., Lenoir, D., Kettrup, A., 1999b. Estimation
of the dioxin emission (PCDD/F I-TEQ) from the concen-
tration of low chlorinated aromatic compounds in the ¯ue
and stack gas of a hazardous waste incinerator. Journal of
Analytical and Applied Pyrolysis 49, 179±190.
Rohl®ng, E.A., 1988. Resonantly enhanced multiphoton ion-
ization for the trace detection of chlorinated aromatics. In:
Proceedings of the 22nd Symposim on Combustion, The
Combustion Institute, 1988, p. 1843.
Blumenstock, M., Zimmermann, R., Schramm, K.-W., Kaune,
A., Nikolai, U.., Lenoir, D., Kettrup, A., 1998a. Estimating
I-TEQ emissions of polychlorinated dibenzo-p-dioxins
(PCDD) and dibenzofurans (PCDF) from lower chlorinated
Schramm, K.-W., Kaune, A., Lehnhardt, R., Hofmaier, A.,
Henkelmann, B., Kettrup, A., 1998. Isokinetic sampling of
PCDD/F response in low and high volatile fractions of
a wood incinerator. Organohalogen compounds 36, 289±
292.
PCDD/F and benzenes (PCBz) at
a hazardous waste
incinerator (HWI). Organohalogen Compounds 36, 47±52.
Blumenstock, M., Zimmermann, R., Schramm K.-W, , Hen-
kelmann, B., Kettrup, A., 1998b. Presence of polychlori-
nated dibenzo-p-dioxins (PCDD), dibenzofurans (PCDF),
biphenyls (PCB), chlorinated benzenes (PCBz) and polycy-
clic aromatic hydrocarbons (PAH) under various combus-
Schramm, K.-W., Merk, M., Henkelmann, B., Kettrup, A.,
1995. Leaching of PCDD/F from ¯y ash and soil with ®re-
extinguishing water. Chemosphere 30, 2249±2250.
Thanner, R., Oser, H., Grotheer, H.H., 1998. Time-resolved
monitoring of aromatic compounds in an experimental
incinerator using an improved jet-resonance-enhanced mul-
tiphoton ionization system Jet-REMPI. European Mass
Spectrometry 4 (3), 215±222.
tion conditions in
a
Organohalogen Compounds 36, 59±63.
post combustion chamber.
Boesl, U., Zimmermann, R., Weickhardt, C., Lenoir, D.,
Schramm, K.-W., Kettrup, A., Schlag, E.W., 1994. Reso-
nance enhanced laser ionisation as a compound-selective ion
source for mass spectrometry: technique and experimental
setup. Chemosphere 29, 1429±1440.
Trembeull, R., Sin, C.H., Li, P., Pang, H.M., Lubman, D.M.,
1985. Applicability of resonant two photon ionisation in
supersonic beam mass spectrometry to halogenated aro-
matic hydrocarbons. Analytical Chemistry 57, 1186±1192.
VDI 3499, 1990. Messen von Emissionen (german guideline for
PCDD/F sampling). Blatt, pp. 1±3.
Boesl, U., 1991. Multiphoton excitation and mass-selective ion
detection for neutral and ion spectroscopy. Journal of
Chemical Physics 95, 29.
Walter, K., Scherm, K., Boesl, U., 1991. Spectroscopy of the
ionic ground state of monohalogenated benzenes. Journal of
Physical Chemistry 95, 1188±1195.
Dellinger, B., Hall, D.L., 1986. Surrogate compounds for
monitoring the e€ectiveness of incineration systems. Haz-
ardous Waste Management 36, 179±183.
Weickhardt, C., Zimmermann, R., Schramm, K.-W., Boesl, U.,
Schlag, E.W., 1994. Laser mass spectrometry of the di-, tri-
and tetrachlorobenzenes: isomer selective ionization and
detection. Rapid Communications in Mass Spectrometry 8,
381±384.

Gullett, B.K., Wikstrom, E., 1999. Use of mono- to tri-
chlorinated dibenzodioxin (CDD) and dibenzofuran (CDF)
congeners/homologues as indicators of tetra- to octa-CDD/
CDF emissions from municipal waste and waste/coal
combustion. Organohalogen Compounds 36, 179±182.
Heger, H.J., Boesl, U., Zimmermann, R., Dorfner, R., Kettrup,
A., 1999a. On-line multi-component analysis of aromatic
compounds in complex gas mixtures with resonant laser
mass spectrometry. European Mass Spectrometry 5, 51±57.
Heger, H.J., Zimmermann, R., Dorfner, R., Beckmann, M.,
Griebel, H., Kettrup, A., Boesl, U., 1999b. On-line emission
analysis of polycyclic aromatic hydrocarbons down to pptv
concentration levels in the ¯ue gas of an incineration pilot
plant with a mobile resonance-enhanced multiphoton ion-
ization time-of-¯ight mass spectrometer. Analytical Chem-
istry 71, 46±57.
Weickhardt, C., Zimmermann, R., Boesl, U., Schlag, E.W.,
1993. Laser mass spectrometry of dibenzodioxin, dibenzofu-
ran and two isomers of dichlorodibenzodioxins: selective
ionisation. Rapid Communications in Mass Spectrometry 7,
183±185.

Wikstrom, E., Persson, A., Marklund, S., 1998. Secondary
formation of PCDDs, PCDFs, PCBs, PCBz, PCPhs and
PAHs during MSW combustion. Organohalogen Com-
pounds 36, 65±68.
Zimmermann, R., Blumenstock, M., Heger, H.J., Schramm,
K.-W., Kettrup, A., 1999a. Emission memories of non-
chlorinated and chlorinated aromatics in the ¯ue gas of
incineration plants after transient disturbances of the
combustion conditions. Environmental Science and Tech-
nology (in press).
Kaune, A., Lenoir, D., Schramm, K.-W., Zimmermann, R.,

Kettrup, A., Jaeger, K., Ruckel, H.G., Frank, F., 1998.
Chlorobenzenes and chlorophenols as indicator parameters
for chlorinated dibenzodioxins and dibenzofurans in incin-
eration processes: In¯uences of various facilities and sam-
pling points. Environmental Engineering Science 15, 85±95.
Zimmermann, R., Heger, H.J., Blumenstock, M., Dorfner, R.,
Schramm, K.-W., Boesl, U., Kettrup, A., 1999b. On-line
measurement of chlorobenzene in waste incineration ¯ue

518
M. Blumenstock et al. / Chemosphere 42 (2001) 507±518
gas as a surrogate for the emission of polychlorinated results. Rapid Communications in Mass Spectrometry 11,
dibenzo-p-dioxins/-furans (I-TEQ) with a mobile resonance
laser ionization-time-of-¯ight mass spectrometer (REMPI±
TOFMS). Rapid Communications in Mass Spectrometry
13, 307±314.
1095±1102.
Zimmermann, R., Rohwer, E.R., Heger, H.J., Schlag, E.W.,
Kettrup, A., Gilch, G., Lenoir, D., Boesl, U., 1996a.
Resonance ionization laser mass spectrometry: New possi-
bilities for on-line analysis of waste incinerator emissions.
In: Proceedings of the Eighth International Symposium on
Resonance Ionization Spectroscopy RIS 96, Penn State
University, USA.
Zimmermann, R., Heger, H.J., Dorfner, R., Boesl, U., Blu-
menstock, M., Lenoir, D., Kettrup, A., 1998. A mobile laser
mass spectrometer (REMPI±TOFMS) for continuous mon-
itoring of toxic combustion byproducts: real-time on-line
analysis of PAH in waste incineration ¯ue gases. Combus-
tion Science and Technology 134, 87±101.
Zimmermann, R., Lenoir, D., Kettrup, A., Nagel, H., Boesl,
U., 1996b. On-line emission control of combustion process-
es by laser induced resonance-enhanced multi-photon ion-
ization combined with a time-of-¯ight mass analyzer. In:
The 26th International Symposium on Combustion, The
Combustion Institute, Pittsburgh, pp. 2859±2868.
Zimmermann, R., Boesl, U., Weickhardt, C., Lenoir, D.,
Schramm, K.-W., Kettrup, A., Schlag, E.W., 1994. Reso-
nance enhanced laser ionisation as a compound-selective ion
source for mass spectrometry: First results for chlorinated
dibenzodioxins, chlorinated biphenyls and chlorinated
benzenes. Chemosphere 29, 1877±1888.
Zimmermann, R., 1998b. Dibenzo-p-dioxin and its chlorinated
congeners: molecular geometry and electronic structure. In:
Proceedings of the Ninth Resonance Ionization Spectros-
copy Symposium, June 1998, Manchester, UK, American
Institute of Physics (AIP) Conference Proceedings Series,
vol. 454, AIP, New York, pp. 131±136.
Zimmermann, R., Heger, H.J., Kettrup, A., Boesl, U., 1997. A
mobile resonance-enhanced multiphoton ionization time-of-
¯ight mass spectrometry device for on-line anlysis of
aromatic pollutants in waste incineration ¯ue gases: First