Study of the characteristics of three high-energy intermediates generated in peroxyoxalate chemiluminescence (PO-CL) reactions

Article abstract of DOI:10.1039/b109210c

Perylene emission intensity generated from peroxyoxalate chemiluminescence (PO-CL) reactions was studied as a function of time and order of reagent addition. Based on 1H-NMR analyses, kinetics of UV absorbance and emission intensity vs. time profiles, we

Full text of DOI:10.1039/b109210c

View Article Online / Journal Homepage / Table of Contents for this issue
Study of the characteristics of three high-energy intermediates
generated in peroxyoxalate chemiluminescence (PO-CL) reactions
Ji Hoon Lee,*†a James C. Rock,a Seung Bum Park,‡b Mark A. Schlautmanc,d
and Elizabeth R. Carrawayd
2
a Occupational Health & Safety Institute, Texas A&M University, College Station,
TX 77843-3133, USA. E-mail: lee@clemson.edu
b Department of Chemistry, Texas A&M University, College Station, TX 77842-3012, USA
c Department of Agricultural & Biological Engineering, Clemson University, Clemson,
SC 29634-0357, USA
d Department of Environmental Toxicology and the Clemson Institute of Environmental
Toxicology, Clemson University, Pendleton, SC 29670, USA
Received (in Cambridge, UK) 10th October 2001, Accepted 13th February 2002
First published as an Advance Article on the web 4th March 2002
Perylene emission intensity generated from peroxyoxalate chemiluminescence (PO-CL) reactions was studied
as a function of time and order of reagent addition. Based on 1H-NMR analyses, kinetics of UV absorbance
and emission intensity vs. time profiles, we conclude that PO-CL reactions in the presence of imidazole (ImH)
can proceed by three distinct reaction pathways depending on how the reagents are mixed together. When
bis(2,4,6-trichlorophenyl) oxalate (TCPO) is mixed simultaneously with H2O2, ImH and perylene, a slowly
decaying emission curve is generated from the interaction between perylene and a high-energy intermediate
(i.e., six- or eight-membered cyclic compound) formed by the ImH-catalyzed nucleophilic reaction (TCPO-CL
reaction). Upon mixing TCPO simultaneously with ImH and perylene in the absence of H2O2, however, distinctly
different CL curves of lower intensity are generated from the interaction between perylene and a new, unknown
high-energy intermediate formed from the reaction between the aryl oxalate and ImH. Finally, using 1H-NMR,
we observed that 1,1Ј-oxalyldiimidazole (ODI) is also formed from the reaction between TCPO and ImH. When
ODI reacts with excess H2O2 in the presence of perylene, a higher intensity and relatively fast decaying emission
curve is generated (ODI-CL reaction) from the interaction between perylene and the high-energy intermediate
produced, which we propose is imidazolylhydroxydioxetanone or 1,2-dioxetanedione.
(1,2-dioxetanedione) producing the fast decay curve resulted
Introduction
from the acid–base reaction between the base catalyst and the
hydroperoxyoxalate ester generated from the reaction between
DNPO and H2O2. They also suggested that the intermediate
(six-membered cyclic intermediate) generating the slow decay
curve was formed by the ImH-catalyzed reaction proposed by
Milofsky and Birks.11 Both decay curves observed under their
experimental conditions13 were dependent on the concentra-
tions of base catalyst and DNPO. However, in the CL intensity
vs. time profile obtained from the bis(2,4,6-trichlorophenyl)
oxalate (TCPO) CL reaction using a higher ImH concentration
than that in the DNPO-CL reaction, only a slow decaying curve
appeared. This is because the hydroperoxyoxalate ester gener-
ated from the reaction between TCPO and H2O2 has a higher
pKa than ImH.13
Peroxyoxalate chemiluminescence (PO-CL) generated from the
reaction between aryl oxalates and H2O2 in the presence of a
catalyst has been widely used to analyze various trace fluoro-
phores because it is more sensitive than fluorescence and UV
absorbance.1–5
Since Rauhut et al. first proposed 1,2-dioxetanedione as a
high-energy intermediate capable of transferring energy to
fluorophores in PO-CL reactions,6 several authors have pro-
posed structures and properties of other possible high-energy
intermediates generated in these reactions.7–20 Alvarez et al.8
proposed that at least two high-energy intermediates are gener-
ated in PO-CL reactions, and several research groups9–20 have
subsequently proposed possible reaction pathways for these
high-energy intermediates. These researchers generally used
imidazole (ImH) to catalyze the PO-CL reaction because it is
an efficient catalyst.21
Neuvonen14 suggested the formation of 1,1Ј-oxalyldiimid-
azole (ODI) as one of the intermediates formed in PO-CL reac-
tions in the presence of ImH based on the ImH-catalyzed
hydrolysis of bis(4-nitrophenyl) oxalate (4-NPO) and the
observation of ImH-catalyzed hydrolysis of p-nitrophenyl acet-
ate.22 Hadd and Birks17 proposed a nucleophilic substitution
reaction of ImH with TCPO or DNPO from the UV absorb-
ance of 2,4,6-trichlorophenol (TCP) or 2,4-dinitrophenol
(DNP) produced in the reactions. They19 also studied the ImH-
catalyzed PO-CL reaction under both reaction conditions
([ImH], [H2O2] ӷ [TCPO] and [ImH] ӷ [TCPO] ӷ [H2O2]) in
the presence of 9,10-diphenylanthracene (DPA). They19 pro-
posed that ODI is formed from the reaction between TCPO and
ImH in the presence of H2O2. Recently, Hadd et al.23 proposed
Lee et al.13 obtained both a fast decay curve and a slow decay
curve from bis(2,4-dinitrophenyl) oxalate (DNPO) CL reac-
tions in the presence of base catalysts (ImH or sodium
salicylate). They proposed that the high-energy intermediate
† Present address: Department of Environmental Toxicology and the
Clemson Institute of Environmental Toxicology, Clemson University,
PO Box 709, 509 Westinghouse Road, Pendleton, SC 29670, USA.
‡ Present address: Howard Hughes Medical Institute (HHMI),
Department of Chemistry and Chemical Biology, and Harvard
Institute of Chemistry and Cell Biology (ICCB), Harvard University,
12 Oxford Street, Cambridge, Massachusetts 02138, USA.
802
J. Chem. Soc., Perkin Trans. 2, 2002, 802–809
This journal is © The Royal Society of Chemistry 2002
DOI: 10.1039/b109210c

View Article Online
two possible types of intermediates formed in ODI-CL reac-
tions based on the results observed from the reactions between
ODI and various H2O2 concentrations in the presence of DPA.
Previous research groups11–13,15,16,18–20,23 have studied ImH-
catalyzed PO-CL reactions when the reagents were mixed
simultaneously. As shown in Scheme 1, the base-catalyzed
system. Perdeuterated DMSO, acetonitrile, benzene, acetone,
and chloroform were purchased from Cambridge Isotope
Laboratories.
2.2 Chemiluminescence measurements
A typical CL reaction was conducted at room temperature (∼22
to 23 ЊC) in a 1.0 cm fluorescence cell inserted in the sample
compartment of a spectrofluorometer (PTI Inc.). Fresh sol-
utions were prepared daily and kept in the dark. The CL inten-
sity vs. time was monitored by the spectrofluorometer at the
maximum emission wavelength (468 nm) of perylene without
an illuminating light source. A small magnetic stir bar in the
fluorescence cell provided continuous mixing during each
experiment.
2.2.1 Simultaneous mixing of PO-CL reagents. A 0.5-ml
H2O2 solution was added to the cell followed by addition of
0.5 ml each of a perylene solution and an ImH solution. The
reaction was initiated by injecting 0.5 ml of an aryl oxalate
(DNPO, TCPO or PCPO) solution, and each experimental
condition was replicated three times.
The rate of TCP formation from the TCPO-CL reaction was
followed by absorbance at 290 nm using the kinetics function of
an 8452A diode array spectrophotometer (Hewlett Packard).
2.2.2 Mixing of perylene with ODI formed from the TCPO–
ImH reaction in the absence and presence of H2O2. After mixing
0.5 ml of an aryl oxalate solution with 0.5 ml ImH solution in
the cell for a specified time period, the CL reaction was initiated
by injecting 1.0 ml of a perylene solution or a mixed perylene–
H2O2 solution into the cell. Each experimental condition was
replicated three times.
TCP generation from the reaction between TCPO and ImH
in the absence of H2O2 was measured as described above.
Scheme 1
2.3 NMR analyses
nucleophilic reaction proposed by two research groups11,13 and
the ODI-CL reaction suggested by other authors12,15,16,18–20,23
have been used to explain observed CL decay curves. X and
Y in Scheme 1 indicate high-energy intermediates proposed by
previous research groups.11–13,15,16,18–20,23
Samples of ODI were dissolved in d6-DMSO, while TCPO,
TCP and ImH were dissolved in d3-acetonitrile or d6-DMSO.
1H-NMR spectra of ODI, TCPO, TCP, ImH, and solution mix-
tures of TCPO and ImH were recorded in flame sealed NMR
tubes at ambient temperature with a UnityPlus 300 NMR
Spectrometer. Spectra were obtained for various mixing times
to determine if one of the products generated from the reaction
between TCPO and ImH in DMSO or acetonitrile was ODI.
The motivation for this study was to gain additional inform-
ation on the complex mechanisms and pathways of PO-CL
reactions, particularly the characteristics of the high-energy
intermediates generated by the reactions. In this research, we
varied the addition of reagents to elucidate the properties of
possible high-energy intermediate(s) formed in the presence of
ImH. First, we mixed TCPO and the other reagents (ImH,
H2O2 and perylene) simultaneously. Second, we mixed TCPO
with ImH and allowed the mixture to react for a specified time
before adding the mixture to a solution containing only peryl-
ene or H2O2 and perylene. The CL decay curve properties
observed for the different reagent addition methods were
distinctly different from one another. Based on these results,
we propose possible reaction pathways for the three different
systems.
2.4 Kinetics analysis
CL intensity vs. time profiles obtained from the TCPO-CL and
ODI-CL reactions were found to fit a simple two-step reaction
scheme:
(1)
where A and C represent reactants and products, respectively,
and CL intensity is proportional to the concentration of inter-
mediate species, B. Both steps in eqn. (1) are assumed to be
irreversible, pseudo first-order reactions with rate constants kr
and kf corresponding to the rise and fall, respectively, in the
observed CL intensity vs. time profiles. The integrated rate
equation then gives
2
Experimental
2.1 Chemicals
DNPO, TCPO, pentachlorophenyl oxalate (PCPO), TCP, H2O2
(50%), ODI (90%), perylene, and ImH were purchased from
Aldrich. H2O2 was stored in a refrigerator at Ϫ10 ЊC and
titrated with KMnO4 to determine the exact concentration
before use. Spectrophotometric grade ethyl acetate, acetonitrile,
and dimethyl sulfoxide (DMSO) purchased from Baker were
used as solvents. Deionized water of resistivity greater than
17.8 MΩ cmϪ1 was obtained from a Barnstead Nanopure
(2)
for the CL intensity vs. time.24 Values for kr, kf and the max-
imum intensity, Imax, for each CL time profile were obtained by
nonlinear least-squares fitting of eqn. (2) to the experimental
J. Chem. Soc., Perkin Trans. 2, 2002, 802–809
803

View Article Online
data using the optimization toolbox in Matlab 5.2. All reported
rate constants are an average of at least three independent
measurements.
3
Results and discussion
The CL intensity vs. time profile for simultaneous mixing of
reagents ([TCPO] = 0.25 mM, [ImH] = 0.50 mM, [H2O2] =
50.0 mM, and [perylene] = 0.04 mM) in acetonitrile is shown in
Fig. 1 (curve a). This reaction, which generated a relatively slow
Fig. 2 UV absorbance (290 nm) of TCP vs. time for the reaction
between 0.05 mM TCPO and 2.4 mM ImH in ethyl acetate in the
presence (56.0 mM, ᭹) and absence (ꢀ) of H2O2.
TCPO-CL reaction in the presence of excess H2O2 (0.050
0.003 sϪ1) is faster than that by the ODI-CL reaction (0.0060
0.0008 sϪ1). In other words, the reaction between TCPO and
H2O2 in the presence of ImH (TCPO-CL reaction) is faster than
that between TCPO and ImH (ODI-CL reaction).
1H-NMR spectra of ODI dissolved in d6-DMSO revealed
chemical shifts at 7.19, 7.78 and 8.43 ppm, whereas the chem-
ical shifts for ImH and TCPO in d6-DMSO were 7.00, 7.59 and
11.70 ppm, and 7.42 and 7.72 ppm, respectively. Although we
could not detect the chemical shifts for ODI from the reaction
of TCPO and ImH in the presence of H2O2 (TCPO-CL reac-
tion), the chemical shifts for ODI from the reaction of TCPO
and ImH without H2O2 (ODI-CL reaction) were observed.
These results indicate that the high-energy intermediate formed
from the TCPO-CL reaction differs from that produced in the
ODI-CL reaction.
Fig. 1 CL intensity vs. time profiles in acetonitrile obtained under
various experimental conditions (see text for details). (a) TCPO-CL
reaction. (b) Addition of H2O2 and perylene after TCPO reaction with
ImH for 90 seconds. (c) ODI-CL reaction. (d) Addition of H2O2 and
perylene after mixing TCPO with ODI (∼10% ImH impurity) for 60
seconds.
CL decay curve, will be referred to here as the TCPO-CL reac-
tion. Fig. 1-b shows the curve obtained when mixing TCPO
with ImH for 90 seconds before adding H2O2 and perylene; note
that the reagent concentrations used for curve b were the same
as those used for curve a. Curve b exhibits a biexponential
decay profile having both a fast decaying component and a
slowly decaying component.
3.2 Effect of mixing time between TCPO and ImH
Fig. 3 shows the effect of reaction time between TCPO
Fig. 1-c shows the CL intensity vs. time profile obtained for
the ODI-CL reaction ([ODI] = 0.25 mM, [H2O2] = 50.0 mM,
[perylene] = 0.04 mM). Several authors12,14–20 have suggested
that ODI would be generated from the reaction between ImH
and TCPO. For this study, we used ODI purchased from
Aldrich, which had a manufacturer-stated 10% (wt) impurity
consisting primarily of ImH. Thus, upon mixing the ODI with
TCPO (0.1 mM) for 60 seconds before adding H2O2 and peryl-
ene for the same ODI concentration used in Fig. 1-c, we
obtained a biexponential decay curve (Fig. 1-d) having both a
fast decaying component and a slowly decaying component,
similar to that previously observed for curve b. Based on these
results, it is reasonable to believe that the slow decay com-
ponent in curves b and d is produced by the TCPO-CL reaction
(e.g., curve a) whereas the fast decay component of b and d is
generated by the ODI-CL reaction (e.g., curve c).
The maximum intensity, Imax, of curve d is 3.1 times larger
than that of curve c. Upon mixing ODI with TCPO for
15 minutes under the otherwise same experimental conditions
used for curve c, we obtained an Imax that was 16 times larger
than that for curve c (data not shown). These results indicate
that ODI is relatively unstable in acetonitrile,25 with most of it
decomposing to ImH before it can react with H2O2. Thus, the
CL curve observed in Fig. 1-c resulted from the reaction
between H2O2 and the remaining fraction of ODI.
Fig. 3 CL intensity vs. time profiles obtained when TCPO was reacted
with ImH for specified periods of time prior to addition of H2O2 and
perylene in ethyl acetate solutions.
(0.05 mM) and ImH (2.4 mM) before adding H2O2 (56.0 mM)
and perylene (0.05 mM) in ethyl acetate solutions. When all
reagents were mixed simultaneously, a CL curve having a very
long lifetime was observed. However, when the reaction time
between TCPO and ImH prior to addition of H2O2 and peryl-
ene was increased, Imax of the CL curves initially increased
3.1 UV and NMR analyses of the TCPO-CL and ODI-CL
reactions
Fig. 2 shows TCP formation vs. time as measured by UV
absorbance. The rate constant for TCP formation by the
804
J. Chem. Soc., Perkin Trans. 2, 2002, 802–809

View Article Online
Table 1 Effect of ImH concentration in TCPO-CL reaction
Table 3 Effect of water content (% volume) in TCPO-CL reaction
[ImH]/mM
Imax
τmax/s
τhalf/s
103kr/sϪ1
103kf/sϪ1
[H2O] (%)
Imax
τmax/s
τhalf/s
0.6
0.9
1.2
1.8
2.4
2.6
2.9
3.1
3.5
3.7
75.0
54.0
41.0
25.0
16.0
451
282
194
122
85
44.2 (0.6)
56.1 (0.3)
70.3 (0.4)
113.4 (0.3)
182.4 (0.4)
1.9 (0.04)
3.3 (0.01)
5.0 (0.02)
7.9 (0.03)
11.0 (0.03)
0.0
4.0
8.0
12.0
20.0
0.22
0.43
0.48
0.52
0.39
52.0
39.0
31.0
24.0
18.0
298.0
205.0
187.0
134.0
85.0
a [TCPO] = 0.05 mM, [H2O2] = 1.80 mM, [ImH] = 1.0 mM, [perylene] =
0.05 mM, solvent: ethyl acetate.
Table 2 Effect of H2O2 concentration in TCPO-CL reaction
[H2O2]/mM
Imax
τmax/s
τhalf/s
10kr/sϪ1
103kf/sϪ1
phenol, pentachlorophenol, DNP) which are final products in
PO-CL systems on the TCPO-CL reaction and observed that
Imax, τmax and τhalf each showed a dependence on the pKa of the
phenols (data not shown).
1.8
7.2
28.0
56.0
0.51
0.59
0.71
0.78
18.0
17.0
16.0
14.5
102
98
91
1.82 (0.05)
1.90 (0.05)
2.12 (0.03)
2.39 (0.06)
7.98 (0.05)
8.21 (0.04)
9.02 (0.03)
10.31 (0.04)
84
3.3.4 Possible reaction mechanism. When the TCPO-CL
reagents are mixed simultaneously, the first intermediate most
likely to appear would be the hydroperoxyoxalate ester (ArCO-
COOOH) because the reaction rate between TCPO and H2O2 is
faster than that between TCPO and ImH (Fig. 2, Tables 1 and
2). Stevani et al.15 suggested that 4-chlorophenyl hydrogen
oxalate, one of the hydroperoxyoxalate esters, is stable in the
presence of fluorescent compounds before adding weak bases
such as ImH and p-chlorophenolate. Thus, 2,4,6-trichloro-
phenyl hydrogen oxalate, another hydroperoxyoxalate ester,
might not be a high-energy intermediate capable of generating
excited singlet states of fluorescent molecules.
for mixing times up to six minutes and then decreased exponen-
tially with longer mixing times.
Because the ImH concentration was in excess over TCPO for
the CL curves shown in Fig. 3, the slowly decaying component
was not observed and thus did not complicate the analyses. It
can be seen that as the reaction time between TCPO and
ImH increased, the time to reach Imax (τmax) and the half-life of
the decay reaction, τhalf, each decreased for the CL curves. In
the following sections, we report the effects of ImH, TCP
and water, and discuss the differences between the TCPO-CL
and ODI-CL reactions.
Alvarez et al.8 observed two peaks having different decay
times in TCPO-CL reactions using triethylamine (pKa 10.7 in
water) as a weak base. Lee et al.13 have studied the possible
reaction mechanism for two CL decay curves generated from
DNPO-CL reactions in the presence of various weak bases
such as ImH and sodium salicylate, and have proposed that the
fast decay curve observed by Alvarez et al.8 is generated from
the acid–base reaction while the slow decay curve is produced
from the base-catalyzed nucleophilic reaction proposed by
Milofsky and Birks.11
As shown in Fig. 1-a, when ImH (pKa 6.95 in water) was used
as a weak base instead of triethylamine, we did not observe a
fast decay curve generated from the acid–base reaction because
the pKa of 2,4,6-trichlorophenyl hydrogen oxalate is higher
than that of ImH.13 In conclusion, the slow decay curve
observed in this system is generated from the interaction
between perylene and a high-energy intermediate (i.e., an
eight- or six-membered cyclic intermediate) formed by the
ImH-catalyzed nucleophilic reaction.
3.3 TCPO-CL reaction
3.3.1 Effects of ImH concentration. The CL intensity vs.
time profile exhibiting a slowly decaying component (Fig. 1-a)
was obtained by simultaneous mixing of the TCPO-CL
reagents. Using this same experimental procedure, we deter-
mined the effect of ImH concentration for TCPO, H2O2 and
perylene concentrations of 0.05, 56.0, and 0.05 mM, respect-
ively, in ethyl acetate solutions (Table 1). As can be seen,
Imax increased whereas τmax and τhalf both decreased with
increasing ImH concentration. In addition, the two pseudo
first-order rate constants kr and kf exhibited second- and first-
order dependences, respectively, on the concentration of ImH.
3.3.2 Effect of H2O2 concentration. Table 2 shows the effect
of varying H2O2 concentration on Imax, τmax and τhalf for ImH,
perylene and TCPO concentrations of 3.0, 0.01 and 0.06 mM,
respectively. As can be seen, Imax increased while τmax and τhalf
both decreased slightly with increasing H2O2 concentration. In
addition, values for kr and kf also showed a slight increase with
increasing H2O2 concentration.
3.4 Chemiluminescence generated from reactions between aryl
oxalates and ImH in the absence of H2O2
3.3.3 Effect of water content. Several research groups26–28
have observed that DNPO undergoes hydrolysis in solvents
with a low water content, forming 2 moles of the corresponding
DNP from 1 mole of DNPO. Based on this information, we
hypothesized that TCPO might also undergo decomposition to
TCP and carbon monoxide in solvents containing water even
though TCPO is much more stable than DNPO.
As shown in Table 3, τmax and τhalf decreased with increasing
water content in the TCPO-CL system, while Imax first increased
and then at a certain water content also began to decrease. We
also investigated the effect of TCP (0.005 mM) for the same
reaction conditions as Table 3 in the absence of water (data not
shown). Upon adding TCP, Imax increased while τmax and τhalf
decreased slightly. Based on these results, we propose that the
TCP generated from TCPO hydrolysis acts as a catalyst or
nucleophile to enhance Imax and reduce τmax and τhalf below a
certain concentration but then acts as an inhibitor in the
TCPO-CL reaction above this concentration. We also studied
the effects of adding various other phenols (e.g., 2,4-dichloro-
We observed distinctly different CL intensity vs. time profiles
when different aryl oxalates were mixed simultaneously with
ImH and perylene in the absence of H2O2. Fig. 4 shows that the
reactivity of the intermediate formed by this reaction depends
on the properties of the substituted leaving group in a particu-
lar aryl oxalate. When perylene was mixed directly with TCPO
or ODI in the absence of H2O2, only noise was detected.
Undoubtedly, the profiles shown in Fig. 4 are generated from
the interaction between perylene and a new high-energy inter-
mediate formed from the reaction between aryl oxalates and
ImH. Fig. 4 implies that the nucleophilic reaction of ester with
ImH is more complicated than previously known.
1
3.4.1 NMR analysis. Fig. 5 shows H-NMR spectra for
TCPO and ImH in acetonitrile for different mixing times.
Although five specific chemical shifts (7.19, 7.72, 7.80, 8.40 and
8.53 ppm) appeared between 2 and 10 min of mixing time, only
three (7.19, 7.72 and 8.40 ppm) of them, all consistent with
ODI, were observed at 30 min. The chemical shift for ODI
J. Chem. Soc., Perkin Trans. 2, 2002, 802–809
805

View Article Online
Table 4 Effects of ImH concentration on kr and kf
[ImH]/mM
10kr/sϪ1
102kf/sϪ1
0.50
0.75
1.00
1.50
2.00
1.01 (0.08)
1.54 (0.06)
2.33 (0.15)
3.53 (0.09)
4.52 (0.12)
0.05 (0.006)
0.12 (0.012)
0.17 (0.021)
0.42 (0.023)
0.65 (0.019)
a [DNPO] = 0.1 mM, [perylene] = 0.05 mM, solvent: ethyl acetate.
Fig. 4 CL generated from the reaction between ImH and three types
of aryl oxalates in the presence of perylene without added H2O2;
[oxalate] = 0.2 mM, [ImH] = 2.0 mM, [perylene] = 0.05 mM.
Scheme 2
Fig. 5 1H-NMR spectra obtained after mixing TCPO with ImH for
a specified time in acetonitrile.
here with an undefined structure, is generated from 1. We pro-
pose that Z is very unstable and has never been isolated, again
as often is the case for other high-energy intermediates formed
in PO-CL reactions.
observed at 7.19 ppm was higher than the other two chemical
shifts (7.72 and 8.40 ppm) between 2 and 10 min of mixing
time, although the height ratio of the chemical shifts after
30 min was constant. This result indicates that the chemical
shifts of the transient intermediate formed from the reaction
between TCPO and ImH before ODI was produced not only
appeared at 7.80 and 8.53 ppm but also at 7.19 ppm (i.e., the
chemical shift of the transient intermediate overlapped with
that of ODI at 7.19 ppm).
Rising rate constants, kr, and falling rate constants, kf, were
obtained from the pseudo-first-order reaction8 (Table 4). A
first-order dependence of kr on the ImH concentration was
observed, whereas a second-order dependence resulted for kf.
kr, necessary to form 1 in the reaction between DNPO and
ImH, illustrates the first-order dependence on ImH concen-
tration because the reaction of the tetrahedral intermediate that
forms 1, k2[ImH], is faster than the reverse reaction, kϪ1. kf,
necessary to form Z and 2 in the reaction between 1 and ImH,
shows a second-order dependence on ImH concentration. This
is because the pathway for forming 2 is predominant compared
to that for generating Z. Also, the base-catalyzed breakdown of
the zwitterionic addition intermediate is the rate-limiting step
of this reaction in organic solvents (kϪ3 ӷ k4 [ImH]).14,17,22,30 kf is
approximately equal to k4Ka[DNPO][ImH],2 where Ka = k3/kϪ3.
3.4.2 Possible mechanism. Neuvonen14 obtained UV spec-
tra of a transient intermediate for the reaction of ImH with
bis(4-nitrophenyl) oxalate and reported that the decompos-
ition rate of the transient intermediate depended on the ImH
concentration. Thus, the three chemical shifts (7.19, 7.80 and
8.53 ppm) observed in Fig. 5 could conceivably be the oxalyl
derivative (ImCOCO2Ar) proposed by Neuvonen.22 However,
the oxalyl derivative cannot explain the results shown in Fig. 4
because it does not generate singlet excited states in fluoro-
phores by the chemically initiated electron exchange lumin-
escence (CIEEL) mechanism.29 Thus, we conclude that a new
high-energy intermediate formed from the reaction between
TCPO and ImH in the absence of H2O2 was not detected by the
1H-NMR spectrometer, most likely because of its very short
lifetime, which would be consistent with other high-energy
intermediates formed in PO-CL reactions. Figs. 4 and 5
together indicate that the new high-energy intermediate can be
formed from an oxalyl derivative before producing ODI, as
demonstrated in Scheme 2. Unfortunately, we were unable to
determine the identity of this new high-energy intermediate
based on our results. The high-energy intermediate, Z, shown
3.5 ODI-CL reaction
As shown in Fig. 1, the CL decay curve obtained from the
ODI-CL reaction is faster than that observed in the TCPO-CL
reaction. The CL decay curve observed in the ODI-CL reac-
tion is generated by the interaction between perylene and a
high-energy intermediate formed from ODI and H2O2.
3.5.1 Effect of ImH concentration. We observed CL inten-
sity vs. time profiles for varying ImH concentration (0.6 to
2.4 mM) using the method of Fig. 3 to understand the details
of the reaction between TCPO and ImH that forms ODI. Imax
shown in Fig. 6 indicates the maximum of the CL intensity vs.
806
J. Chem. Soc., Perkin Trans. 2, 2002, 802–809

View Article Online
Table 5 Effect of ImH concentration on the reactions between TCPO
and ImH before adding H2O2 and perylene in ODI-CL reactions
Table 6 Effect of water content (% volume) in ODI-CL reactions
[H2O] (%)
0
8.0
12.0
16.0
20.0
[ImH]/mM
Imax
τmax/s
τhalf/s
103kr/sϪ1
103kf/sϪ1
Relative Imax
STD
0.64
0.03
2.26
0.12
3.99
0.21
1.42
0.15
0.43
0.05
0.6
0.9
1.2
1.8
2.4
18.9
31.0
25.3
19.7
15.6
∼1800
∼1260
∼840
∼540
∼360
∼5220
∼3420
∼2400
∼1740
∼1140
0.38 (0.03)
0.72 (0.04)
1.30 (0.04)
2.77 (0.05)
5.33 (0.03)
0.82 (0.03)
0.87 (0.04)
0.95 (0.03)
1.05 (0.03)
1.17 (0.03)
a Reaction conditions: [TCPO] = 0.05 mM, [ImH] = 0.5 mM, [H2O2] =
20.0 mM, [perylene] = 0.03 mM, TCPO mixed with ImH for 120 s
before adding H2O2 and perylene.
a Reaction conditions are shown in Fig. 6.
Fig. 7 Effect of H2O2 concentration on the reactions between TCPO
and ImH before adding H2O2 in ODI-CL reactions at 22–23 ЊC.
[TCPO] = 0.03 mM, [ImH] = 5.0 mM, [perylene] = 0.05 mM.
Fig. 6 Effect of ImH concentration on the reaction between TCPO
and ImH before adding H2O2 and perylene in ODI-CL reactions
in ethyl acetate. [H2O2] = 56.0 mM, [TCPO] = 0.05 mM, [perylene] =
0.05 mM.
first increased and then began to decrease with increasing H2O2
concentration because ODI formed by the reaction between
TCPO and ImH reacts with H2O2 to produce a high-energy
time profile observed for each experimental condition. As can
be seen, Imax increased up to a certain point and then decreased
exponentially with an increase in the mixing time for the reac-
tion between TCPO (0.05 mM) and each ImH concentration
before adding perylene (0.05 mM) and H2O2 (56.0 mM) in ethyl
acetate. τmax and τhalf for the data presented in Fig. 6 decreased
as the ImH concentration increased. This is because the excess
ImH used in the ODI-CL reaction catalyzes the breakdown
of ODI formed from the reaction between TCPO and ImH.
Neuvonen14 and Stigbrand et al.25 have suggested that the
decomposition rate of ODI is dependent on the concentration
of ImH.
intermediate. However, kr (0.031
0.003 sϪ1), kf (0.0015
0.0002 sϪ1), τmax, and τhalf for the profiles shown in Fig. 7 were
independent of H2O2 concentration under these experimental
conditions because H2O2 reacts with ODI.
3.5.3 Effect of water content. Table 6 shows the effect of
water content in the ODI-CL reaction solution. As the water
content increased, Imax initially increased and then began to
decrease above 16%. In particular, Imax obtained in the solvent
containing 12% water was approximately 6.25 times higher than
that observed in acetonitrile alone. The decay time of the CL
intensity vs. time profile measured with water present was faster
than that observed without water, a result consistent with that
reported by Stigbrand et al.25 Under our experimental con-
ditions, the reaction rate between TCPO and ImH in aqueous
solvent would be faster than that in a purely nonaqueous sol-
vent because water is widely known to accelerate PO-CL reac-
tions.6,31,32 In other words, the reaction kinetics between TCPO
and ImH depends on the concentration of water in ODI-CL
reaction systems. The concentration of ODI generated by the
120-s reaction between TCPO and ImH in the organic solvent
containing 12% water is likely the highest in our experiments.
Thus, when TCPO reacts with ImH for 120 s in an aqueous
solvent containing 16 or 20% water, the ODI formed is under-
going a fast decomposition during that reaction time. If this
hypothesis is correct, the reaction time between TCPO and
ImH would have to be reduced to measure higher Imax values for
the ODI-CL system using a solvent composed of 20% : 80%
(v/v) water and acetonitrile.
Several research groups14,16,17 have obtained rate constants
for TCP production from the reaction between TCPO and ImH
using a pseudo-first-order reaction method and then used UV
absorbance of TCP to predict indirectly the rate constants of
ODI formation in their systems. With the CL intensity vs. time
profiles shown in Fig. 6, we were able to directly determine the
rate constants of ODI production from the reaction between
TCPO and ImH (Table 5). We are unaware of any prior use of
this method. The upward curvature of the trend for kr obtained
from this new kinetics method indicates a second-order
dependence on the ImH concentration, a result consistent with
those obtained by Neuvonen14 and Hadd and Birks.17 The fall
rate constant, kf, exhibits a first-order dependence on the ImH
concentration. We interpret this result to mean that the decom-
position of ODI is accelerated with increasing ImH concen-
tration, an interpretation consistent with results observed by
Stigbrand.25
3.5.2 Effect of H2O2 concentration. Fig. 7 shows the effect of
H2O2 concentration on Imax determined when H2O2 and peryl-
ene were added to the solution after premixing TCPO and ImH
in ethyl acetate for various times similarly to Figs. 3 and 6. Imax
The results shown in Table 6 likely are observed because of
the formation of TCP from TCPO hydrolysis. To determine the
effect of TCP concentration in the ODI-CL reaction, we added
TCP (0.01 to 0.05 mM) to the mixture of TCPO (0.5 mM) and
J. Chem. Soc., Perkin Trans. 2, 2002, 802–809
807

View Article Online
ImH (5.0 mM) in dried acetonitrile and waited 2 minutes before
adding H2O2 (20.0 mM) and perylene (0.05 mM). Imax observed
under these reaction conditions was equal to or slightly lower
than those obtained for ODI-CL reactions without added TCP
We conclude that Imax obtained from the ODI-CL reaction is
not enhanced by adding TCP even though Imax observed from
the TCPO-CL reaction is extremely dependent on TCP
concentration.
over a wide range of concentrations. To determine the effect of
H2O2 concentration in detail, H2O2 and perylene were added to
solutions after TCPO had reacted with excess ImH in ethyl
acetate for 240 s. The slowly decaying CL curve disappeared.
However, we were not able to show an effect of H2O2 con-
centration and are not yet able to propose a high-energy inter-
mediate formed by ODI-CL reaction. This is because the τmax
and τhalf of the CL decay curves were too short to calculate the
rate constants (kr and kf) for each peak by the presumed
pseudo-first-order reaction. The high-energy intermediate gen-
erated by the ODI-CL reaction might come from imidazolyl
peracid (ImCOCOOOH), analogous to hydroperoxyoxalate
ester (ArCOCOOOH) formed in the first reaction step of
mechanisms proposed for PO-CL reactions. We showed that
Imax, τmax and τhalf in ODI-CL reactions decrease with increasing
H2O2 concentration in Fig. 8. This is because H2O2 substitution
for ImH of imidazolyl peracid to form dihydroperoxyoxalate
(HOOCCOCOOOH), known to be a low-energy intermediate,
occurs competitively with the reaction to form high-energy
intermediate(s) from imidazolyl peracid when ODI reacts with
excess H2O2. Based on the results obtained, the most plausible
structures of high energy intermediate(s) formed from the
intramolecular cyclization of imidazolyl peracid may be imid-
azolylhydroxydioxetanone or 1,2-dioxetanedione because of
the relatively excess H2O2 added in our system. As Hadd et al.23
suggested recently, however, when H2O2 concentrations much
lower than those of TCPO and ImH are added in the ODI-CL
reaction system, the possible structure of the high-energy
intermediate (ImH-substituted six-membered cyclic intermedi-
ate) may be similar to that suggested for the TCPO-CL reaction
in Section 3.3 of this paper.
3.5.4 Possible reaction mechanism. Using 1H-NMR spectro-
scopy, we confirmed the formation of ODI by the reaction
between TCPO and ImH in the absence of H2O2. As shown in
Table 5, when TCPO reacts with excess ImH to form ODI the
upward trend for kr indicates a second-order dependence on the
ImH concentration, consistent with the results reported for
reactions of esters with ImH.27
Orosz33 investigated the hydrolysis of various diaryl oxalates
using organic solvent–water CL systems and gas chrom-
atography. Neuvonen14 also measured rate constants for the
hydrolysis of 4-NPO using UV absorbance. Both researchers
postulated that the release of the first phenol group for each
oxalate is much faster than that of the second one. Neuvonen14
also proposed that the reaction step leading to the formation of
the oxalyl derivative by the reaction between NPO and ImH is
fast compared with the second reaction step to form ODI.
However, results obtained by other research groups16,17 for the
reaction between TCPO and ImH in the absence of H2O2 are in
contrast with those of Neuvonen.14 For example, Hadd and
Birks17 have suggested that substitution of ImH for one TCP in
TCPO will be slower than that of ImH for the second substi-
tution because of the order of activation by the acyl groups
toward substitution on the adjacent carbonyl (Im(C᎐O) >
᎐
TCP(C᎐O)). We also discussed the detailed mechanism to form
᎐
4
Conclusions
ODI for the reaction between TCPO and ImH in the previous
section of this paper. Based on the references and our results
obtained in Section 3.4, kr values shown in Table 5 are related to
the nucleophilic reaction of ImH toward TCPO molecules.
ODI formed from the reaction between TCPO and ImH in
the absence of H2O2 subsequently reacts with added H2O2 to
produce a high-energy intermediate capable of generating
excited singlet states of perylene. We investigated the effect of
H2O2 concentration added to the solution after the reaction
between TCPO and ImH in ethyl acetate for 60 s (Fig. 8). The
In this paper, we present evidence for three distinct PO-CL reac-
tionpathwayscatalyzedbyImH.BycomparingUVabsorbances,
we show that the reaction between TCPO and H2O2 to form
hydroperoxyoxalateesterismuchfasterthanthatbetweenTCPO
and ImH to form ODI. When all TCPO-CL reagents are mixed
simultaneously (TCPO-CL reaction), the observed CL intensity
vs. time profiles consistently generate relatively long decay curves
for a wide range of H2O2 concentrations. We propose that the
high-energy intermediate formed in TCPO-CL reactions is the
same six- or eight-membered cyclic intermediate proposed by
Milofsky and Birks11 and Lee et al.13 We also observe relatively
low CL intensities when aryl oxalates are mixed simultaneously
withImHandperyleneintheabsenceofH2O2.Unfortunately,the
high-energyintermediategeneratedundertheseconditionscould
not be resolved based on our experimental results. Finally, we
confirmed the formation of ODI from the reaction between
TCPOandImHusing1H-NMR.WhenTCPOreactswithImHin
thepresenceofperylenebeforetheotherCLreagentsareaddedto
the solution, a relatively fast decaying curve appears under excess
H2O2 conditions. Based on these results, the ODI-CL reaction
pathway appears to differ from the pathways reported by Alvarez
et al.8 and Lee et al.13 However, the possible structures of high-
energy intermediates formed in the ODI-CL reaction may be
analogous to those suggested for PO-CL. Table 7 summarizes the
differences between TCPO-CL and ODI-CL reactions.
Fig. 8 Effect of H2O2 concentration on ODI-CL reactions (fast
decaying curves) and TCPO-CL reactions (slowly decaying curves)
after TCPO reacted with ImH in ethyl acetate for 60 s. [TCPO] =
0.25 mM, [ImH] = 0.5 mM, [perylene] = 0.05 mM. [H2O2] (mM) = (a)
2.0, (b) 7.0, (c) 16, and (d) 56.
Acknowledgements
This research was partially funded with the support of the U.S.
Department of Energy (DOE) Cooperative Agreement No.
DE-FC04-95AL85832 with the Amarillo National Resource
Center for Plutonium (ANRCP). However, any opinions, find-
ings, conclusions or recommendations expressed herein are
those of the authors and do not necessarily reflect the views of
DOE or ANRCP.
fast decaying CL curves generated by the ODI-CL reaction are
more sensitive to varying H2O2 concentration than are the
slowly decaying CL curves produced by the TCPO-CL reaction
808
J. Chem. Soc., Perkin Trans. 2, 2002, 802–809

View Article Online
Table 7 Comparison between the TCPO-CL reaction and the ODI-CL reaction in the presence of ImH
TCPO-CL reaction
ODI-CL reaction
τmax
Sensitivity
H2O2
Slow
Low
Very fast
High
CL intensity increases and τmax and τhalf decrease with its increase.
CL intensity increases to a certain point and then
decreases and τmax and τhalf decrease with its increase.
CL intensity increases to a certain point and then
decreases with its increase.
CL intensity increases with its addition to a certain
point and then begins to decrease.
ImH
H2O
CL intensity increases and τmax and τhalf decrease with its increase.
CL intensity increases with its addition to a certain point and then
begins to decrease.
15 C. V. Stevani, I. P. d’A. Campos and W. J. Baader, J. Chem. Soc.,
Perkin Trans. 2, 1996, 1645.
16 C. V. Stevani, D. F. Lima, V. G. Toscano and W. J. Baader, J. Chem.
Soc., Perkin Trans. 2, 1996, 1989.
17 A. G. Hadd and J. W. Birks, J. Org. Chem., 1996, 61, 2657.
18 C. V. Stevani and W. J. Baader, J. Phys. Org. Chem., 1997, 10, 593.
19 A. G. Hadd, A. L. Robinson, K. L. Rowlen and J. W. Birks, J. Org.
Chem., 1998, 63, 3023.
References
1 K. W. Sigvardson and J. W. Birks, Anal. Chem., 1980, 55, 432.
2 K. Honda, K. Miyaguchi and K. Imai, Anal. Chim. Acta, 1985, 177,
111.
3 J. K. DeVasto and M. L. Grayeski, Analyst, 1991, 116, 443.
4 P. J. Kwakman and U. A. Th. Brinkman, Anal. Chim. Acta, 1992,
266, 175.
5 J. Cepas, M. Silva and D. Pérez-Bendito, J. Chromatogr., 1996, 749,
73.
6 M. M. Rauhut, L. J. Bollyky, B. G. Roberts and M. Loy, J. Am.
Chem. Soc., 1967, 89, 6515.
7 C. L. R. Catherall, T. F. Palmer and R. B. Cundall, J. Chem. Soc.,
Faraday Trans. 2, 1984, 80, 823.
8 F. Alvarez, N. J. Parekh, B. Matuszeski, R. S. Givens, T. Higuchi
and R. L. Schowen, J. Am. Chem. Soc., 1986, 108, 6435.
9 M. Orlovic, R. L. Schowen, R. S. Givens, F. Alvarez, B. Matuszeski
and N. Parekh, J. Org. Chem., 1989, 54, 3606.
10 N. Hanaoka, H. Tanaka, A. Nakamoto and M. Takada, Anal.
Chem., 1991, 63, 2680.
11 R. E. Milofsky and J. K. Birks, J. Am. Chem. Soc., 1991, 113, 9715.
12 M. Stigbrand, E. Ponten and K. Irgum, Anal. Chem., 1994, 66,
1766.
20 H. Neuvonen, J. Biolumin. Chemilumin., 1997, 12, 241.
21 N. Hanaoka, R. S. Givens, R. L. Schowen and T. Kuwana, Anal.
Chem., 1988, 60, 2193.
22 H. Neuvonen, J. Chem. Soc., Perkin Trans. 2, 1995, 951.
23 A. G. Hadd, A. Seeber and J. W. Birks, J. Org. Chem., 2000, 65, 2675.
24 J. H. Esperson, Chemical kinetics and reaction mechanism, 2nd edn.,
McGraw-Hill, New York, 1995, p. 71.
25 M. Stigbrand, E. Pontén and K Irgum, Anal. Chem., 1997, 69, 2109.
26 R. N. Jennings and A. C. Capomacchia, Anal. Chim. Acta, 1988,
205, 207.
27 H. Neuvonen, J. Chem. Soc., Perkin Trans. 2, 1994, 89.
28 G. Orosz and E. Dudar, Anal. Chim. Acta, 1991, 247, 141.
29 G. B. Schuster, Acc. Chem. Res., 1979, 12, 366.
30 J. F. Kirsch and W. P. Jencks, J. Am. Chem. Soc., 1964, 86, 833.
31 R. Weinberger, J. Chromatogr., 1984, 314, 155.
32 G. J. D. Jong, N. Lammers, F. Spruit, R. W. Frei and U. A. Th.
Brinkman, J. Chromatogr., 1986, 353, 249.
13 J. H. Lee, S. Y. Lee and K.-J. Kim, Anal. Chim. Acta, 1996, 329,
177.
14 H. Neuvonen, J. Chem. Soc., Perkin Trans. 2, 1995, 945.
33 G. Orosz, Tetrahedron, 1989, 11, 3493.
J. Chem. Soc., Perkin Trans. 2, 2002, 802–809
809