A colorimetric ligand for mercuric ion

Article abstract of DOI:10.1021/ol990730f

(equation presented) Practical materials are needed to expedite the detection and screening of heavy metals. This Letter describes the synthesis of a new phosphorodithioate-based ligand 1 that specifically communicates a color change when exposed to mercu

Full text of DOI:10.1021/ol990730f

ORGANIC
LETTERS
1
999
Vol. 1, No. 3
15-418
A Colorimetric Ligand for Mercuric Ion
Oliver Br u1 mmer, James J. La Clair,* and Kim D. Janda*
4
Skaggs Institute for Chemical Biology and the Departments of Chemistry and
Molecular Biology, The Scripps Research Institute, 10550 North Torrey Pines Road,
La Jolla, California 92037
kdjanda@scripps.edu
Received June 11, 1999
ABSTRACT
Practical materials are needed to expedite the detection and screening of heavy metals. This Letter describes the synthesis of a new
phosphorodithioate-based ligand 1 that specifically communicates a color change when exposed to mercuric ion. The origin of this recognition
arises through a charge transfer and solubility change. This ligand now extends a new means to visually screen and spectroscopically
quantify mercuric ion.
Human health risks associated with the increased release of
toxic metals over the last century have drawn attention to
their vectors of presentation. This concern has gained an
added dimension due to the fact that elemental or ionic
mercury is converted by aquatic organisms to methylmercury,
extent of their utility and ultimate throughput can be limited
by the need for instrumentation.
1
Materials that change their color upon recognition of
6
macromolecules are widely accepted for fingerprinting or
7
conducting pregnancy tests at home. Clearly, the throughput
2
which subsequently bioaccumulates through the food chain.
of this testing was profoundly advanced by the elimination
of instrumental requirements. We sought to extend a visual
test for mercuric ion. Prior to our investigation, dyes based
on thiocarbazone moieties, such as dithizone and â-naphth-
ylthiocarbazone, were known to respond colorimetrically to
2
+
Recent synthetic activity has focused on tailoring a Hg -
specific sensor. The resulting materials typically contain a
platform for ion recognition as well as an optical transducer
to measure the binding event. In this context, fluorescence
has been used extensively.3 Even though these probes are
-5
2+ 8
Hg . However, the reliability of these systems was limited
2
+
very sensitive and show excellent selectivity for Hg , the
by a complex equilibrium which presented an array of
differently colored complexes. We now demonstrate how to
circumvent this problem through perturbing such equilibria
by selective precipitation.
(
1) (a) Shotyk, W.; Weiss, D.; Appleby, P. G.; Cheburkin, A. K.; Frei,
R.; Gloor, J. D.; Kramers, M.; Reese, S.; Vanderknaap, W. O. Science 1998,
81, 1635. (b) Benoit, J. M.; Fitzgerald, W. F.; Damman, A. W. EnViron.
Res. 1998, 78, 118. (c) McKeown-Eyssen, G. E.; Ruedy, J. Am. J. Epidemiol.
983, 118, 461. (d) Harada, M. Crit. ReV. Toxicol. 1995, 25, 1. (e) Renzoni,
A.; Zino, F.; Franchi, E. EnViron. Res. 1998, 77, 68.
2) (a) Nendza, M.; Herbst, T.; Kussatz, C.; Gies, A. Chemosphere 1997,
2
1
(5) For fluorescent sensors that monitor metal binding in general, see:
(a) Walkup, G. K.; Imperiali, B. J. Org. Chem. 1998, 63, 6727. (b) Walkup,
G. K.; Imperiali, B. J. Am. Chem. Soc. 1997, 119, 3443. (c) Yoon, J. Y.;
Ohler, N. E.; Vance, D. H.; Aumiller, W. D.; Czarnik, A. W. Tetrahedron
Lett. 1997, 38, 3845. (d) Kimura, M.; Horai, T.; Hanabusa, K.; Shirai, H.
AdV. Mater. 1998, 10, 459. (e) Corradini, R.; Dossena, A.; Galaverna, G.;
Marchelli, R.; Panagia, A.; Sartor, G. J. Org. Chem. 1997, 62, 6283. (g)
Desantis, G.; Fabbrizzi, L.; Licchelli, M.; Mangano, C.; Sacchi, D.; Sardone,
N. Inorg. Chim. Acta 1997, 257, 69.
(
3
5, 1875. (b) Boudou, A.; Ribeyre, F. Metal Ions Biol. Systems 1997, 34,
89. (c) Clarkson, T. W. Am. J. Clin. Nutrit. 1995, 61, 682S.
2
(3) For mercuric ion sensitive fluorescent probes, see: (a) Sasaki, D.
Y.; Padilla, B. E. Chem. Commun. 1998, 1581. (b) Hennrich, G.;
Sonnenschein, H.; Resch-Genger, U. J. Am. Chem. Soc. 1999, 121, 5073
and references therein.
(4) For other methods to detect mercury, see: (a) Uria, J. E. S.;
Sanzmedel, A. Talanta 1998, 47, 509. (b) Krenkel, P. A. in HeaVy metals
in the aquatic enVironment; Pergamon Press: Oxford, 1997. (c) Wagemann,
R.; Trebacz, E.; Boila, G.; Lockhart, W. L. Science Total EnViron. 1998,
(6) Hauze, D.; Joulli e´ , M. M. Tetrahedron 1997, 53, 4239.
(7) (a) Bastian, L. A.; Piscitelli, J. T. JAMA 1998, 279, 1264. (b) Bose,
R. J. Assist. Reproduct. Gen. 1997, 14, 497. (c) Alfthan, H.; Stenman, U.
H. Cell. Endocrinol. 1996, 125, 107.
2
18, 19. (d) Moreton, J. A.; Delves, H. T. J. Anal. At. Spectrom. 1998, 13,
6
59. (e) Hardy, S.; Jones, P. J. Chrom. A. 1997, 791, 333. (f) Flamini, A.;
(8) Woidich, H.; Pfannhauser, W. Z. Lebensm. Unters. Forsch. 1972,
149, 1.
Panusam, A. Sens. Actuators 1997, B42, 39.
1
0.1021/ol990730f CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/15/1999

The design was based on monitoring the regulation of a
2
+
dye’s charge transfer upon binding of Hg to a pendant
Scheme 2. Synthesis of Chemosensor 1a
sulfur-containing ligand. Chemosensor 1 contains both a soft
metal binding site, a phosphorodithioate, and an adjacent
9
amine, which is intimately involved in charge transfer
(
Scheme 1).10 We envisioned that the binding of metal ion
Scheme 1a
a
(
a) I
Pd(PPh
rt, 6 h, 89%; (d) add NaHMDS in THF to 6 in DMF, 0 °C to rt, 1
CN, rt, 4
2
, NaHCO
3
, CH
2
Cl
2
, H
2
O, 0 °C, 92%; (b) propargyl alcohol,
Cl
2
3
)
2
, CuI, Et
3
N, THF, rt, 18 h, 93%; (c) MnO
2
, CH Cl
2
2
,
h; add 5 in THF, -20 °C to rt, 8 h, 74%; (e) DBU, CH
h, 24%. Unreacted 2 (55%) from the final step was collected and
3
recycled.
this route, as 7 can act as a cassette wherein other functional-
ity may be readily inserted for the purpose of examining
other metal-ligand interactions. The host chromophore 7
was synthesized in four steps from N-2-hydroxyethyl-N-
methylaniline. Iodination in the para-position, Heck coupling
a
with propargyl alcohol, and oxidation with MnO
aldehyde 5, which was subsequently condensed with diethyl
-hydroxy-4-nitrobenzylphosphonate (6) using a Wadsworth-
2
led to
(A) Indicator 1 senses mercury through subsequent regulation
of its ability to undergo charge transfer. Most metal ions exist in
equilibrium between metal-bound and free states of 1. Certain
metals, such as mercury, form strong complexes that subsequently
precipitate, leaving the aqueous phase colorless (bleached). The
apparent color of the solution is provided next its chemical
equivalent. Precipitated complexes leave the solution colorless. (B)
Acylation of amine 2 diminishes its intramolecular charge transfer.
This acylation was accompanied by a shift in the absorption
maximum from 425 to 398 nm in methylene chloride.
2
Horner-Emmons reaction. Final conversion to 1 was
achieved by phosphorylation of 7 with 2-phenoxy-2-thio-
1
1
1
,3,2-dithiophospholane 8 in the presence of DBU.
Given their high degree of homology, compounds 1 and
could not be distinguished by their absorption of visible
7
light. However, this was not the case when metal was added.
Addition of metal salts to 1 resulted in significant reduction
and hypsochromic shift of the absorption between 415 and
to the phosphorodithioate would alter the electron-donating
capability of the terminal nitrogen in 1 either through ligand
participation or environmental alteration (Scheme 1A),
ultimately diverting the charge transfer and hypsochromically
shifting the absorption maximum.
4
D
50 nm. Dissociation constants (K ’s) were calculated by
comparing the decrease in absorption at 415 nm relative to
that at 250 nm over a series of metal concentrations ranging
2+
3+
3+
2+
from 0 to 35 equiv of metal nitrate (Cd , Eu , Gd , Hg ,
3+
3+
2+
3+
2+
2+
+
12
In , La , Pb , Tb , Co , Zn , K ) (Table 1). Weak
Examination of this approach began with the synthesis of
, an immediate precursor to 1 (Scheme 2). We opted for
2+
2+
+
binding metals (Co , Zn , K ) required up to 150 equiv
7
to alter the absorption of 1. The latter effect was independent
(
9) For metal binding to phosphorodithioates, see: Hayashi, K.; Sasaki,
Y.; Tagashira, S.; Yanagidani, T. Bull. Chem. Soc. Jpn. 1986, 59, 1255.
10) Reichardt, C. Chem. ReV. 1994, 94, 2319.
(11) Br u¨ mmer, O.; Gao, C.; Mao, S.; Weiner, D. P.; Janda, K. D. Lett.
Pept. Sci. 1999, in press. For the synthesis of phosphorodithioates, see
also: Martin, S. F.; Wagman, A. S. J. Org. Chem. 1996, 61, 8016.
(
416
Org. Lett., Vol. 1, No. 3, 1999

D
Table 1. Dissociation Constants, K , for the Interaction of 1
with Various Metalsa
KD (10- M)
9
metal
KD (10-9 M)
metal
Cd2
Co2
Eu3
Gd3
Hg2
+
4200
92500
2500
2600
1.6
In3+
La3+
Pb2+
Tb3+
12200
2300
79.6
+
+
+
2700
+
a
As determined by examining the addition of 0.1-35 equiv of metal
nitrate to 10.5 µM 1 in 35 mM HEPES in 30% acetonitrile/water at pH
7.0. The values presented were established by an average of four repetitions
and were within 2% error
of metal and likely originated from adjustment of the media’s
ionic strength. Titration curves for several metals are
provided in the Supporting Information.
Figure 2. Interaction of 1 with metals after 30 min. The plate was
loaded as described in Figure 1. The final metal concentration is
provided in µM along the y-axis; all metals were added as their
nitrate salts. (A) Solutions resulting after mother liquor was
transferred to another well. The loss of color originates from
precipitation of 1. (B) Transfer of the remaining liquid to another
well leaves behind the metal complex of 1 that provides a yellow
color when dissolved in acetone. Note: cadmium and lead began
to bleach these solutions only after 2 h.
Mercuric ion demonstrated the greatest affinity to 1,
thereby permitting detection of nanomolar quantities in the
presence of millimolar levels of other common metal ions.
The fact that absorption between 415 and 450 nm was not
altered upon exposing 0.1 equiv of any metal to 7 or
nonthiophilic metals to 1 indicated that complexation with
the phosphorodithioate was an absolute requirement for this
response (Scheme 1A).
This system was advanced to provide a visual test. When
2+
examined at greater than 8 µM in 1 and 5 µM Hg , the
addition of mercuric ion resulted in a distinct color change
from yellow to red. Observation of a comparable transition
upon acylation of 2 suggested that this effect originated from
reduction of charge transfer (Scheme 1B). The red complex
precipitated immediately, leaving the aqueous solution color-
Figure 1. Colorimetric response arising from the addition of metal
salts to 1. A total of 10 µL of a 200 µM stock of each metal was
added to a distinct well in a 96-welled white Teflon plate loaded
with 60 µL of a 35 µM acetonitrile solution of 1 and 130 µL of 50
mM HEPES (pH 7.0). The counterions of each metal are indicated
by the color provided in the key: nitrates in black, chlorides in
blue, and sulfates in red. Within seconds of addition, only the well
containing mercury was clear, at which point the mother liquor
was transferred to empty wells. (A) Yellow color appears upon
redissolving the precipitate left in the original well containing Hg2
with acetone. (B) Key. Blank space indicates sample without metal.
13
(12) For general metal ligand interactions, see: O’Sullivan, W. J.;
Smithers, G. W. In Methods in Enzymology; Purich, D. L., Ed.; Academic
Press: New York, 1979; pp 294-336.
(13) (a) La Clair, J. J. J. Am. Chem. Soc. 1997, 119, 7676. (b) La Clair,
J. J. Angew. Chem., Int. Ed. Engl. 1998, 37, 325.
+
Org. Lett., Vol. 1, No. 3, 1999
417

3
1
less or bleached. HPLC, P NMR, and mass spectral analysis
conventional spectrophotometer. Changes in absorption that
are not distinguishable by the naked eye were apparent upon
inspection at a single wavelength between 415 and 450 nm,
2
+
indicated the formation of a 2:1 complex of 1 to Hg (see
1
4
2+
Supporting Information). The precipitation of the 1-Hg
complex15 and thereby removal of the metal ion from the
solution now presents an irreversible color change.
2+
extending the detection limit to 1 ( 0.1 µM Hg .
The interaction of metals with the phosphorodithioate
recognition group in ligand 1 can be effectively quantified
through the spectral changes of a neighboring chromophore.
Probe 1 now presents two dimensions of metal sensing. First,
The selectivity of this chemosensor was demonstrated by
screening 42 different metal salts (Figure 1). Within a second
after addition, only Hg2+ produced an immediate red color
and underwent precipitation, thereby satisfying the required
selectivity. This observation was further verified by transfer-
ring the mother liquor to opaque white wells and by the
reappearance of a yellow color upon dissolving the remaining
precipitate with acetone (Figure 1A). Within 2 h, a similar
2+
3+
the interaction with eleven different metals (Cd , Eu ,
3+
2+
3+
3+
2+
3+
2+
2+
+
Gd , Hg , In , La , Pb , Tb , Co , Zn , K ) could
be quantified in terms of the dissociation constant, K , by
D
using a conventional spectrophotometer. Second, reaction of
2+
1
with Hg is accompanied by a color change and
+
2+
2+
response also occurred in wells containing Ag , Cd , Cu ,
precipitation, providing an irreversible visual test. The
material required for a million assays can be prepared in
approximately 1 week from inexpensive materials, as a single
visual analysis requires less than 150 ng of 1. If desired, a
microtiter plate reader may be used to extend this method
for high-throughput screening and analysis.
2+
2+
Pb , and Pd . The relative visual sensitivity was determined
to be Hg2 >>> Ag , Pd >> Cu > Pb > Cd
Figure 2). These data correlated with the dissociation
constants (K ) of these interactions (Table 1).
+
+
2+
2+
2+
2+
(
D
This analysis provided a confident visual signal when
adding 5 µM mercuric ion to 8 µM 1 as judged by the
observation of colored or bleached dye solutions (Figure 2).
Further quantification of this approach was possible using a
Acknowledgment. This study was supported in part by
the Skaggs Institute for Chemical Biology and the Deutsche
Forschungsgemeinschaft.
(
14) (a) Lawton, S. L. Inorg. Chem. 1971, 10, 328. (b) Burn, A. J.;
Dewan, S. K.; Gosney, I.; McKendrick, K. G.; Warrens, C. P.; Wastle, J.
P.; Watson, C. W. J. Chem. Soc., Perkin Trans. 2 1994, 373.
Supporting Information Available: Experimental pro-
cedures, spectroscopic characterization, metal titration curves,
determination of the K 's, and details about colorimetric
D
assays. This material is available free of charge via the
Internet at http://pubs.acs.org.
(15) Metal ions also form a variety of higher order complexes in both
metal and ligand with phosphorodithiolates. While spectroscopic and mass
determinations indicate that the precipitation under the given conditions
resulted in a 2:1 dye to metal complex, complete precipitation also occurs
with far more metal per ligand. For reports describing the formation and
precipitation of such complexes, see: (a) Drew, M. G. B.; Hobson, R. J.;
Mumba, P. P. E. M.; Rice, D. A. J. Chem. Soc., Dalton Trans. 1987, 1569.
(b) Dakternieks, D. R.; Graddon, D. P. Aust. J. Chem. 1971, 24, 2077.
OL990730F
418
Org. Lett., Vol. 1, No. 3, 1999