Mercuric chloride-catalyzed hydrolysis of the new antifouling compound Irgarol 1051

Article abstract of DOI:10.1016/S0043-1354(98)00186-9

Irgarol 1051, 2-methylthio-4-tert-butylamino-6-cyclopropylamino-s-triazine is a newly developed herbicidal additive for use in copper-based antifouling paints. It is intended to replace the antifouling agent tributyltin, which has been regulated internationally due to its severe impact on the aquatic ecosystem. However, there is no information in the open literature on the abiotic degradation of Irgarol, a fact that hinders the assessment of its ultimate impact on the environment. This study showed that mercuric chloride was capable of rapidly catalyzing the hydrolysis of Irgarol 1051 in distilled water and buffer solutions. The degradation appeared to follow the reaction of a catalyzed hydrolysis and was not significantly affected by the pH tested (5 to 9). All other 5 heavy metal salts tested (AgNO₃, CdCl₂, CuSO₄, PbCl₂ and ZnCl₂) had practically no catalytic property on Irgarol hydrolysis, implying the involvement of a specific activity for Hg²⁺ in this reaction. The mechanism for the catalyzed hydrolysis may be the formation of bidentate chelation through nitrogen (No. 5) on the ring and the nitrogen on the cyclopropylamino side chain in Irgarol 1051 with the Hg²⁺ ion. The resulting four-member chelate complex would weaken the cyclopropyl-amino bond considerably, thus facilitating the hydrolysis reaction. Ultraviolet spectroscopy of the reaction mixtures and the identification of Irgarol hydrolysis product M1 (2-methylthio-4-tert-butylamino-6-amino-s-triazine) by GC-MS and LC-MS provided the basis for the proposed mechanism on the HgCl₂-catalyzed hydrolysis of Irgarol 1051. M1 appeared to be more stable than the parent compound Irgarol 1051, thus implying its possible accumulation in the environment. One practical aspect of this work is that HgCl₂ should not be used in preserving water samples in Irgarol 1051 monitoring programs.

Full text of DOI:10.1016/S0043-1354(98)00186-9

Wat. Res. Vol. 33, No. 1, pp. 155±163, 1999
1998 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
#
PII: S0043-1354(98)00186-9
0043-1354/98 $ - see front matter
MERCURIC CHLORIDE-CATALYZED HYDROLYSIS OF
THE NEW ANTIFOULING COMPOUND IRGAROL 1051
1
D. LIU *, G. J. PACEPAVICIUS , R. J. MAGUIRE , Y. L. LAU ,
1
1
1
2
H. OKAMURA and I. AOYAMA
2
*
M
1
2
National Water Research Institute, Burlington, Ont., Canada L7R 4A6 and Research Institute for
Bioresources, Okayama University, Kurashiki 710, Japan
(
First received January 1998; accepted in revised form April 1998)
AbstractÐIrgarol 1051, 2-methylthio-4-tert-butylamino-6-cyclopropylamino-s-triazine, is a newly devel-
oped herbicidal additive for use in copper-based antifouling paints. It is intended to replace the anti-
fouling agent tributyltin, which has been regulated internationally due to its severe impact on the
aquatic ecosystem. However, there is no information in the open literature on the abiotic degradation
of Irgarol, a fact that hinders the assessment of its ultimate impact on the environment. This study
showed that mercuric chloride was capable of rapidly catalyzing the hydrolysis of Irgarol 1051 in dis-
tilled water and bu€er solutions. The degradation appeared to follow the reaction of a catalyzed hy-
drolysis and was not signi®cantly a€ected by the pH tested (5 to 9). All other 5 heavy metal salts tested
(
AgNO
implying the involvement of a speci®c activity for Hg
lyzed hydrolysis may be the formation of bidentate chelation through nitrogen (No. 5) on the ring and
3 2 4 2 2
, CdCl , CuSO , PbCl and ZnCl ) had practically no catalytic property on Irgarol hydrolysis,
2+
in this reaction. The mechanism for the cata-
2+
the nitrogen on the cyclopropylamino side chain in Irgarol 1051 with the Hg ion. The resulting four-
member chelate complex would weaken the cyclopropyl-amino bond considerably, thus facilitating the
hydrolysis reaction. Ultraviolet spectroscopy of the reaction mixtures and the identi®cation of Irgarol
hydrolysis product M1 (2-methylthio-4-tert-butylamino-6-amino-s-triazine) by GC-MS and LC-MS pro-
vided the basis for the proposed mechanism on the HgCl
appeared to be more stable than the parent compound Irgarol 1051, thus implying its possible accumu-
lation in the environment. One practical aspect of this work is that HgCl should not be used in preser-
2
-catalyzed hydrolysis of Irgarol 1051. M1
2
ving water samples in Irgarol 1051 monitoring programs. # 1998 Elsevier Science Ltd. All rights
reserved
Key words: Irgarol 1051, antifouling compound, herbicide, hydrolysis, abiotic degradation, chelate,
mercuric chloride, heavy metals.
INTRODUCTION
about 100 and 200 days, respectively (Ciba
Geigy, 1995). It is also immune to bacterial
Organotin biocides (e.g. tributyltin and triphenyl-
degradation, but can be biotransformed by the
tin) have been widely used in the formulations
white rot fungus Phanerochaete chrysosporium via
of antifouling paint. However, their wide en-
the mechanism of N-dealkylation to yield
stable metabolite M1 (Liu et al., 1997). Irgarol
051 is highly toxic to algae and can cause a
a
vironmental distribution (Langston et al., 1990)
and non-selective biotoxicity (Beaumont and
Newman, 1986) have led the OECD countries in
1
signi®cant change in structure of the marine per-
iphyton community at an ambient concentration
1
988 to restrict their use in antifouling paints
Evans et al., 1995). Since then copper-based
(
�1
of 0.063±0.25 mg l
This concentration is well within the ranges
(
(Dahl and Blanck, 1996).
biocides, forti®ed with a herbicide, have largely
replaced organotins in antifouling paints world-
wide (Readman et al., 1993). Irgarol 1051, 2-
methylthio-4-tert-butylamino-6-cyclopropylamino-s-
triazine, is a newly developed herbicidal additive for
use in copper-based antifouling paints.
�1
0.002±0.19 mg l ) detected in the contaminated
areas of Southern England and the coastal
waters of France (Dahl and Blanck, 1996).
Thus, the continuous use of Irgarol 1051 in
antifouling paints will likely damage microalgal
communities in contaminated coastal waters.
Little is known to date about the long-term
toxicity and degradability of Irgarol 1051 (Toth
et al., 1996). Irgarol 1051 is not considered to
be readily biodegradable and its degradation in
sea and freshwater is slow, with half-lives of
To better understand the ultimate impact of
Irgarol 1051 on the aquatic ecosystem, factors and
processes a€ecting the environmental persistence of
this chemical have become one of the major
research topics in our laboratories since 1995. In
this communication, we report on the rapid hy-
drolysis of Irgarol 1051 in aqueous solution by mer-
*
Author to whom all correspondence should be addressed.
Fax: +1-905-3364420].
[
155

1
56
D. Liu et al.
For GC analysis, the toluene extracts were analyzed on
a Hewlett Packard 5890 gas chromatograph equipped with
a nitrogen±phosphorus detector (3008C) and a ¯ame ioniz-
ation detector (3008C) utilising an oven program with a
curic chloride. The possible mechanism involved in
the catalyzed hydrolysis of Irgarol 1051 is also dis-
cussed.
2
min hold at 808C and a temperature ramp of 108C/min
to 1508C followed by a temperature ramp of 48C/min to
808C and a ®nal temperature ramp of 88C/min to 3008C.
MATERIAL AND METHODS
2
The columns used were dual DB5 coated capillary col-
umns (0.25 mm 27 m) which had been installed into the
injector (2008C) in the splitless mode with a constant
helium carrier ¯ow of 0.8 ml/min.
For mass spectral analysis of the toluene extracts, the
work was performed via two GC-MS instruments. The
Chemicals
Irgarol 1051 [(2-methylthio-4-tert-butylamino-6-
cyclopropylamino-s-triazine), identi®cation No.
8
4611.0] of high grade (95%) was a gift of Ciba-
Geigy Canada, Mississauga, Ontario. All the test ®rst one used the same temperature program and column
stationary phase (0.25 mm 30 m) as the above GC analy-
inorganic and organic chemicals were obtained
sis on a Hewlett Packard 5971A mass selective detector
from BDH Chemicals (Toronto, Canada). The inor-
(
electron impact (EI) mode with an ionization potential of
70 eV and a source temperature of 1908C. The scan range
MSD) and MS Chem Station. The MSD was operated in
ganic salts used in the test experiments included
AgNO CdCl O, CuSO O, HgCl
PbCl and ZnCl
were obtained from Caledon Laboratories,
Georgetown, Ontario. The sodium sulphate used
for drying organic extracts was heated to 5008C for
3
,
2
Á2.5H
2
4
Á5H
2
2
,
. Pesticide grade organic solvents was 50±500 amu. The second GC-MS instrument was a
HP 5890 Series II, coupled to a Jeol GC Mate. The dual
2
2
columns used were HP-5MS (0.25 mm Â30 m), with a
temperature programming isothermally at 508C for 2 min,
then at 408C/min to 2008C and at 108C/min to 3008C.
For HPLC analysis, an Hitachi L7000 system equipped
with a Develosil ODS UG-5 column (2 Â 150 mm) and
photodiode array detector (225 nm) was used. The mobile
phases were 10 mM ammonium acetate (pH 4.6) and
MeCN, with a linear gradient increase of MeCN concen-
tration from 30 to 100%. The ¯ow rate was 0.2 ml/min
and the volume of sample injected was 10 ml. For LC-MS
analysis, an Hitachi M-1200H system equipped with
2
4 h before use. All glassware were also rinsed with
pesticide grade solvents before use. All other chemi-
cals used in the experiments were reagent grade or
better.
Experimental procedure
Hydrolysis of Irgarol 1051 was tracked by observing the
disappearance of the parent compound and the appear- Develosil ODS UG-5 column (4.6Â 150 mm) and photo-
ance of its degradation products in the reaction mixture. diode array detector (225 nm) was used. The mobile phase
The measurement was made by using HPLC (high per-
consisted of 50% ammonium acetate and 50% MeCN.
formance liquid chromatography), LC-MS (liquid chroma- The ¯ow rate was at 1.0 ml/min with the sample volume
tography mass spectrometry), GC (gas chromatograph), injected being 10 or 20 ml. The ionization was APCI (ato-
GC-MS (gas chromatography mass spectrometry) and UV mospheric pressure chemical ionization), with the aperture
(ultraviolet) spectrophotometric analysis. Typically, 5 mg
temperature at 1308C and the needle voltage at 3000 V.
of Irgarol 1051 in 1 l of distilled water or in bu€ered sol-
ution (20 mM ammonium acetate±acetic acid±sodium hy-
The scan range was 10±500 amu.
Interactions between heavy metals and Irgarol 1051
droxide for pH 5.0, 7.0 and 9.0) was treated with HgCl
20 mg/l). The reaction mixture was kept at room tempera- Shimadzu model UV-260 double beam UV-VIS recording
2
were also studied spectrophotometrically using
a
(
ture (218C) for 2 h before analyzing for Irgarol 1051 and
its hydrolysis products.
spectrophotometer. An absorbance spectrum was typically
taken between 190 and 400 nm.
To study the e€ect of various heavy metals (in the salt
form) on the hydrolysis of Irgarol 1051, a ®nal concen-
tration of 10, 20 and 100 mg/l of AgNO Á2.5H
3
, CdCl
2
2
O,
, was each
RESULTS AND DISCUSSION
CuSO O, HgCl , MeHgCl, PbCl or ZnCl
2
4
Á5H
2
2
2
individually tested in the Irgarol solution (5 mg/l). To
assess the concentration e€ect of mercuric chloride on the
hydrolysis of Irgarol 1051, HgCl at ®nal concentrations
2
of 0, 0.1, 1, 10, 20 and 100 mg/l was used. Kinetic studies
of hydrolysis were conducted by removing an aliquot (50±
To exclude as much as possible the interference
among chemicals in the reaction mixture, distilled
water was used as the medium in which the e€ect
of pH on the stability of Irgarol 1051 was investi-
1
00 ml) of the reaction mixtures at di€erent times (0, 1, 10 gated. Fresh distilled water was adjusted to pH
and 60 min, 2, 6, 24, 48 and 120 h). The percentages of values of 5.0, 7.0 and 9.0 with diluted acid (0.05 N
remaining Irgarol 1051 and the hydrolysis products
formed (mainly M1) were measured through GC, GC-MS,
cing Irgarol 1051 to the test solutions. Figure 1
HPLC and LC-MS analysis.
HCl) and/or base (0.05 N NaOH) prior to introdu-
shows that Irgarol 1051 was very stable in distilled
Sample preparation and chemical analysis
water at the acidic, neutral and alkaline pH tested
For the analysis of Irgarol 1051 and its degradation (5.0, 7.0, 9.0). No signi®cant change in Irgarol con-
products by GC or GC-MS, an aliquot (100 ml) of the
reaction mixture was transferred to a 500-ml separatory
tures during the 48-h test period, implying its
funnel containing 50 ml of DCM (dichloromethane). The
centrations was observed in all the reaction mix-
inherent stability in aqueous solution. Since the pH
extraction was repeated once with another 50 ml of DCM
and the combined DCM extracts with a toluene keeper for most natural waters ranges from
5 to 9
were reduced to 5 ml on a rotary evaporator. Further con- (Miyamoto et al., 1990), the continuous use of
centration and solvent exchange into toluene were per-
formed under a nitrogen stream. For analysis by HPLC or
result in its accumulation in the natural aquatic en-
LC-MS, the combined DCM extracts were evaporated to
dryness and then the residue was re-dissolved into 0.3 ml
of MeCN (acetonitrile).
Irgarol 1051 as an antifouling agent may eventually
vironment. Irgarol 1051 has been found to be
highly stable in the marine environment and has

Hydrolysis of the antifouling agent, Irgarol
157
Fig. 2. Formation of the Irgarol's degradation product
M1 in the presence and absence of mercuric chloride.
Fig. 1. Time course of the Irgarol's concentration pro®les
in the presence and absence of mercuric chloride.
et al., 1997). Results from this study and our pre-
vious investigation (Liu et al., 1997) strongly
been detected in areas close to intense boating ac- suggest that M1 could be a major and perhaps ulti-
tivity (Scarlett et al., 1997).
mate degradation product during the chemical and
Mercuric chloride is a microbial inhibitor, com- biological degradation of Irgarol 1051. The obser-
monly used in the control of biodegradation studies vation of M1 formation from Irgarol 1051 via
to di€erentiate a chemical reaction from a micro- chemical or biological processes is of interest, par-
biologically mediated process. During the early ticularly in view of the fact that hydroxyatrazine,
stage of our study on microbial degradation of an atrazine metabolite, could also be formed from
Irgarol 1051, substantial amounts of breakdown atrazine by either chemical or biological reactions
products from Irgarol 1051 were found in the con- (Mandelbaum et al., 1993). Both atrazine and
trol ¯asks containing mercuric chloride. This led to Irgarol 1051 are s-triazine herbicides and thus they
the replacement of HgCl
nide) as an inhibitor in all subsequent experiments
2
by KCN (potassium cya- likely share this similar property.
GC-MS con®rmation of M1 as a hydrolytical
on microbial transformation of Irgarol 1051 (Liu et product in the reaction mixture is shown in Fig. 3.
al., 1997). The process and mechanism involved in Figure 3(a) shows the total ion chromatogram of
this chemical degradation of Irgarol 1051 by mercu- the DCM extract from an initial 10-min sample
ric chloride were further investigated in the present taken from the neutral distilled water containing
study. Figure 1 clearly shows that, in the presence 5 mg/l Irgarol 1051 and 10 mg/l HgCl
of a low concentration of HgCl
(20 mg/l), Irgarol matogram indicates the presence of two major com-
051 would undergo a rapid chemical degradation pounds in the extract, Irgarol 1051 and the
2
. The chro-
2
1
in distilled water. 10 min after the initiation of the suspected hydrolysis product M1, which eluted
experiment, more than 90% of the added Irgarol from the column approximately 1.3 min earlier than
had disappeared from the reaction mixture. The the parent compound Irgarol. Figure 3(c) shows the
very fast disappearance of Irgarol 1051 strongly (EI)GC-MS spectrum for Irgarol 1051. The mass
suggests that the degradation of Irgarol 1051 in dis- spectrum is characterized by a molecular ion at m/z
tilled water by HgCl
2
closely follows the reaction of 253 and major fragment ions at m/z 238, 196, 182
a catalyzed chemical hydrolysis. Reactions of a cat- and 112. These fragment ions can be attributed to
alyzed hydrolysis are typically very fast to ®nish fragmentation and/or rearrangements in the mol-
(
Gamble and Khan, 1988), as was the case in the ecular ion and fragment ions (Liu et al., 1997).
present investigation.
Figure 3(b) shows the mass spectrum of the sus-
The HgCl
-catalyzed hydrolysis of Irgarol 1051 in pected new hydrolysis product M1 formed during
distilled water appears to be pH-independent the HgCl -catalyzed hydrolysis of Irgarol 1051 in
Fig. 1). Virtually all the Irgarol 1051 had disap- neutral distilled water. It can be seen that M1 has a
2
2
(
peared from the acidic, neutral and alkaline distilled molecular ion at m/z 213 and with major fragment
water (pH 5.0, 7.0 and 9.0) in less than 2 h. The ions at m/z 198, 157 and 111. The mass spectrum
enhanced hydrolysis of Irgarol 1051 by HgCl was pro®le of this M1 mirrors exactly the one isolated
2
also con®rmed by another experiment, in which the during the fungal biotransformation of Irgarol 1051
appearance of the Irgarol's hydrolysis product M1 (Liu et al., 1997). Thus, the M1 in the DCM extract
(
2-methylthio-4-tert-butylamino-6-amino-s-triazine)
2
of the HgCl -catalyzed hydrolysis of Irgarol 1051 in
in acidic, neutral and alkaline distilled water was the neutral distilled water is very likely to be the
followed (Fig. 2). It can be seen that the formation compound 2-methylthio-4-tert-butylamino-6-amino-
of M1 was also not in¯uenced by the pH of the dis- s-triazine.
tilled water used. M1 is a major metabolite resulting Many triazines and their metabolites are ther-
from the biotransformation of Irgarol 1051 by the mally labile and cannot be always analyzed directly
white rot fungus Phanerochaete chrysosporium (Liu or reliably by GC or GC-MS. LC-MS and HPLC-

1
58
D. Liu et al.
Fig. 3. Chromatogram and mass spectra: (a) total ion chromatogram of the DCM extract from an in-
itial 10-min sample taken from the neutral distilled water containing 5 mg/l Irgarol 1051 and 10 mg/l
HgCl
2
, (b) low-resolution mass spectrum of M1 from the peak M1 in (a), and (c) low-resolution mass
spectrum of Irgarol from the Irgarol peak in (a).
+
MS o€er alternatives for the direct analysis of tria- ion [M + H ] of Irgarol 1051 and the latter (m/z
zine herbicides and their degradation products 214) being the result of the loss of a cyclopropyl
Voyksner et al., 1987). In general, many of these group from the protonated molecular ion m/z 254.
(
degradation products have a polar structure, and APCI-MS employs soft ionization and thus the
thus can only be handled by GC or GC-MS after mass spectra acquired by this technique are very
+
derivatization (Schroder, 1997). Figure 4 demon- simple, mainly consisting of only [M + H ] ions
strates the application of LC-MS in the identi®- (Voyksner et al., 1987). The peak at 4.03 min in the
cation of Irgarol 1051 and its degradation products. LC pro®le [Fig. 4(a)] is the degradation product M1
2
The LC pro®le of the DCM extract from a reaction resulting from the HgCl -catalyzed hydrolysis of
mixture containing Irgarol 1051 and HgCl is Irgarol 1051. Its APCI-MS spectrum is shown in
2
shown in Fig. 4(a). The high background noise is Fig. 4(b). The spectrum exhibits only two major
mainly due to salt (ammonium acetate) in the ions, the m/z 214 being the protonated M1 molecu-
mobile phase. The peak at 11.32 min is the parent lar ion and the m/z 158 being the result of the loss
compound Irgarol 1051 and its APCI (atmospheric of a tert-butyl group from the molecular ion m/z
pressure chemical ionization) mass spectrum is 214. The peak at 6.95 min in the LC pro®le in
shown in Fig. 4(d). The spectrum is very simple Fig. 4(a) is probably the M1 monopotassium
showing only two major ions, m/z 254 and 214, the adduct, an artifact formed under the conditions of
former (m/z 254) being the protonated molecular LC-APCI/MS analysis. Its mass spectrum is shown

Hydrolysis of the antifouling agent, Irgarol
159

1
60
D. Liu et al.
in Fig. 4(c). The ion m/z 252 is a M1 monopotas- mately 50%. The amount of M1 produced was
+
sium adduct, i.e. [M1 + K ] = (213 + 39 = 252). found to be only about half of what was produced
The ion m/z 214 is the protonated M1 molecular from the reaction mixture containing no EDTA.
Direct evidence of the ability of mercuric chloride
+
ion [M1 + H ] = (213 + 1 = 214). The ions m/z
1
96 and 158 can be attributed to reactions involving to react with Irgarol 1051 came from the UV (ultra
+
adduct formation; [M1 + K � isobutylene] = violet) spectra (Fig. 6) of the reaction mixtures con-
+
213 + 39 � 56 = 196) and [M1 + K + H
+
(
taining a ®xed amount (7 mg/l) of Irgarol 1051 and
+
�
K � isobutylene] = (213 + 39 + 1 � 39 � 56 = various amounts (0±22 mg/l) of HgCl . This evi-
2
1
58). During ionization of pesticides by APCI, par- dence was further supported by the gas chromato-
ticularly those pesticides containing nitrogen (e.g. graphic determination of Irgarol 1051 and its
atrazine, metolachlor, etc.), molecular and molecu- hydrolysis product M1 in the reaction mixtures
lar adduct ions are formed preferentially over the (Fig. 7). It should be noted here that UV analysis
fragmented ions (Schroder, 1997). Both Irgarol for organics is less sensitive than that by GC and
1
051 and its degradation product M1 contain nitro- thus the amount of HgCl used in the reaction mix-
2
gen in their molecular structures and thus their tures for the experiment of UV spectral analysis
APCI-MS spectra exhibit molecular and adduct was doubled in comparison with the experiment for
ions accordingly (Fig. 4).
The reaction of mercury-catalyzed hydrolysis of minute quantity of HgCl
GC analysis. The results of Fig. 6 indicate that a
(0.22 mg/l) was able to
2
Irgarol 1051 was also studied in bu€er solutions at shift the absorbance maximum for Irgarol 1051
pH 5.0, 7.0 and 9.0 for an extended period of 120 h from 226.4 to 225.6 nm. A signi®cant 8.8-nm down
to assess the in¯uence of pH on the reaction of shift in absorbance maximum from 226.4 to
such a catalyzed hydrolysis (Fig. 5). Formation of 217.6 nm was observed when a higher concentration
2
precipitate was immediately observed when mercu- of HgCl (21.84 mg/l) was used in the reaction mix-
ric chloride was added to the pH 9.0 bu€er sol- ture. Formation of the hydrolysis product M1 from
ution. The depletion of mercuric ions in the pH 9.0 Irgarol could be gas chromatographically demon-
bu€er solution greatly increased the stability of strated in the reaction mixture containing 0.1 mg/l
Irgarol 1051 in that bu€er solution. After a 24-h HgCl . When the HgCl was increased to 1 mg/l,
2 2
contact time, all Irgarol in the pH 5.0 and 7.0 buf- approximately 50% of Irgarol 1051 in the reaction
fer solutions had been completely hydrolyzed to mixture had been hydrolyzed to M1. All the Irgarol
yield the corresponding hydrolysis product M1. 1051 had been hydrolyzed to M1 in the presence of
Only an insigni®cant amount of M1 was found in 10 mg/l HgCl
the pH 9.0 bu€er solution after a 120-h contact
2
.
The e€ectiveness of various heavy metal salts and
time (Fig. 5). The critical role of mercuric ions one organomercury compound on the hydrolysis of
involved in enhanced hydrolysis of Irgarol 1051 was Irgarol 1051 was also investigated in order to assess
further demonstrated in another experiment, in whether the observed mercuric chloride-catalyzed
which the addition of EDTA (50 mg/l) to the reac- hydrolysis of Irgarol 1051 also occurs with other
tion mixture containing HgCl
051 (5 mg/l) was found to be capable of suppres- are known to accelerate the hydrolysis of some or-
sing the hydrolysis of Irgarol 1051 by approxi- ganic phosphate pesticides in a pH range between 5
2
(10 mg/l) and Irgarol heavy metals (Table 1). Metal ions such as Cu(II)
1
Fig. 5. Time course of Irgarol 1051 and M1 concentration pro®les in bu€er solutions containing 5 mg/l
Irgarol and 20 mg/l HgCl at pH 5.0. 7.0 and 9.0.
2

Hydrolysis of the antifouling agent, Irgarol
161
Fig. 6. E€ect of HgCl
2
on the UV absorbance spectra of
Irgarol 1051. The reactants were incubated at room tem-
Fig. 7. Concentration e€ect of HgCl
2
on Irgarol degra-
dation and M1 formation. The reactants were incubated
at room temperature (218C) for 1 h before extraction with
DCM for GC analysis.
perature (218C) for 1 h before spectra taken.
and 6 where they are normally considered relatively
stable (Mortland and Raman, 1967; Miyamoto et
al., 1990). The results of Table 1 show that the four
action increases the electrophilic character of the
cyclopropyl carbon adjacent to the amino nitrogen
of the cyclopropylamino side chain and would sim-
ultaneously weaken the cyclopropyl±NH bond, thus
facilitating the cleavage of this bond by the nucleo-
philic attack of water. This reaction would lead to
the formation of two degradation products M1 and
cyclopropanol. Being more stable than the parent
compound, M1 has been frequently identi®ed in the
reaction mixtures. However, our attempts to ident-
ify cyclopropanol as a degradation product were
not successful. Cyclopropanol is an extremely un-
stable compound, primarily due to its angle strain.
Many attempts to prepare cyclopropanol resulted
only in the formation of allyl alcohol (Fieser and
Fieser, 1964).
3 2 4 2
heavy metal salts AgNO , CdCl , CuSO and PbCl
could not e€ect a hydrolysis of Irgarol 1051 in dis-
tilled water. No Irgarol disappearance and M1 for-
mation in the reaction mixtures were observed. At
the test concentration of 100 mg/l, ZnCl
2
was
capable of catalyzing about 10% hydrolysis of
Irgarol 1051. The ability of the inorganic mercury
salt HgCl
1
1
2
to promote the hydrolysis of Irgarol
051 was self-evident (Table 1). At the level of
0 mg/l, mercuric chloride was able to e€ect a com-
plete hydrolysis of Irgarol 1051 in distilled water.
However, the organic mercury compound MeHgCl
was less e€ective in catalyzing the hydrolysis of
2
+
Irgarol 1051. The cation Hg
sponsible for the enhanced hydrolysis of Irgarol
051, since the inclusion of EDTA (50 mg/l) in the
reaction mixture would substantially lower the rate
of Irgarol hydrolysis by HgCl
appears to be re-
1
Table 1. Hydrolysis of Irgarol 1051 by heavy metals
2
.
%
Irgarol hydrolyzed
20 mg/l
Based on the limited information in the present
study, a degradation pathway for the mercuric
chloride-catalyzed hydrolysis of Irgarol 1051 is pro-
posed (Fig. 8). The pathway includes the initiation
of the reaction by mercuric ion to form a four-
membered chelate complex involving the Irgarol
ring nitrogen (number 5 position) and the nitrogen
at the 6-cyclopropylamino side chain. This inter-
Metals
10 mg/l
100 mg/l
CuCl
AgNO
CdCl
PbCl2
ZnCl
HgCl
MeHgCl
2
0
0
0
0
0
0
0
0
0
0
0
0
0
3
2
0
2
10
100
62
2
100
13
100
33

1
62
D. Liu et al.
2
Fig. 8. Proposed mechanism for the HgCl -catalyzed hydrolysis of Irgarol 1051.
Although there are several reports delineating the more accurately estimate the potential impact of
hydrolysis of pesticides by metal ions (Miyamoto et these compounds, further interaction studies must
al., 1990; Munch and Frebis, 1992; Noblet et al., be undertaken. For this reason, abiotic degradation
1
996), the mechanism of metal ion-catalyzed hy- of priority chemicals is also one of the major and
drolysis of pesticides in the aquatic environment is long term studies being conducted in our labora-
poorly understood. Mortland and Raman (1967) tories. Probably one of the most important and
studied the Cu(II)-catalyzed hydrolysis of organo- practical aspects of this work is the demonstration
phosphate pesticides and proposed an elegant of rapid decomposition of Irgarol 1051 by mercuric
chemical degradation mechanism. The proposed chloride. Consequently, HgCl should never be used
2
mechanism involved the bidentating chelation of as a microbial inhibitor in preserving water samples
the metal ion Cu(II) through nitrogen in the ring in Irgarol monitoring programs. Mercuric chloride
structure and sulfur on the phosphate side chain in has been widely used as microbial inhibitor to con-
the studied pesticides Dursban and Diazinon. The trol biological degradation in environmental
resultant cyclic resonance system would weaken the samples including the US EPA National Pesticide
bonding of the side chain through electron shifts, Survey Program (Munch and Frebis, 1992).
thus promoting the hydrolysis. Pusino et al. (1988)
also proposed a similar degradation mechanism
involving the bidentate complex to explain the cata-
lytic hydrolysis of the pesticide Quinaphos by
CONCLUSIONS
Cu(II), Fe(III) and other metal ions. Hydrolysis is
Speci®c conclusions established from this study
an important abiotic degradation process for many are as follows:
pesticides in aquatic environments. Thus, under-
(1) Mercuric chloride was capable of rapidly cata-
standing the factors a€ecting the rate of hydrolysis lyzing the hydrolysis of the new antifouling com-
is essential in estimating a pesticide's environmental pound Irgarol 1051 in distilled water and bu€er
persistence. Although the studies presented here solutions.
deal with a pure herbicide tested in the laboratory
(2) The abiotic degradation appeared to follow
setting, and although it may be dicult to extrap- the reaction of a catalyzed hydrolysis, and was not
olate the results to the natural environment, the signi®cantly a€ected by the pH tested (5 to 9).
data identify a signi®cant de®ciency in information
regarding the impact of heavy metals on the en- CdCl , CuSO , PbCl and ZnCl ) had practically no
(3) All other 5 heavy metal salts tested (AgNO3,
2
4
2
2
vironmental persistence of pesticides. Heavy metal catalytic property on Irgarol hydrolysis, implying
2
+
and pesticide contaminants can be expected to be the involvement of a speci®c activity for Hg
found together in the environment and in order to this reaction.
in

Hydrolysis of the antifouling agent, Irgarol
163
Gamble D. S. and Khan S. U. (1988) Atrazine hydrolysis
in aqueous suspension of humic acid at 25.08C. Can. J.
Chem. 66, 2605±2617.
(
4) The mechanism for the catalyzed hydrolysis
may be the formation of bidentate chelation
through nitrogen (No. 5) on the ring and the nitro-
gen on the cyclopropylamino side chain in Irgarol
Langston W. J., Bryan G. W., Burt G. R. and Gibbs P.
E. (1990) Assessing the impact of tin and TBT in estu-
aries and coastal regions. Func. Biol. 4, 433±443.
Liu D., Maguire R. J., Lau Y. L., Pacepavicius G. J.,
Okamura H. and Aoyama I. (1997) Transformation of
the new antifouling compound Irgarol 1051 by
Phanerochaete chrysosporium. Water Res. 31, 2363±
2369.
2
+
1
051 with the Hg ion. The resulting four-member
chelate complex would weaken the cyclopropyl-
amino bond considerably, thus facilitating the hy-
drolysis reaction.
(
5) Ultraviolet spectroscopy of the reaction mix-
Mandelbaum R. T., Wackett L. P. and Allan D. L. (1993)
Rapid hydrolysis of atrazine to hydroxyatrazine by soil
bacteria. Environ. Sci. Technol. 27, 1943±1946.
Miyamoto J., Mikami N. and Takimoto Y. (1990) The
fate of pesticides in aquatic ecosystems. In
Environmental Fate of Pesticides, eds. D. H. Hutson and
T. R. Roberts, pp. 123±147. John Wiley and Sons,
Chichester.
Mortland M. M. and Raman K. V. (1967) Catalytic hy-
drolysis of some organic phosphate pesticides by
copper(II). J. Agric. Food Chem. 15, 163±167.
Munch D. J. and Frebis C. P. (1992) Analyte stability stu-
dies conducted during the national pesticide survey.
Environ. Sci. Technol. 26, 921±925.
tures and the identi®cation of Irgarol hydrolysis
product M1 (2-methylthio-4-tert-butylamino-6-
amino-s-triazine) by GC-MS and LC-MS provided
the basis for the proposed mechanism on the
2
HgCl -catalyzed hydrolysis of Irgarol 1051. M1
appeared to be more stable than the parent com-
pound Irgarol 1051, thus implying its possible ac-
cumulation in the environment.
(
HgCl should not be used in preserving water
6) One practical aspect of this work is that
2
samples in Irgarol 1051 monitoring programs.
Noblet J. A., Smith L. A. and Su€et I. H. (1996) In¯uence
of natural dissolved organic matter, temperature, and
mixing on the abiotic hydrolysis of triazine and organo-
phosphate pesticides. J. Agric. Food Chem. 44, 3685±
3693.
AcknowledgementsÐThis work was supported in part by
the PESTMYOP and Great Lakes 2000 funds of the
Department of Environment (Canada) and by a research
grant provided by the Ministry of Education, Science,
Sports and Culture (Japan) for the International Pusino A., Gessa C. and Kozlowski H. (1988) Catalytic
Comparative Study on Toxicity Assessment. The authors
would also like to thank Dr C. E. Cerniglia and Dr T. M.
Heinze of the National Center for Toxicological Research,
Je€erson, for the interpretation of some of the mass spec-
tral data.
hydrolysis of quinalphos on homoinonic clays. Pestic.
Sci. 24, 1±8.
Readman J. W., Kwong L. L. W., Grondin D., Bartocci
J., Villeneuve J. P. and Mee L. D. (1993) Coastal water
contamination from a triazine herbicide used in anti-
fouling paints. Environ. Sci. Technol. 27, 1940±1942.
Scarlett A., Donkin M. E., Fileman T. W. and Donkin
P. (1997) Occurrence of the marine antifouling agent
Irgarol 1051 within the Plymouth Sound locality:
Implicatioons for the green marcroalga Enteromorpha
intestinalis. Mar. Pollut. Bull. 34, 645±651.
REFERENCES
Beaumont A. R. and Newman P. B. (1986) Low levels of
tributyltin reduce growth of microalgae. Mar. Pollut.
Bull. 19, 294±296.
Ciba Geigy (1995) IRGAROL
Paints. Technical Information Bulletin. Ciba Geigy,
Basel, Switzerland, 14 pp.
Schroder H. F. (1997) Mass spectrometric detection and
identi®cation of polar pesticides and their degradation
products: a comparison of di€erent ionization methods.
Environ. Monitor. Assess. 44, 503±513.
1
1051 in Antifouling
Dahl B. and Blanck H. (1996) Toxic e€ects of the anti-
fouling agent Irgarol 1051 on periphyton communities
in coastal water microcosms. Mar. Pollut. Bull. 32, 342±
Toth S., Slooten K. K., Spack L., de Alencastro L. F. and
Tarradellas J. (1996) Irgarol 1051, an antifouling com-
pound in freshwater, sediment and biota of Lake
Geneva. Bull. Environ. Contam. Toxicol. 57, 423±426.
3
50.
Evans S. M., Leksono T. and McKinnell P. D. (1995) Voyksner R. D., McFadden W. H. and Lammert S. A.
Tributyltin pollution: A diminishing problem following
legislation limiting the use of TBT-based anti-fouling
paints. Mar. Pollut. Bull. 30, 14±21.
Fieser L. F. and Fieser M. (1964) Advanced Organic
Chemistry. McGraw-Hill, New York, 1158 pp.
(1987) Application of thermospray HPLC/MS/MS for
determination of triazine herbicides. In Applications of
New Mass Spectrometry Techniques in Pesticide
Chemistry, ed. J. D. Rosen, pp. 247±258. John Wiley
and Sons, New York.