Journal of Chromatographic Science, Vol. 40, September 2002
N-chloramines from individual solutions, each containing an
excess of free amine. However, the peak areas for the four compo-
nents were not equal. Two factors contributed to the differences
in the peak areas: equivalent molar responses were not obtained
between the four analytes, and N-chloramines in the presence of
excess free amines redistributed the active chlorine with time
that describes disproportionation is second order in chloramine
concentration, at least in the case of NHCl (19). The linearity of
2
the calibration curves (Table II) and the linearity of the
ln(chloramine) versus time plot (Figure 5) would appear to rule
out a second order loss process. Thus, if oncolumn losses occur
through disproportionation, a special mechanism, probably
involving catalysis by a column component, must be at work.
Defining a clear source of chlorine-consuming reductants in
the LC experiment has proven elusive. Experiments with different
frit materials, columns, and NaOCl preoxidation of columns seem
to exonerate the LC column as the direct source of the reductants.
Although organic modifiers in the mobile phase might reasonably
be suspected as consumers of oxidizing analytes, this hypothesis
is contradicted by the observations that the amount of loss dimin-
ished as the amount of modifier increased, and that losses
occurred with chemically diverse modifiers.
Fortunately, we have identified a set of LC operating conditions
in which the losses of monochloramine, N-Cl-AlaAla, and N-Cl-
LeuAla are minimized and linear calibration curves are obtained.
The nearly identical slopes indicate that a calibration based on
any of these chloramines could be used to estimate active chlo-
rine in a wastewater sample. Calibration of analytes exhibiting
higher losses (such as N-chloropiperidine) would not be
amenable to such an approach. If it were desirable to determine
N-Cl-piperidine, it could be calibrated independently. However, in
dechlorinated wastewaters, most residual chloramines are
expected to be peptide and protein chloramines (6,7,17); these
should display the favorable behavior we have found for N-Cl-
AlaAla and N-Cl-LeuAla.
(15).
Discussion
A goal of this research has been to develop an LC system that
provides an equivalent molar response for unstable chloramines
in complex matrices. As judged by identical signals for all of the
test compounds using FIA (Figure 3A), the postcolumn detection
system has been optimized to the point that equivalent molar
responses are achieved. Unfortunately, the introduction of the LC
separation column results in analyte-dependent signal losses that
are consistent with the known reactivity of the different chlo-
ramines. Losses are most severe with monochloramine and N-Cl-
piperidine. These chloramines are better proton acceptors and are
consequently more reactive than the N-chloropeptides in many
reactions, including reduction reactions (6). Greater losses at
lower pH (Figure 3B versus 7) and at comparable concentrations
of methanol versus acetonitrile (Figure 6 versus 3B) are consis-
tent with Brønsted acid catalysis, which is common in chlo-
ramine reactions.
For all analytes except N-chloropiperidine, the first step in reac-
tive losses may proceed through chloramine disproportionation:
2RNHCl
RNCl + RNH2
Eq. 2
2
¡
References
This disproportionation is favored by acid catalysis at low pH
1
. R.A. Issac and J.C. Morris. “Transfer of active chlorine from chlo-
ramine to nitrogenous organic compounds”. In Water Chlorination,
Environmental Impact and Health Effects, Vol. 4. R.L. Jolley, W.A.
Brungs, R.B. Cotruvo, R.B. Cumming, J.S. Mattice, and V.A. Jacobs,
Eds. Ann Arbor Science, Ann Arbor, MI, 1983, pp. 63–75.
and inhibited by excess amine (19,20). In order to minimize this
reaction, the chloramine samples were always prepared at neutral
pH and with excess free amine. In the FIA mode, in which the ana-
lyte and excess amine are not separated, this reaction is not
favored, resulting in maximum detector responses. In the LC
mode, the separation of the chloramine from the parent amine
may initiate disproportionation, particularly with the pH 2.75
mobile phase. Lower detector signals in the LC mode would occur
after disproportionation for two reasons. Dichloramines formed
as the chloramines traverse the column and would be dissipated
and “lost” to the chromatographic peak area because of greater
retention, thereby contributing only to the background detector
current. Relative to monochloramines, dichloramines also could
show enhanced reactivity toward reducing contaminants or
exhibit autodecomposition. Organic dichloramines autodecom-
pose by dehydrohalogenation, which produces chlorimines and
nitriles with the loss of one and both equivalents of oxidizing
capacity, respectively (21,22). N-chloropiperidine also autode-
2
. J.M. Antelo, F. Arce, and M. Parajó. Kinetic study of the formation of
N-chloramines. Int. J. Chem. Kinetics 27: 637–47 (1995).
3
. T.I. Bieber and M.L. Trehy. “Dihaloacetonitriles in chlorinated natural
waters”. In Water Chlorination, Environmental Impact and Health
Effects, Vol. 4. R.L. Jolley, W.A. Brungs, R.B. Cotruvo, R.B. Cumming,
J.S. Mattice, and V.A. Jacobs, Eds. Ann Arbor Science, Ann Arbor, MI,
1983, pp. 85–96.
4
. G.C. White. The Handbook of Chlorination and Alternate
Disinfectants, 4th Ed. John Wiley, New York, NY, 1999.
5
. J.S. Mattice and S.C. Tsai. “Total residual chlorine as a regulatory
tool”. In Water Chlorination, Environmental Impact and Health
Effects, Vol. 4. R.L. Jolley, W.A. Brungs, R.B. Cotruvo, R.B. Cumming,
J.S. Mattice, and V.A. Jacobs, Eds. Ann Arbor Science, Ann Arbor, MI,
1983, pp. 901–12.
6
. R.H. Jameel and G.R. Helz. Organic chloramines in disinfected
wastewaters: rates of reduction by sulfite and toxicity. Environ.
Toxicol. Chem. 18: 1899–1904 (1999).
composes to an imine (23). In contrast, NHCl autodecomposes
7. J.S. Jensen and G.R. Helz. Rates of reduction of N-chlorinated pep-
tides by sulfite: relevance to incomplete dechlorination of wastewa-
ters. Environ. Sci. Technol. 32: 516–22 (1998).
2
primarily to N (20,24).
2
Two objections can be raised to the hypothesis that dispropor-
tionation is involved in the primary loss initiation step, based on
the known behavior of chloramines in free solution. First, dispro-
portionation is kinetically relatively slow, and second, the rate law
8
. T. Kotiaho, M.J. Hayward, and R.G. Cooks. Direct determination of
chlorination products of organic amines using membrane introduc-
tion mass-spectrometry. Anal. Chem. 63: 1794–1801 (1991).
9. C. Shang and E.R. Blatchley. Differentiation and quantification of free
454