A chemiluminescence sensor with signal amplification based on a self-immolative reaction for the detection of fluoride ion at low concentrations

Article abstract of DOI:10.1016/j.tet.2017.05.084

A sensory system incorporated with an amplification function was developed for the detection of trace-level fluoride ions. This sensory system comprises two steps: amplification and chemiluminescence. These steps were linked with chemical reactions and we

Full text of DOI:10.1016/j.tet.2017.05.084

Tetrahedron 73 (2017) 3993e3998
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
A chemiluminescence sensor with signal amplification based on a
self-immolative reaction for the detection of fluoride ion at low
concentrations
,
Shugo Hisamatsu a *, Shinichi Suzuki a, Shigeo Kohmoto b, Keiki Kishikawa b,
Yusuke Yamamoto b, Ryuhei Motokawa a, Tsuyoshi Yaita a
a Actinide Chemistry Group, Materials Sciences Research Center (MSR), Japan Atomic Energy Agency (JAEA), Ibaraki 319-1195, Japan
b Department of Applied Chemistry and Biotechnology, Graduate School of Engineering, Chiba University, 1-33, Yayoi-cho, Inage-ku, Chiba 263-8522, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
A sensory system incorporated with an amplification function was developed for the detection of trace-
level fluoride ions. This sensory system comprises two steps: amplification and chemiluminescence.
These steps were linked with chemical reactions and were induced continuously. The process from
amplification to chemiluminescence was accomplished in this system using fluoride ions. The amplifi-
cation is based on a self-immolative system that permits the detection of emissions even at low fluoride
ion concentrations for systems in which chemiluminescence cannot be induced in the absence of fluoride
ions. The optimal ratio of the chemiluminescent compound and the amplifier was calculated to achieve
efficient amplification.
Received 14 March 2017
Received in revised form
23 May 2017
Accepted 28 May 2017
Available online 29 May 2017
Keywords:
Chemiluminescence sensor
Amplification
© 2017 Elsevier Ltd. All rights reserved.
Self-immolative reaction
Detection of fluoride ions
Sensory system comprises two steps
1. Introduction
and highly recommended method of visual detection.
Chemiluminescence is an attractive option for visual detection
Technology for the detection of trace-level substances is useful
and has been applied in various fields. Specifically, convenient
detection of specific harmful materials, such as fluoride ions, is very
valuable. Fluoride ions cause diseases of the teeth and bones. For
example, the U.S. Environmental Protection Agency recommends a
fluoride concentration of 2.0 ppm in drinking water and has
mandated an upper limit of 4.0 ppm.1 Therefore, sensors and
probes for detecting trace-level fluoride ions are needed in the
food, farming, and environmental research industries.
Inductively coupled plasma mass spectrometry is generally used
to detect trace-level substances. Signal amplification plays an
important role in highly sensitive detection tools such as photo-
multipliers and secondary electron multipliers. However, these
devices require electricity to operate, and they should not be moved
frequently because movement affects their precision. Therefore, a
chemical sensor that can detect objective substances on site would
be very convenient. In particular, luminescent sensors are a clear
of trace-level fluoride ions. Chemiluminescence is induced by the
decomposition of energized compounds, such as peroxides; it can
be controlled using chemical reactions and by changing the reac-
tion conditions. In addition, the induction of chemiluminescence
does not require electricity or excitation light. Emission, including
chemiluminescence, is generally a more sensitive detection method
than absorption, reflection, and light scattering. However, at
extremely low sensor concentrations, even effective emissive sen-
sors almost fail to detect objective substances. Therefore, signal
amplification via substrate recognition is necessary for these sen-
sory systems.
Recently, in terms of amplification based on chemical reactions,
much attention2e9 has been paid to sequential disaggregative re-
actions, known as self-immolation. Self-immolative reactions have
been applied in luminescent sensors and probes,10e15 release and
amplification of low molecular weight compounds,16e19 and drug
delivery systems.20e23 For example, Shabat's group reported the
fluorescence amplification of low-molecular weight compounds by
disaggregation of dendritic compounds or polymers.24e26 Attention
has been focused on signal amplification by cyclical reactions with
* Corresponding author.
E-mail address: shugo.hisamatsu.2012@gmail.com (S. Hisamatsu).
http://dx.doi.org/10.1016/j.tet.2017.05.084
0040-4020/© 2017 Elsevier Ltd. All rights reserved.

3994
S. Hisamatsu et al. / Tetrahedron 73 (2017) 3993e3998
fluoride ions or hydrogen peroxide.27e31 Phillips's group reported
the amplification of chromophore concentration by cyclical re-
actions with fluoride ions or piperidine and the release of alcoholic
compounds from acetals.32e35 Furthermore, Huang's group36 and
Gabrielli's group37 recently reported fluorescent probes with fluo-
ride ion amplification functions.
Our idea was to fabricate a sensory system comprising two
distinct steps. The resulting sensory system is based on a self-
immolative system, as shown in Fig. 1, into which the function of
amplifying fluoride ions is incorporated. The two steps are linked
with chemical reactions and are induced continuously. The first
step involves the recognition of trace-level fluoride ions and the
amplification of fluoride ions. In this step, the amplifier recognizes
trace-level fluoride ions as initiators and self-destructs to release
several fresh fluoride ions. This self-destruction is caused repeat-
edly by the generated fluoride ions, resulting in the creation of
more fluoride ions. This cyclical reaction involving self-destruction
and recombination with fluoride ions leads to amplification of the
fluoride ions. The second step involves the induction of chem-
iluminescence by the amplified fluoride ions. In this second step,
the deprotection of the luminophore precursor by fluoride ions
induces chemiluminescence.
Scheme 1. Reaction sequence of amplification of the chemiluminescence 1 by 2.
In this work, as a chemiluminescence source, we applied com-
pound 1, which was reported previously (Scheme 1).38e41 Com-
pound
1
contains
a
phenol group protected with tert-
butyldimethylsilyl ether and a dioxetane unit stabilized with a
bulky adamantyl group. Compound 1 was deprotected by fluoride
ion to induce chemiluminescence. However, the chem-
iluminescence of 1 is so weak that it can not be detected by the
analyzer or by the naked eye at very low fluoride ion concentra-
tions. Therefore, at sufficiently low fluoride ion concentrations, the
detection of the chemiluminescence of 1 requires appropriate
signal amplification. Compound 2 was prepared as the amplifier of
fluoride ions in the developed sensory system. Fluoride ion
amplification was achieved by disaggregation of 2. Compound 2
was prepared from 4-cresol by protection with tert-butyldime-
thylsilyl ether as the triggering group for fluoride ions and was
equipped with fluoromethyl groups at its 2- and 6-positions. The
protecting group of 2 was cleaved by a fluoride ion, leading to
disaggregation of 2 and the release of two fluoride ions through 1,4-
elimination. The freshly released fluoride ions again reacted with 2.
When this cyclical reaction is performed N times (N: natural
number), 2N fresh fluoride ions are released. Finally, compound 2
was fully consumed and converted into an equivalent amount of
fluoride ions. Using this amplified amount of fluoride ions, chem-
iluminescence of 1 was induced. Compounds 1 and 2 were pre-
pared on the basis of literature reports, as shown in Schemes 2 and
3, respectively.15,30
Scheme 2. Synthesis of chemiluminescent compound 1.
Scheme 3. Synthesis of compound 2 as the amplifier of fluoride ions.
2. Results and discussion
The chemiluminescence performance of 1 was investigated to
determine the lowest amount of fluoride ions required to induce
chemiluminescence of 1. Compound
1 shows blue chem-
iluminescence, with a peak emission at 471 nm in dichloromethane
(DCM), as shown in Fig. 2a. This result is similar to a previously
reported result for another solvent.15,38e40 Time-resolved chem-
iluminescence spectra of 1 were measured by monitoring the
luminescence intensity at the wavelength of 471 nm (Fig. 2b). It was
found that the luminescence intensity depended on the amount of
tetrabutylammonium fluoride (TBAF) added as a fluoride ion
initiator. The addition of 2.1 mM (0.21
detectable chemiluminescence; however, the addition of 0.53 mM
(0.53 mol) of TBAF to 1 caused detectable chemiluminescence.
mmol) of TBAF to 1 caused no
m
Therefore, more than 0.53 mM of fluoride ions from the initiator or
the amplifier is required to maintain constant intensity of the
chemiluminescence of 1. Conversely, inducing chemiluminescence
of 1 with the amplifier 2 resulted in improved detectability of
fluoride ions in the sensory system, despite the addition of less than
0.21 mM of fluoride ions from the initiator.
To investigate the self-immolative reaction of 2, fluorine-19
nuclear magnetic resonance (19F NMR) spectroscopy was con-
ducted with the addition of various amounts of chloroform-d so-
lutions of TBAF; the conversion was then calculated. Fig. 3 indicates
that the self-immolative reaction of 2 completed within a short
Fig. 1. Induced chemiluminescence with signal amplification of fluoride ions in the
developed sensory system.

S. Hisamatsu et al. / Tetrahedron 73 (2017) 3993e3998
3995
The selectivity of 2 for other anions was also investigated
(Fig. 4). The investigation was performed under the same condi-
tions as with TBAF. A series of tetrabutylammonium compounds
were used as sources of other anions instead of TBAF. The con-
centration of each anion used was 0.032 mM. This is the concen-
tration that initiates the reaction of 2 and consumes 2 completely
within 2 h when the anion source is TBAF. However, in each case,
the reaction of 2, which was monitored by 19F NMR spectroscopy
was not initiated 2 h after the addition of each anion. This result
was very similar to previously reported data for analogous
compounds.30,37
Considering the chemical properties of 1 and 2, the fluoride ion
generation was amplified to induce subsequent chem-
iluminescence with the generated fluoride ions. To clarify whether
the reaction was influenced by either 1 or 2, their amounts were
varied, and the induced chemiluminescence intensities were
measured by monitoring the intensity at 471 nm. Fig. 4 shows the
time-dependent chemiluminescence intensities with the addition
Fig. 2. Chemiluminescence performance of 1. (a) Time-resolved chemiluminescence
spectra of 1 (9.0 mg, 22 mol) in DCM (ca. 1 mL). Chemiluminescence was induced
with a dichloromethane solution of TBAF (2.1 mM, 2.1 mol). The lines with different
colors correspond to respective measurement times. (b) Time-resolved chem-
iluminescence spectra monitored at 471 nm, which is the maximum intensity of the
m
m
spectrum of 1 (9.0 mg, 22
lines correspond to 2.1 mM (2.1
0.21 mM (0.21 mol) of TBAF added to a solution of 1, respectively.
m
mol) in DCM (ca. 1 mL). The red, yellow, green, and blue
m
mol), 1.1 mM (1.1 mol), 0.53 mM (0.53 mol), and
m
m
m
of (5a) 21e63
of 2 was increased, the luminescence intensity tended to increase
within the 21e63 mol range of added 2 (Fig. 5a). In contrast, when
mmol and (5b) 63e349 mmol of 2. When the amount
m
the amount of 2 exceeded this range, the luminescence intensity
began to decrease (Fig. 5b). Fig. 5c shows plots of the relative
emission yields against the amount of 2 during the initial 2 h after
addition. The yields were calculated by integrating the lumines-
cence intensity at the wavelength of 471 nm. The profile indicates
that 1.9 equivalent amounts of 2 to 1 induced the maximum
luminescence intensity; an excessive amount of 2 caused the in-
tensity to decrease.
Similar luminescence behavior was observed as a function of the
amount of 1. An increase in the amount of 1 within the range of
7.2e36 mmol led to intensification of the luminescence (Fig. 6a).
Fig. 3. Conversion of 2 versus reaction time during the self-immolative reaction. The
conversions were calculated from the 19F NMR spectra of 2. The red, orange, green,
From the time-dependent intensities of 1, a profile was plotted
using the integrated values of luminescence intensity for 2 h for
each amount of 1 added (Fig. 6b). The profile indicates that 2.4
equivalent amounts of 2 to 1 induces the maximum luminescence
blue, and black lines indicate the conversion patterns of 0.32 mM (0.20
mmol), 32 mM
(20 nmol), 3.2 M (2.0 nmol), and 0.32 M (0.20 nmol) of TBAF and the solution
m
m
without fluoride ions, respectively. The sample solution was prepared by mixing 2
(25 mg, 0.087 mmol), triethylamine (6% solution in CDCl3, 0.35 mL, 0.15 mmol), and
CDCl3 (0.25 mL). An appropriate amount of chloroform-d solution (20
added to each solution as a fluoride ion initiator.
mL) of TBAF was
intensity, and further addition of 1 (36e86
mmol) causes the
luminescence intensity to decrease.
Considering the results of these two cases, we deduced that the
optimum ratio of 2 to 1 is 1.9e2.4 for efficient amplification of
chemiluminescence. The existence of an optimum mixture ratio
during the addition of 1 and 2 to the sensory system may be due to
chemiluminescence quenching caused by the byproducts of 1 or 2
or by the radical coupling reaction of 1 when its amount is increase
in the amount of 1. Therefore, when a large excess of 1 was used, the
chemiluminescence intensity decreased due to the byproducts of 1
and the radical coupling reaction of 1. In contrast, when a large
period (15e30 min) because of the exponential release of fluoride
ions. Furthermore, it shows that the reaction of 2 completed even
with a small amount of TBAF (0.32 mM, 0.20 nmol), although the
reaction required a long time to start. The reason for the long delay
of the reaction with 2 is that the amplification of fluoride ions oc-
curs exponentially. The amplification of fluoride ions can be divided
into two phases. The first phase is the initial reaction phase, which
occurs immediately after the start of amplification. In this phase,
the fluoride ions from the TBAF initiator account for the majority of
the fluoride ions in the system. The chain reaction involved in the
amplification of fluoride ions is the intermolecular reaction be-
tween 2 and a fluoride ion. Therefore, in the initial-reaction phase,
the rate of the chain reaction is slow; the fluoride ions gradually
increase exponentially (2N). The second phase is the state in which
the quantity of fluoride ions is sufficient to react with 2. In this
phase, the rate of the chain reaction increases. Therefore, the chain
reaction depends on the amount of initiator in the first phase; thus,
the smaller the amount of initiator added, the slower the rate of
amplification of fluoride ions, and the more time required to reach
the second phase. Meanwhile, 2 was consumed rapidly by the chain
reaction regardless of the amount of initiator. Therefore, in all cases,
2 was consumed completely within a short period (15e30 min), as
shown in Fig. 3. This result also indicates that the amplification of
fluoride ions with 2 can occur even with a lower amount of initiator
Fig. 4. Selectivity of the amplifier 2 for other anions. The conversions were calculated
from the 19F NMR spectra of 2 2 h after the addition of each anion. The sample solution
was prepared by mixing 2 (25 mg, 0.087 mmol), triethylamine (6% solution in CDCl3,
0.35 mL, 0.15 mmol), and CDCl3 (0.25 mL). Chloroform-d solutions (0.032 mM, 20 mL) of
the tetrabutylammonium series (Clꢀ, Brꢀ, Iꢀ, AcOꢀ, NO3ꢀ, HSOꢀ4 , H2PO4ꢀ, Nꢀ3 , and SCNꢀ
were added to each solution instead of TBAF as the initiator.
)
(0.32
mM) than the amount of fluoride ions (0.21 mM, see Fig. 2)
that causes no detectable chemiluminescence of 1.

3996
S. Hisamatsu et al. / Tetrahedron 73 (2017) 3993e3998
Fig. 7. Chemiluminescence performance of 1 with amplified Fꢀ from a small amount of
TBAF (0.20 M 2.0 nmol) as the initiator. Time-resolved chemiluminescence spectra of
m
1 monitored at 471 nm. The red and blue lines correspond to the spectra with and
without amplifier, respectively. The amounts of 1, 2, and triethylamine were fixed as
follows: 1 (15 mg, 36 mmol), 2 (25 mg, 87 mmol), and triethylamine (0.22 mmol) in
DCM solution (1 mL).
(100
mM) and is synthesized in two steps. While, the probe of
Akkaya's group was a chemiluminescence sensor for fluoride ions.
Their probe was equipped with two 1,2-dioxetane units as a
chemiluminescence source. Therefore, the chemiluminescence
quantum yield of their probe was almost twice that of the single
1,2-dioxetane unit. Furthermore, their probe can detect a low
concentration of fluoride ions. The detection limit of their probe
Fig. 5. Chemiluminescence performance of 1 with varying amounts of 2. The amounts
of 1, triethylamine, and TBAF were fixed as follows: 1 (9.0 mg, 22
(0.22 mmol) and TBAF (20 M, 20 nmol). (a) Time-resolved chemiluminescence
spectra of 1 monitored at 471 nm while varying the amount of 2 (21e63 mol). The
color of the line corresponds to the amount of 2 added. (b) Timeeresolved chem-
iluminescence spectra of monitored at 471 nm by varying the amount of
(63e349 mol). The color of the line corresponds to the amount of 2 added. (c) Plots of
mmol), triethylamine
m
m
was determined to be 47
mM in DMSO and 0.67 mM in DMSO/
aqueous buffer mixture (90/10, pH 7.2). Compared to the probe of
Gabrielli's group, our sensory system is superior in the following
two ways. One is that our probe can detect fluoride ions in lower
1
2
m
emission yields during the initial 2 h versus the amount of 2.
amounts i.e., 0.20 mM (0.20 nmol), than their probe. The other is
that our probe can detect fluoride ions by chemiluminescence
without electricity or excitation light. On the other hand, compared
to the probe of Akkaya's group, our sensory system is superior
slightly in terms of the detectable concentration for fluoride ions
and further improvement is possible with modification of each the
steps. However, in this work, a series of process from amplification
of fluoride ions to inducement of chemiluminescence was per-
formed in non-protonic solvents rather than mixed aqueous sol-
vents used by Gabrielli's group and Akkaya's group. The reason is to
prevent the chemiluminescence of 1 from being quenched by a
protonic solvent such as water and alcohol. Using a mixed solvent
used by Akkaya's group or an enhancer reported in the past39 may
reduce the quenching of chemiluminescence of 1. However, the
factors involved in our sensory system increase and it led to become
complicated our sensory system. We focused on the process from
the amplification of fluoride ions to inducement of chem-
iluminescence in this work and therefore we performed with non-
protonic solvents in order to simplify the sensor system.
Fig. 6. Chemiluminescence performance of 1 with varying amounts of 1. The amounts
of 2, triethylamine, and TBAF were fixed as follows: 2 (25 mg, 87
(0.22 mmol), and TBAF (20 M, 20 nmol). (a) Time-resolved chemiluminescence
spectra of 1 monitored at 471 nm with varying amounts of 1 (7.2e36 mol). The color
mmol), triethylamine
m
m
of the line corresponds to the amount of 2 added. (b) Plots of emission yields during
the initial 2 h versus the amount of 1.
excess of 2 was used, the chemiluminescence intensity also
decreased due to the byproducts of 2.
Considering the optimum ratio of 2 to 1 of 1.9e2.4 for efficient
amplification of chemiluminescence, it was verified that the
inducement of chemiluminescence was possible via amplification
3. Conclusions
We accomplished the processes of amplification and chem-
iluminescence in a sensory system. Using signal amplification, the
developed system can detect fluoride ions at extremely low con-
centrations, even lower than those required to observe the chem-
iluminescence of 1. We determined an optimum mixture ratio of 2
and 1, which are the fluoride ion amplifier and the chem-
iluminescence source, respectively. In addition to functioning as a
fluoride ion sensor, the processes of amplification and chem-
iluminescence in the developed sensory system can be applied to
the detection of various substances involved in the disaggregation
process in which fluoride ions are released when an objective
substance is recognized. In our laboratory, further study of this
sensory system for recognition of radioactive nuclides is under-
way. We believe that the present concept of linking two or more
with the lowest amount of fluoride ions (0.32
mM 0.20 nmol)
detectable by 19F NMR. Fig.
7
shows time-resolved chem-
iluminescence spectra of 1 in the presence and absence of the
amplifier 2. When the amplifier 2 was present, the induced
chemiluminescence was clearly detected, in contrast to the absence
of amplifier 2. Therefore, fluoride ions were detectable by the
chemiluminescence of 1 using amplifier 2 even for very low
amounts such as 0.20 m
M (0.20 nmol). Recently, Gabrielli's group37
and Akkaya's group15 reported a self-immolative probe for fluoride,
respectively. The probe of Gabrielli's group was a multimodal
detection probe with two modes: UVevis absorption and fluores-
cence. Furthermore, their probe can detect amounts of fluoride ions

S. Hisamatsu et al. / Tetrahedron 73 (2017) 3993e3998
3997
reactions, such as amplification and chemiluminescence, is highly
useful for detecting trace-level substances.
room temperature and was stirred at room temperature for 23 h.
After the addition of 15 mL aqueous methanol (2:1, v/v), the
resulting solution was diluted with water (100 mL) and extracted
with chloroform (20 mL ꢁ 2). Organic layer was washed with a
saturated aqueous solution of NaHCO3 (20 mL ꢁ 2). Organic layer
was dried over anhydrous sodium sulfate. After evaporation of
solvent, the residue was purified by chromatography on silica gel
using hexane/ethyl acetate (3:1, v/v) and with continuous supply of
ethyl acetate as the eluent to afford 5 (1.98 g, 97%) as a pale yellow
4. Experimental section
4.1. General procedures
4.1.1. Materials and methods
All reagents employed here were obtained commercially and
were used as received without further purification. All solvents
used were of special grade. Column chromatography was per-
formed with silica gel (0.063e0.2 mm). Melting points were
determined by introducing a tiny amount into a small capillary
tube, attaching this tube to MEL-TEMP II (LABORATORY DEVICE INC,
USA), heating the bath slowly, and observing the temperatures at
which melting begins and is complete. IR spectra were recorded on
a JASCO FT-IR spectrometer with polytetrafluoroethylene (PTFE)
plates. Preparative GPC was performed using GPC columns (JAIGEL
1H-40 and 2H-40, Japan Analytical Industry Co., Ltd). 1H NMR
spectra were recorded on a JEOL JNM-ECA600 spectrometer in
CDCl3 with Me4Si as an internal standard. Chemiluminescence
spectra were recorded on a HITACHI F-7000 fluorescence spectro-
photometer in DCM. High-resolution ESI mass spectra were
measured using a Thermo Fisher Exactive mass spectrometer.
oil. 1H NMR (600 MHz, CDCl3)
d
8.19 (d, J ¼ 8.4 Hz, 2H), 7.63 (t,
J ¼ 7.8 Hz, 1H), 7.50 (t, J ¼ 7.5 Hz, 2H), 7.44 (t, J ¼ 7.8 Hz, 1H), 7.35 (d,
J ¼ 7.8 Hz, 1H), 7.31 (s, 1H), 7.20 (d, J ¼ 7.2 Hz, 1H), 4.57 (d,
J ¼ 16.2 Hz, 2H), 3.71 (dd, J ¼ 12.3 Hz, 10.5 Hz, 6H), 3.42 (s, 3H); 13
C
NMR (150 MHz, CDCl3)
d 165.1, 151.2, 136.2, 133.8, 130.2, 129.7,
129.5, 128.7, 125.5 (d, J ¼ 5.7 Hz, P-C coupling), 122.1 (d, J ¼ 2.9 Hz,
P-C coupling), 121.2 (d, J ¼ 5.7 Hz, P-C coupling), 80.5, 79.3, 59.0 (d,
J ¼ 16 Hz, P-C coupling), 54.0 (dd, J ¼ 24 Hz, J ¼ 5.8 Hz, P-C
coupling); HRMS (ESI), calcd for C17H20O6P [M þ H]þ 351.0992,
found 351.0992.
Compound 6. To a solution of 5 (1.05 g, 3.0 mmol) in THF (18 mL)
was added LDA (1.5 M solution in THF, 10 mL, 15 mmol) at ꢀ78 ꢂC.
After stirring of the reaction mixture for 30 min at ꢀ78 ꢂC, 2-
adamantanone (495 mg, 3.3 mmol) was added to the reaction
mixture at ꢀ78 ꢂC. The reaction mixture was allowed to warm to
room temperature and stirred at room temperature for 24 h. The
reaction was quenched with a saturated aqueous solution of
NaHCO3 (130 mL). The resulting solution was extracted with ethyl
acetate (20 mL ꢁ 6). Organic layer was dried over anhydrous so-
dium sulfate. After evaporation of solvent, the residue was purified
by chromatography on silica gel (hexane/ethyl acetate, 5:1).
Furthermore, the crude was purified by preparative GPC (in chlo-
roform, 65 min) to afford 6 (231 mg, 28%) as a colorless oil. IR (PTFE)
3347 (m), 3063 (w), 2905 (s), 2846 (m), 1580 (m), 1445 (m), 1306
4.2. Synthesis
4.2.1. Synthesis of the chemiluminescence source (compound 1)
Compound 3. To a solution of 3-hydroxybenzaldehyde (3.00 g,
25 mmol), triethylamine (10 mL, 74 mmol), and DMAP (50 mg,
0.41 mmol) in THF (60 mL) was added benzoyl chloride (3.2 mL,
28 mmol) at room temperature. The resulting mixture was stirred
at room temperature for 18 h. After the reaction, the resulting so-
lution was diluted with ethyl acetate (60 mL) and washed with
water (40 mL ꢁ 3), aqueous 1 M HCl solution (20 mL ꢁ 2) and
saturated aqueous solution of NaHCO3 (40 mL). Organic layer was
dried over anhydrous magnesium sulfate. After evaporation of
solvent, the residue was purified by chromatography on silica gel
(hexane/ethyl acetate, 5:1) to afford 3 (5.47 g, 98%) as a white solid;
(m), 1230 (m) cmꢀ1; 1H NMR (600 MHz, CDCl3)
d
7.22 (t, J ¼ 7.8 Hz,
1H), 6.89 (dd, J ¼ 7.8 Hz, 1H), 6.81e6.80 (m, 1H), 6.77 (dd, J ¼ 8.4 Hz,
2.4 Hz, 2H), 4.84 (s, 1H), 3.30 (s, 3H), 3.24 (s, 1H), 2.66 (s, 1H),
1.98e1.76 (m, 12H); 13C NMR (150 MHz, CDCl3)
d 155.9, 143.0, 136.9,
132.4, 129.2, 121.9, 116.0, 114.7, 57.9, 39.3, 39.1, 37.2, 32.3, 30.4, 28.4;
HRMS (ESI), calcd for C18H23O2 [M þ H]þ 271.1693, found 271.1694.
Compound 7. To a solution of 6 (205 mg, 0.76 mmol) and
imidazole (67 mg, 0.99 mmol) in DCM (6.0 mL) was added tert-
butyldimethylsilyl chloride (149 mg, 0.99 mmol) at room temper-
ature. The resulting mixture was stirred at room temperature for
17 h. After the reaction, the resulting solution was diluted with
chloroform (40 mL) and washed with water (20 mL ꢁ 4). Organic
layer was dried over anhydrous magnesium sulfate. After evapo-
ration of solvent, the residue was purified by chromatography on
silica gel by using hexane/ethyl acetate (5:1, v/v) as the eluent to
afford 7 (242 mg, 83%) as a colorless oil. IR (PTFE) 3062 (w), 2907(s),
mp 136e137 ꢂC. 1H NMR (600 MHz, CDCl3)
d 10.03 (s, 1H), 8.21 (dd,
J ¼ 8.4 Hz, 1.8 Hz, 2H), 7.80 (d, J ¼ 7.8 Hz, 1H), 7.75 (t, J ¼ 1.8 Hz, 1H),
7.66 (t, J ¼ 7.2 Hz, 1H), 7.61 (t, J ¼ 7.2 Hz, 1H), 7.53 (t, J ¼ 7.8 Hz, 1H),
7.51e7.49 (m, 1H); HRMS (ESI), calcd for C14H11O3 [M þ H]þ
227.0703, found 227.0709.
Compound 4. A solution of 3 (3.83 g, 17 mmol), p-toluene-
sulfonic acid monohydrate (583 mg, 3.4 mmol) in methanol
(100 mL) and 2,2-dimethoxypropane (20 mL) was dehydrated with
4A molecular sieves. The molecular sieves were removed from the
mixture solution by filtration. The filtrate was stirred at 80 ꢂC for
27 h. After the reaction, the resulting solution was concentrated
and diluted with ethyl acetate (120 mL). The solution was filtered
and washed with water (50 mL ꢁ 4). Organic layer was dried over
anhydrous magnesium sulfate. After evaporation of solvent, the
residue was purified by chromatography on silica gel (hexane/ethyl
acetate, 5:1) to afford 4 (3.29 g, 71%) as a yellow oil. 1H NMR
2848 (s), 1597 (m), 1578 (m), 1480 (m), 1286 (m) cmꢀ1
(600 MHz, CDCl3)
;
1H NMR
d
7.20 (t, J ¼ 7.8 Hz, 1H), 6.91 (d, J ¼ 7.2 Hz, 1H),
6.81e6.79 (m, 1H), 6.78e6.75 (m, 1H), 3.29 (s, 3H), 3.24 (s, 1H), 2.63
(s, 1H), 1.97e1.75 (m, 12H), 0.98 (s, 9H), 0.20 (s, 6H); 13C NMR
(150 MHz, CDCl3)
d 155.4, 143.4, 136.9, 131.3, 129.0, 122.6, 121.2,
119.4, 57.7, 39.3, 39.1, 37.3, 32.3, 30.2, 28.4, 25.8, 18.3, ꢀ4.4; HRMS
(ESI), calcd for C24H37O2Si [M þ H]þ 385.2557, found 385.2557.
Compound 1. A solution of 7 (178 mg, 0.46 mmol) and methy-
lene blue (7.0 mg, 0.022 mmol) in DCM (11 mL) was irradiated with
a halogen lamp at 0 ꢂC while oxygen gas was passed through it; the
reaction was monitored by TLC. When the TLC showed no presence
of compound 7, the resulting solution was concentrated. The res-
idue was purified by chromatography on silica gel by using hexane/
ethyl acetate (20:1, v/v). Furthermore, the crude residue was puri-
fied by preparative GPC (in chloroform, 59 min) as the eluent to
afford 1 (151 mg, 78%) as a colorless oil. IR (PTFE) 3365 (w), 3066
(600 MHz, CDCl3)
d
8.20 (d, J ¼ 7.2 Hz, 2H), 7.63 (t, J ¼ 7.2 Hz, 1H),
7.51 (t, J ¼ 7.2 Hz, 2H), 7.43 (t, J ¼ 8.4 Hz, 1H), 7.36 (d, J ¼ 7.2 Hz, 1H),
7.34 (s, 1H), 7.19 (dm, J ¼ 6.6 Hz, 1H), 5.44 (s, 1H), 3.36 (s, 6H); 13C
NMR (150 MHz, CDCl3)
d 165.2, 151.0, 140.1, 133.7, 130.3, 129.6,
129.4, 128.7, 124.3, 121.8, 120.3, 102.3, 52.7; HRMS (ESI), calcd for
C
16H16O4Na [M þ Na]þ 295.0941, found 295.0938.
Compound 5. To a solution of 4 (1.60 g, 1.7 mmol) and trimethyl
phosphite (1.0 mL, 8.8 mmol) in DCM (17 mL) was added TiCl4
(1.0 M solution in DCM, 8.7 mL, 8.8 mmol) at ꢀ78 ꢂC. After stirring
at ꢀ78 ꢂC for 30 min, the resulting mixture was allowed to warm to

3998
S. Hisamatsu et al. / Tetrahedron 73 (2017) 3993e3998
(w), 2934 (s), 2858 (s),1601 (m),1584 (m),1472 (m),1291 (m) cmꢀ1
1H NMR (600 MHz, CDCl3)
;
chemiluminescence of 1 was performed as follows.
d
7.50e6.70 (br, 3H), 6.87 (d, J ¼ 6.6 Hz,
Appropriate amounts of 2 and triethylamine (6% solution in
DCM, 0.50 mL, 0.22 mmol) were added to a fluoropolymer tube. To
this mixture solution was added TBAF (1.0 ꢁ 10ꢀ3 M solution in
1H), 3.22 (s, 3H), 3.01 (s, 1H), 2.23 (s, 1H), 1.91 (d, J ¼ 13 Hz, 1H),
1.79e1.46 (m, 10H), 1.23 (d, J ¼ 15 Hz, 1H), 0.97 (s, 9H), 0.18 (s, 6H);
13C NMR (150 MHz, CDCl3)
d
155.7, 136.2, 129.3, 121.4, 112.1, 95.6,
DCM, 0.020 mL, 0.020 mmol). The resulting mixture was stirred at
50.0, 36.5, 34.8, 33.2, 32.5, 31.7, 31.6, 26.1, 26.0, 25.8, ꢀ4.3; HRMS
room temperature for 2e4 h; the reaction was monitored by TLC.
When TLC showed no presence of compound 2, the mixture was
added to a solution with an appropriate amount of 1 in DCM
(0.48 mL). Immediately, thereafter, the chemiluminescence spectra
of 1 were recorded.
(ESI), calcd for C24H37O4Si [M þ H]þ 417.2456, found 417.2455.
4.2.2. Synthesis of the amplifier (compound 2)
Compound 8. To a solution of 2,6-bis(hydroxymethyl)-p-cresol
(2.00 g, 12 mmol) and imidazole (2.67 g 39 mmol) in DMF (15 mL)
was added tert-butyldimethylsilyl chloride (5.91 g, 39 mmol) at
room temperature. The resulting mixture was stirred at room
temperature for 16 h. After the reaction, the resulting solution was
diluted with ethyl acetate (100 mL) and washed with water
(50 mL ꢁ 4). Organic layer was dried over anhydrous magnesium
sulfate. After evaporation of solvent, 8 was obtained as a colorless
oil (6.14 g). The crude product was used in subsequent reactions
Acknowledgements
This research was supported in part by a Grant-in-Aid for Young
Scientists B (15K18322) from Japan Society for the Promotion of
Science (JSPS). SH is grateful to Dr. T. Sekiguchi of JAEA for fruitful
discussions.
without further purification. 1H NMR (600 MHz, CDCl3)
d 7.15 (s,
1H), 4.68 (s, 4H), 2.32 (s, 3H), 1.02 (s, 9H), 0.94 (s, 18H), 0.15 (s, 6H),
0.08 (s, 12H); HRMS (ESI), calcd for C27H54O3NaSi3 [M þ Na]þ
533.3269, found 533.3273.
References
1. EPA. National Primary Drinking Water Standards. 2009.
2. Li S, Szalai ML, Kevwitch RM, McGrath DV.
J Am Chem Soc. 2003;125:
Compound 9. To a solution of 8 (6.14 g, 12 mmol) in MeOH
(25 mL) was added p-toluenesulfonic acid monohydrate (69 mg,
0.36 mmol) at room temperature. The resulting mixture was stirred
at room temperature; the reaction was monitored by TLC. When
TLC showed no presence of compound 8, the resulting solution was
diluted with ethyl acetate (100 mL) and neutralized with a satu-
rated aqueous solution of NaHCO3 (20 mL ꢁ 3). Organic layer was
washed with water (40 mL ꢁ 4) and dried over anhydrous sodium
sulfate. After evaporation of solvent, the residue was purified by
chromatography on silica gel (hexane/ethyl acetate, 5:1) to afford 9
(2.19 g, 65%) as a white solid; mp 80e81 ꢂC. IR (PTFE) 3245 (w),
10516e10517.
3. Blencowe CA, Russell AT, Greco F, Hayes W, Thornthwaite DW. Polym Chem.
2011;2:773e790.
4. Kevwitch RM, Shanahan CS, McGrath DV. New J Chem. 2012;36:492e505.
5. Zhang H, Yeung K, Robbins JS, et al. Angew Chem Int Ed. 2012;51:2400e2404.
6. Gnaim S, Shabat D. Acc Chem Res. 2014;47:2970e2984.
7. Kim H, Baker MS, Phillips ST. Chem Sci. 2015;6:3388e3392.
8. Mohapatra H, Kim H, Phillips ST. J Am Chem Soc. 2015;137:12498e12501.
9. Roth ME, Green O, Gnaim S, Shabat D. Chem Rev. 2016;116:1309e1352.
10. Richard JA, Jean L, Romieu A, Massonneau M, Noack-Fraissignes P, Renard PY.
Org Lett. 2007;9:4853e4855.
11. Meyer Y, Richard JA, Massonneau M, Renard PY, Romieu A. Org Lett. 2008;10:
1517e1520.
12. Sella E, Shabat D. Chem Commun. 2008:5701e5703.
13. Richard J-A, Jean L, Schenkels C, Massonneau M, Romieu A, Renard P-Y. Org
Biomol Chem. 2009;7:2941e2957.
14. Weinstain R, Baran PS, Shabat D. Bioconjug Chem. 2009;20:1783e1791.
15. Turan IS, Akkaya EU. Org Lett. 2014;16:1680e1683.
2926 (w), 2883 (w), 2855 (w), 1459 (m), 1200 (s) cmꢀ1
(600 MHz, CDCl3)
7.12 (s, 1H), 4.61 (d, J ¼ 4.8 Hz, 4H), 2.28 (s, 3H),
1.02 (s, 9H), 0.17 (s, 6H); 13C NMR (150 MHz, CDCl3)
147.8, 131.8,
;
1H NMR
d
d
131.7, 129.1, 61.1, 26.1, 20.7, 18.7, ꢀ3.6; HRMS (ESI), calcd for
16. Montano LA, Ingle JD. Anal Chem. 1979;51:919e926.
15H26O3NaSi [M þ H]þ 305.1543, found 305.1540.
17. Szalai ML, Kevwitch RM, McGrath DV. J Am Chem Soc. 2003;125:15688e15689.
18. Weinstain R, Sagi A, Karton N, Shabat D. Chem A Eur J. 2008;14:6857e6861.
19. Dharmaraja AT, Ravikumar G, Chakrapani H. Org Lett. 2014;16:2610e2613.
20. De Groot FMH, Loos WJ, Koekkoek R, et al. J Org Chem. 2001;66:8815e8830.
21. Damen EWP, Nevalainen TJ, Van den Bergh TJM, De Groot FMH, Scheeren HW.
Bioorg Med Chem. 2002;10:71e77.
22. Haba K, Popkov M, Shamis M, Lerner RA, Barbas CF, Shabat D. Angew Chem Int
Ed. 2005;44:716e720.
23. Esser-Kahn AP, Sottos NR, White SR, Moore JS. J Am Chem Soc. 2010;132:
10266e10268.
C
Compound 2. To a solution of 4 (711 mg, 2.5 mmol) in DCM
(5.0 mL) was added (diethylamino)sulfur trifluoride (0.99 mL,
7.5 mmol) at 0 ꢂC. The resulting mixture was stirred at 0 ꢂC; the
reaction was monitored by TLC. When TLC showed no presence of
compound 9, the reaction was quenched with water (10 mL).
Organic layer was washed with a saturated aqueous solution of
NaHCO3 (20 mL) and brine (20 mL ꢁ 2) and subsequently dried over
anhydrous sodium sulfate. After evaporation of solvent, the residue
was purified by preparative GPC (in chloroform, 62 min) to afford 2
(268 mg, 37%) as a colorless oil. IR (PTFE) 2956 (s), 2932 (s), 2898
(m), 2860 (s), 1479 (s), 1464 (s), 1257 (s) cmꢀ1; 1H NMR (600 MHz,
24. Amir RJ, Pessah N, Shamis M, Shabat D. Angew Chem Int Ed. 2003;42:
4494e4499.
25. Shamis M, Shabat D. Chem A Eur J. 2007;13:4523e4528.
26. Sagi A, Weinstain R, Karton N, Shabat D. J Am Chem Soc. 2008;130:5434e5435.
27. Sella E, Shabat D. J Am Chem Soc. 2009;131:9934e9936.
28. Avital-Shmilovici M, Shabat D. Bioorg Med Chem. 2010;18:3643e3647.
29. Seila E, Lubelski A, Klafter J, Shabat D. J Am Chem Soc. 2010;132:3945e3952.
30. Perry-Feigenbaum R, Sella E, Shabat D. Chem A Eur J. 2011;17:12123e12128.
31. Karton-Lifshin N, Shabat D. New J Chem. 2012;36:386e393.
32. Baker MS, Phillips ST. J Am Chem Soc. 2011;133:5170e5173.
CDCl3)
d
7.21 (s, 2H), 5.35 (d, J ¼ 48.0 Hz, 4H), 2.31 (s, 3H), 1.04 (s,
9H), 0.18 (s, 6H); 13C NMR (150 MHz, CDCl3)
d 148.9, 131.6, 127.3,
127.2, 80.2 (d, J ¼ 165 Hz, C-F coupling), 26.0, 20.6, 18.8, ꢀ4.0; 19F
NMR (565 MHz, CDCl3) ꢀ206.9 (t, J ¼ 47.7); HRMS (ESI), calcd for
~
33. Nunez SA, Yeung K, Fox NS, Phillips ST. J Org Chem. 2011;76:10099e10113.
F]þ 267.1575, found 267.1556; calcd for
34. Mohapatra H, Schmid KM, Phillips ST. Chem Commun. 2012;48:3018e3020.
35. Baker MS, Phillips ST. Org Biomol Chem. 2012;10:3595e3599.
36. Gu J-A, Mani V, Huang S-T. Analyst. 2015;140:346e352.
C
C
15H24OFSi [M
ꢀ
15H24OF2NaSi [M þ Na]þ 309.1457, found 309.1486.
37. Gabrielli L, Mancin F. J Org Chem. 2016;81:10715e10720.
38. Edgeworth J, Freemont P, Hogg N. Nature. 1989;342, 599e192.
4.3. Amplification of fluoride ions and inducement of
chemiluminescence
€
39. Beck S, Koster H. Anal Chem. 1990;62:2258e2270.
40. Adam W, Bronstein I, Trofimov AV. J Phys Chem A. 1998;102:5406e5414.
41. Hananya N, Eldar Boock A, Bauer CR, Satchi-Fainaro R, Shabat D. J Am Chem Soc.
2016;138:13438e13446.
The amplification of fluoride ions with 2 and inducement of