Molecular anchors - Mimicking metabolic processes in thiol analysis

Article abstract of DOI:10.1039/b611471g

The interaction between various novel naphthoquinones and glutathione was explored using a variety of electrochemical techniques. An adamantane derivatised quinone was found to adsorb strongly to carbon surfaces providing a robust film possessing a distin

Full text of DOI:10.1039/b611471g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
Molecular anchors—mimicking metabolic processes in thiol analysis
Robert B. Smith,*a Claire Canton,a Nathan S. Lawrence,b Callum Livingstonec and
James Davisa
Received (in Montpellier, France) 8th August 2006, Accepted 8th September 2006
First published as an Advance Article on the web 26th September 2006
DOI: 10.1039/b611471g
The interaction between various novel naphthoquinones and glutathione was explored using a
variety of electrochemical techniques. An adamantane derivatised quinone was found to adsorb
strongly to carbon surfaces providing a robust film possessing a distinct and consistent
voltammetric profile which is markedly different from solution based species. The anodic peak
process was found to respond to increasing additions of glutathione and the analytical merits
have been assessed. In contrast to conventional thiol–quinone electrochemistry, the detection
pathway involves the solubilisation and removal of the hydrophobic quinone derivative through
conjugation with glutathione. The peak potential (+0.2 V vs. Ag/AgCl) is such that an
unambiguous response to glutathione can be achieved in the presence of a significant excess
(1 mM) of common physiological components.
can be attributed to the reaction scheme highlighted in Fig. 1B.
Introduction
The naphthoquinone (I) is subject to a 1,4-Michael addition
Glutathione is a sulfur containing tripeptide and is one of the
by the nucleophilic thiol species (II) resulting in the generation
prime protective agents within physiological systems. It facil-
of the corresponding reduced conjugate (III). As such, when
itates detoxification through a number of mechanistic path-
the cathodic scan commences after each addition there is
ways ranging from direct and enzyme mediated electron
progressively less oxidised naphthoquinone (I) available with-
transfer through to metal complexation and more elaborate
in the bulk of the solution and hence the magnitude of the
conjugation processes.1–3 The latter is best exemplified
reductive electrode process diminishes accordingly. It is worth-
through its role in paracetamol metabolism whereby the
while noting that the oxidation process does not increase but
reactive quinone imine intermediates resulting from metabo-
rather broadens out. This is in line with the emergence of the
lism of the analgesic are speedily removed through Michael
new redox couple associated with the conjugate (III) rather
addition by the glutathione present within the liver.3 The
than the simple redox cycling of the native naphthoquinone.
toxicity associated with overdose is related to the effective
depletion of the glutathione reserves. The hydrophilic nature
of the peptide also facilitates the excretion of conjugated
metabolites through increasing the solubility of hitherto hy-
drophobic moieties.1–3 The present communication has sought
to explore and exploit the native metabolic action outlined
above as a means of quantifying thiols using a novel electro-
chemical approach.
Quinone–thiol interactions have been investigated as both a
spectroscopic and electrochemical diagnostic tool for the
quantification of reduced thiol species.4–7 Naphthoquinone,
in particular, has been shown to exhibit higher selectivity than
most benzoquinone derivatives as the former is largely un-
affected by the presence of ascorbate—a prime interferent in
electroanalytical investigations.7 The basic assay is highlighted
in Fig. 1A where cyclic voltammograms detail the typical
responses obtained when glutathione is added to a solution
containing naphthoquinone. The magnitude of the reduction
peak decreases within increasing concentration of the thiol and
a School of Biomedical and Natural Sciences, Nottingham Trent
University, Nottingham, UK NG11 8NS. E-mail:
Robert.smith02@ntu.ac.uk; Tel: 0115 848 3910
b Schlumberger Cambridge Research, High Cross, Madingley Road,
Cambridge, UK CB3 0EL
c Royal Surrey County Hospital NHS Trust, Guildford, Surrey,
Fig. 1 (A) Response of naphthoquinone (1 mM, pH 7) to increasing
additions of glutathione (25 mM). (B) Reaction scheme.
UK GU2 7XX
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
1718 | New J. Chem., 2006, 30, 1718–1724

View Article Online
Table 1 Naphthoquinone derivatives investigated
The key parameter in most assays of this type is that both
indicator and target analyte are present within the solution.4–7
In this instance, however, we have sought to explore the
potential for localising the naphthoquinone indicator directly
at the electrode surface. This would have the advantage of
generating an unambiguous signal which would be modified in
the presence of thiol.
Our strategy was to develop a number of bifunctional
molecular assemblies through tethering naphthoquinone deri-
vatives to an appropriate pendant group. The former serves as
the thiol selective hook with the latter providing the hydro-
phobic anchor through which the assembly could reversibly
adhere to the electrode substrate. The principal species inves-
tigated are shown in Table 1.
Fig. 2 Response of VIII (1 mM, pH 7) to increasing glutathione
concentration (25 mM) and the emergence of a conjugate redox process
(dotted circle).
Results and discussion
There was considerable variation in the response character-
istics for the various derivatives, with IV through to VIII all
producing an initial voltammetric behaviour analogous to that
of the naphthoquinone (I). There was, however, no appreci-
able change in the profile with IV and V upon the addition of
glutathione and this can be attributed to the electron donating
properties inherent to the hydroxy and methoxy functional
groups. The increase in electron density across the double
bond effectively reduces the susceptibility to nucleophilic
attack. The converse is true with the remaining derivatives—
all of which show marked reactivity towards glutathione.
The response of the p-tosyl derivative (VIII) is highlighted in
Fig. 2 where the emergence of a new redox peak process
associated with the reduction of the conjugate can be clearly
observed at À0.2 V.
Voltammograms detailing the electrode response to the
solution based acyloxy (1 mM, pH 7) derivative are detailed
in Fig. 3. The profile is substantially more complex than that
observed with the tosyl systems and is attributed to two
factors. Hydrolysis of the ester bond is relatively rapid,
resulting in the emergence of peaks (À0.4 V and À0.29 V)
associated with the parent 2-hydroxy-1,4-naphthoquinone
(IV). The identity of the latter was confirmed by adding an
aliquot of a standard solution of IV to the cell (dotted line,
Fig. 3).
Fig. 3 Repetitive cyclic voltammograms of IX (1 mM, pH 7) high-
lighting the hydrolysis of the ester linkage. Cyclic voltammogram of
IV is included for comparison.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 1718–1724 | 1719

View Article Online
The second factor is that the peak separation between the
reduction (À0.18 V) and associated oxidation (+0.19 V)
process for the ester derivative is 370 mV which is significantly
greater than that found for the previous investigations (I -
VIII). The peak separations observed with the 2-hydroxy (IV),
2-methoxy and p-tosyl naphthoquinone (VIII) were 110, 95, 92
mV respectively and, while larger than that expected of a 2
electron system, are nevertheless consistent with the quasi
reversible behaviour of the simple NQ derivatives at carbon
substrates. The separation is a consequence of both the ester
functionality and the increased adsorption of the derivative at
the electrode surface. This is clear from the Gaussian shaped
profile observed in the ester peaks (+0.19 V) which contrast
those of the previous examples where diffusion is still an
important contributing factor.
analogous to the behaviour observed with the simple naphtho-
quinone solution detailed in Fig. 1A.
The emergence of a second oxidation process (À0.12 V) in
the cyclic voltammograms detailed in Fig. 4B is attributed to
the resulting glutathione–quinone conjugate. This exhibits the
broad diffusion limited profile expected of a solution based
system and clearly contrasts the sharp adsorption profile
obtained in the absence of the thiol. Square wave voltammo-
grams detailing the response of the oxidation peak processes to
increasing glutathione are shown in the inset in Fig. 4C.
The solubilising effect of the glutathione conjugate as the
principal mode of action at the electrode surface was con-
firmed by removing the electrode from the test solution (after
holding at +0.4 V for 30 s) and then re-scanning the voltam-
mogram in fresh buffer solution. In the absence of glutathione
this could be done repeatedly with only a minimal decrease
(less than 1%) in the quinone peak heights. In the case where
the glutathione (25 mM) is allowed to react with the quinone
assembly, the subsequent fresh buffer test resulted in the
complete removal of the quinone signature. Derivative X
was further modified with a methyl group substituted into
the free quinone ring position (derivative XI) to preclude
reaction with glutathione. The entire process was repeated—
with the new indicator deposited onto the electrode from an
acetone solution and placed in fresh buffer. The voltammetric
profile (not shown) was analogus to that for X and, as in the
previous case, the electrode could be repeatedly removed and
replaced in fresh buffer without accruing any significant loss in
signal intensity. In this case, however, there was no change in
the profile upon the introduction of glutathione. This con-
firmed the necessity of having the free proton as the thiol
selective hook and that the process is not a simple redox
interchange.
The derivatives investigated thus far have provided some
valuable insights into the nature of the tether that needs to
bind the quinone to the substrate anchor but none are
sufficiently hydrophobic in themselves to serve as our detec-
tion indicator. The proposed detection technique requires a
quinone derivative that is effectively insoluble within aqueous
solution, hence the inclusion of the hydrophobic adamantane
pendant group (X) within Table 1.
The adamantane substituent was purposefully attached to a
2-hydroxy-1,4-naphthoquinone core through ester linkages
(X) rather than an ether linkage. While the latter would be
inherently more stable to hydrolysis, the preliminary investi-
gations of V implied that there would be little reaction between
the available ring position and thiol targets whereas the
electron withdrawing properties of the ester functionality
would retain the native quinone reactivity towards thiol
moieties. It was hoped that the greater steric bulk of the
adamantane pendant would impede hydrolysis.
Derivative X was found to be essentially insoluble in aqu-
eous buffer (pH 7) but could be anchored to the electrode
substrate through either solvent casting or dip coating (from
either an acetone solution of the ester or aqueous suspension).
Cyclic voltammograms detailing the response of an electrode
modified with X in fresh buffer (pH 7) are detailed in
Fig. 4A. The profile obtained is again distinct from that
obtained with the simple solution based naphthoquinone
(Fig. 1) and is characterised by the sharp reduction and
oxidation peak processes expected for adsorbed redox species.
The peak separation (410 mV) is again substantially greater
than that expected for such a couple and mirrors the response
obtained with the simpler, methyl ester (Fig. 3). The major
advantage is that in this instance there is no hydrolysis—
confirmed by the absence of any 2-hydroxy-1,4-naphthoqui-
none peaks (IV).
The reduction of quinones within well buffered aqueous
media ordinarily will tend to follow a 2 electron/2 proton
The quinone peaks were found to be stable in the buffer
solution and did not diminish in magnitude, thus confirming
the anchor like properties of the derivative. Cyclic voltammo-
grams detailing the response of the modified electrode to
increasing glutathione concentration (25 mM) are shown in
Fig. 4B. The introduction of the thiol serves to solubilise the
adsorbed quinone and effectively lifts the anchor into the bulk
of the solution. This is shown by the marked decreases in the
heights of both the oxidation and reduction peaks associated
with the adsorbed quinone–adamantane assembly and is
Fig. 4 (A) Voltammograms detailing the response of adsorbed X in
fresh pH 7 buffer. (B) Influence of glutathione (25 mM additions) on
the cyclic voltammetric behaviour of adsorbed IV in pH 7 buffer. (C)
Square wave voltammograms detailing the removal of adsorbed X
with increasing glutathione concentration.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
1720 | New J. Chem., 2006, 30, 1718–1724

View Article Online
route—whereby the initial reduction to the anion is rapidly
followed by protonation leading to the production of the
hydroquinone analogue through an ECEC process. In aprotic
or poorly buffered medium, then, it could be expected that
sequential 1 electron conversions to the semiquinone anion
and lastly to the dianion would be observed.8–12 Interestingly,
the oxidation of hydroquinones within such systems tends to
exhibit electrochemically irreversible behaviour as a conse-
quence of the localised lowering of pH. The voltammograms
detailed in Fig. 4A show a number of similarities with the
latter and may arise as a consequence of the film structure on
the electrode surface. This may be attributed to a combination
of the hydrophobic nature of the film and the possible exclu-
sion of the quinone moieties from the bulk solution. This
would mean that the redox species experience an environment
akin to that of the poorly buffered systems and hence explain
the deviation of the profile from conventional reversible
systems normally expected in buffered aqueous media. Close
inspection of the voltammetric profile detailed in Fig. 4A
reveals the presence of a small anodic process at À0.05 V.
The intensity of this process was found to increase markedly
relative to the existing, and hitherto principal, anodic process
(+0.2 V) when the scan rate was increased (Fig. 5). This can
be attributed to the re-oxidation of the anionic form of the
semiquinone prior to protonation. This is now rate limiting as
indicated by the scan rate dependence. At slow scan rates there
is sufficient time for the protonation of the anionic sites within
the film before scan reversal. The increased scan rate effec-
tively captures the transition between semiquinone anion and
quinone with the oxidation of the hydroquinone formed in
competition by the ECEC process appearing as the anodic
peak at higher potentials (+0.2 V). A similar process is
observed with the acyloxy (IX) derivative shown in Fig. 3.
Upon reducing the quinone to the hydroquinone form the
derivative becomes momentarily more soluble as a conse-
quence of the conversion from quinone to anion to the
Fig. 6 Cyclic voltammograms detailing the response obtained after
rotation (3000 rpm, 2 min) at open circuit (solid line) and at À0.8 V
(dotted line).
protonated form. As such it is possible that in a convective
system significant losses of material may occur. In order to test
this hypothesis a rotating disk experiment was conducted. The
rationale behind the experiment is that if there is a significant
difference in the solubility of the components as a consequence
of the reduction process then the rotation should remove the
material from the electrode before it can reassemble within the
layer. The derivative was adsorbed onto a 3 mm glassy carbon
disk electrode and placed within fresh pH 7 buffer. The
electrode was rotated repeatedly at 3000 rpm for two minute
periods until a stable voltammogram was obtained. This was
done under open circuit and was designed to effectively
remove loosely bound material. The cyclic voltammogram
obtained is shown in Fig. 6 and was found to be stable,
indicating that the oxidised form is strongly adsorbed to the
electrode surface. The electrode was again rotated at 3000 rpm
but in this instance the electrode potential was held at À0.8 V—
the potential being applied after commencing the rotation. The
voltammograms show that upon rotation at À0.8 V substantial
losses of material for the adsorbed layer occurred. It was not,
however, possible to completely remove the quinone layer.
Analytical characterisation
Essentially identical responses were obtained to glutathione,
cysteine and homocysteine and they were linear up to 250 mM
of thiol (peak current/A = À0.001 28; [glutathione/mol LÀ1
]
À6.6 Â 10À9; N = 5; R2 = 0.992) with a detection limit in the
order of 5 mM (based on 3sb). The narrow linear range is due to
the film coverage as the signal is based on a decreasing peak
intensity. Increasing the adsorption time had little effect on
increasing the height of the oxidation peak but it is susceptible
Fig. 5 Influence of scan rate on the voltammetric response of
adsorbed X in fresh pH 7 buffer.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 1718–1724 | 1721

View Article Online
to variation in electrode area. This will obviously influence the
inter-electrode reproducibility (410%) within the present
system given the ad hoc nature of the preparation.
sence of the antioxidant should not compromise thiol
detection. There is another feature of the film and the pro-
posed detection method which could be of analytical value and
is significantly different from conventional electrochemical
thiol measurements. Should there be a case where the voltam-
metric peak of the quinone oxidation is obscured by the
presence of another electroactive component, matrix exchange
could be employed to counter the lack of initial resolu-
tion—providing the interfering substance does not chemically
interact with the film. It can be seen from Fig. 7 (dotted line)
that the peak integrity is retained upon removing the electrode
from the solution and placing it in fresh buffer. It could be
anticipated that upon exposing the electrode to the test solu-
tion any thiols present would react and induce the removal of
the film constituents. Upon removal of the electrode to fresh
solution the concentration of thiol could be unambiguously
quantified from the change in the film peak height. This is
admittedly more clumsy than a direct analysis. Nevertheless,
the chief concept of thiol solubilisation as a diagnostic tool has
been shown to be, in principle, viable with a sensitive signal
that is within a dynamic range potentially suitable for a host of
biomedical applications.
The selectivity of the approach was assessed in terms of the
influence of common functional groups that could have nu-
cleophilic character and hence could act in much the same way
as the thiol targets. There was no response to lysine (up to
1 mM) but this could be expected as the alkyl amino will be
predominantly protonated at physiological pH. The response
to histidine was also assessed as the imidazole substituent
possesses amine functionality that has a pKa of 5 and hence
nucleophile potential.13–15 There was essentially no response
to the amino acid (1 mM). In this case the lack of response can
be attributed to the resonance within the imidazole ring which
under the conditions of the experiment did not present an
effective challenge to the film. It is likely, however, that
reaction would occur over a longer timescale.13 The influence
of ascorbate, paracetamol and urate was also assessed. Only
the former presented problems for the interpretation of the
voltammetric profile as the others are oxidised at a potential
significantly more positive than the quinone oxidation peak.
Square wave voltammograms detailing the response of the film
to increasing additions of ascorbate (200 mM) are shown in
Fig. 7.
Conclusions
The peak separation between the quinone and ascorbate is
close but resolution is still possible. It is also important to note
that the concentration of ascorbate in the test solution is
significantly higher than those expected within normal phy-
siological fluid (typically 100–200 mM14) and hence the pre-
A range of novel naphthoquinones have been investigated
with a view to assessing the molecular architecture needed to
induce their speedy and electrochemically quantifiable reac-
tion with reduced thiols. The potential applicability of a
wholly new electroanalytical strategy for the determination
of glutathione, cysteine and homocysteine has been evaluated
and two distinct approaches outlined. The system has been
shown to be selective for RSH functionality with little inter-
ference from other physiological agents. A strategy has been
developed that outlines a potential route for further exploita-
tion and a molecular framework highlighting the key structur-
al and electronic properties required for future development
has been highlighted. While the detection rationale is not
suitable for in vivo applications, there is increasing interest in
the use of thiols as biomarkers for oxidative stress processes in
a wide range of health care contexts and it is possible that the
approach advocated here may be of value within in vitro based
decentralised testing contexts.7 As the detection pathway
involves the cumulative removal of the indicating species,
any potential sensor would necessarily be a single shot device.
Nevertheless, there is increasing interest in the analysis of
antioxidant thiols in a host of diverse biochemical processes
such that their role can be fully understood and disease
pathways elucidated.16–20 This present communication offers
the possibility of adding another diagnostic tool through
which these can be investigated.
Experimental
Methods and materials
Electrochemical measurements were conducted using
a
Fig. 7 Square wave voltammograms detailing the response obtained
in the presence of ascorbate. The response of the same film after
replacing the solution with fresh buffer is also shown (dotted line).
mAutolab type III computer controlled potentiostat (Eco-
Chemie, Utrecht, The Netherlands) using a three electrode
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
1722 | New J. Chem., 2006, 30, 1718–1724

View Article Online
configuration consisting of a glassy carbon working electrode
(3 mm diameter, BAS Technicol, UK), a platinum wire
counter electrode and a 3 M NaCl Ag | AgCl half cell reference
electrode (BAS Technicol, UK). All measurements were done
under nitrogen. NMR spectra were measured on a JEOL ECX
400 MHz spectrometer. Chemical shifts are reported in parts
per million (ppm) downfield from tetramethylsilane (TMS).
Mass spectra were recorded at the EPSRC Mass Spectrometry
Centre at Swansea University. Microanalysis was carried out
using a Leeman CE 440 automatic elemental analyser. Flash
chromatography was performed on 40–63 silica gel (Merck).
All chemicals and solvents were bought from either Acros
Organics or Sigma Aldrich and used without further purifica-
tion. Quinone derivatives IV and V were obtained from Sigma.
130.72, 131.72, 132.36, 134.30, 134.66, 146.52, 151.54,
184.13, 184.31.
CIMS (m/z) 346.1 [M
+
NH4]+. Anal. Calcd for
C17H12O5S: C, 62.19; H, 3.68. Found: C, 61.80; H, 3.68%.
Derivative IX: 2-Acetoxy-1,4-naphthoquinone. Yield 0.21 g
(84.6%) of 2-acetoxy-1,4-naphthoquinone (as the first frac-
tion), as a bright yellow solid.
1H NMR (d-CDCl3) d 2.39 (s, 3 H), 6.75 (s, 1 H), 7.76–7.77
(m, 2 H), 8.08–8.10 (m, 2 H). 13C NMR (d-CDCl3) d 20.67,
126.13, 126.60, 127.06, 131.30, 134.50, 134.10, 134.50, 154.42,
167.83, 178.70, 184.64. EIMS (m/z) 216. Anal. Calcd for
C17H12O5S: C, 66.67; H, 3.73. Found: C, 66.48; H, 4.00%.
Derivative X: 2-(Tricyclo[3.3.1.13,7]decane-1-carboxylic
acid), 1,4-naphthoquinone ester. Yield 0.61 g (63.2%) of 2-
(tricyclo[3.3.1.13,7]decane-1-carboxylic acid), 1,4-naphthoqui-
none ester (as the second fraction), as a bright yellow solid.
1H NMR (d-CDCl3) d 1.68 (brs, 6 H), 1.84 (brs, 3 H), 2.00
(brs, 6 H), 6.63 (s, 1 H), 7.65–7.68 (m, 2 H), 7.97–7.99 (m, 2H).
13C NMR (d-CDCl3) d 27.86, 36.38, 38.73, 126.03, 126.50,
126.90, 131.14, 132.01, 134.00, 134.38, 155.22, 173.39, 174.82,
178.67, 184.54. Accurate LCMS (m/z) 354.170 [M + NH4]+.
Anal. Calcd for C21H20O4: C, 74.90; H, 5.99. Found: C, 74.75;
H, 5.98%.
Syntheses
General procedure. To a stirred solution of the appropriate
starting quinone (typically 1 g) and a two fold excess of acid
chloride (toluenesulfonyl chloride, acetyl chloride or adaman-
tane-1-carbonyl chloride) in acetone (160 ml) was added
anhydrous potassium carbonate (1.58 g). The reaction was
stirred overnight at room temperature. The red reaction
mixture was poured into dichloromethane (50 ml) and washed
with distilled water (3 Â 100 ml). The organic layer was
isolated, dried with sodium sulfate and evaporated under
reduced pressure. The crude product obtained was purified
by column chromatography using silica gel as adsorbent and
chloroform as eluent.
2-Hydroxy-3-methyl-1,4-naphthoquinone. This was synthe-
sised from 2-hydroxy-1,4-naphthoquinone, hydrogen peroxide
and sulfuric acid using the procedure recorded by Kaushik.15
Derivative VI: 1,4-Naphthalenedione, 2-[[(2-methylphenyl)-
sulfonyl]oxy]. Yield 1.15 g (27%) of 1,4-naphthalenedione, 2-
[[(2-methylphenyl)sulfonyl]oxy] (as the second fraction), as a
bright yellow solid.
1H NMR (d-CDCl3) d 2.81 (s, 3 H), 6.94 (s, 1 H), 7.34–7.38
(t, 1 H), 7.43–7.45 (d, 1 H), 7.57–7.60 (t, 1 H), 7.74–7.77 (m, 2
H), 7.98–8.00 (d, 1 H), 8.06–8.08 (m, 2 H). 13C NMR (d-
CDCl3) d 20.48, 125.22, 126.31, 126.49, 126.88, 130.18, 130.61,
131.50, 132.95, 133.90, 134.14, 134.55, 134.90, 139.51, 151.25,
178.11, 184.10. CIMS (m/z) 346.1 [M + NH4]+. Anal. Calcd
for C17H12O5S: C, 62.19; H, 3.68. Found: C, 62.34; H, 3.37%.
Derivative XI: 3-Methyl-2-(tricyclo[3.3.1.13,7]decane-1-car-
boxylic acid), 1,4-naphthoquinone ester. Yield 0.26 g (67.9%)
of 3-methyl-2-(tricyclo[3.3.1.13,7]decane-1-carboxylic acid),
1,4-naphthoquinone ester (as the second fraction), as a bright
yellow solid.
1H NMR (d-CDCl3) d 1.78 (brs, 6 H), 1.94 (brs, 3 H), 2.07
(s, 3 H), 2.11–2.13 (brs, 6 H), 7.72–7.73 (m, 2 H), 8.06–8.08 (d,
1 H), 8.08–8.10 (d, 1 H). 13C NMR (d-CDCl3) d 9.70, 27.80,
36.33, 38.73, 41.34, 126.49, 126.61, 130.91, 132.06, 133.74,
133.95, 135.93, 151.48, 174.62, 178.01, 184.77. EIMS (m/z) 351
(M+). Accurate CI+ (ammonia) (m/z) 368.185 [M + NH4]+.
Anal. Calcd for C21H20O4: C, 75.41; H, 6.33. Found: C, 75.31;
H, 6.21%.
Derivative VII: 1,4-Naphthalenedione, 2-[[(3-methylphenyl)-
sulfonyl]oxy]. Yield 0.51 g (27%) of 1,4-naphthalenedione, 2-
[[(3-methylphenyl)sulfonyl]oxy] (as the second fraction) as a
bright yellow solid.
1H NMR (d-CDCl3) d 2.47 (s, 3 H), 6.91 (s, 1 H), 7.47–7.54
(m, 2 H), 7.75–7.77 (m, 2 H), 7.83 (brs, 1 H), 7.85 (brs, 1 H),
8.07–8.09 (m, 2 H). 13C NMR (d-CDCl3) d 21.31, 125.60,
125.68, 126.52, 126.94, 128.67, 129.28, 130.69, 131.59, 134.18,
134.55, 135.13, 135.82, 139.94, 151.38, 178.15, 184.16. Accu-
rate LCMS (m/z) 346.074 [M + NH4]+. Anal. Calcd for
C17H12O5S: C, 62.19; H, 3.68. Found: C, 62.05; H, 3.40%.
Acknowledgements
The authors thank the EPSRC for financial support and the
Mass Spectrometry Centre at Swansea University for the mass
spectral characterisation of the quinone derivatives.
References
1 M. Valko, C. J. Rhodes, J. Moncol, M. Izakovic and M. Mazur,
Chem.-Biol. Interact., 2006, 160, 1.
2 J. M. Estrela, A. Ortega and E. Obrador, Crit. Rev. Clin. Lab. Sci.,
2006, 43, 143.
3 D. M. Townsend, K. D. Tew and H. Tapiero, Biomed. Pharmac-
other., 2003, 57, 145.
4 A. Digga, S. Gracheva, C. Livingstone and J. Davis, Electrochem.
Commun., 2003, 5, 732.
Derivative VIII: 1,4-Naphthalenedione, 2-[[(4-methylphenyl)-
sulfonyl]oxy]. Yield 0.45 g (23.8%) of 1,4-naphthalenedione, 2-
[[(4-methylphenyl)sulfonyl]oxy] (as the second fraction), as a
bright yellow solid.
1H NMR (d-CDCl3) d 2.47 (s, 3 H), 6.91 (s, 1 H), 7.36–7.40
(d, 2 H), 7.75–7.77 (m, 2 H), 8.06–8.08 (m, 2 H). 13C NMR
(d-CDCl3) d 21.93, 125.82, 126.63, 127.07, 128.64, 130.06,
5 P. C. White, N. S. Lawrence, J. Davis and R. G. Compton, Anal.
Chim. Acta, 2001, 447, 1.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 1718–1724 | 1723

View Article Online
6 O. Nekrassova, P. C. White, S. Threlfell, G. Hignett, A. J. Wain,
N. S. Lawrence, J. Davis and R. G. Compton, Analyst, 2002, 127,
797–802.
7 S. Gracheva, C. Livingstone and J. Davis, Anal. Chem., 2004, 76,
3833.
8 H. Park, J. S. Park and T. B. Shim, J. Electroanal. Chem., 1997,
13 C. Escolastico, M. D. S. Maria, R. M. Claramunt, M. L. Jimeno,
I. Alkorta, C. Foces-Foces, F. H. Cano and J. Elguero, Tetrahe-
dron, 1994, 50, 12489.
14 S. Bittner, Amino Acids, 2006, 30, 205.
15 V. K. Srivastava, N. K. Kaushik and B. Khera, Synth. React.
Inorg. Met.-Org. Chem., 1987, 17, 15.
438, 113.
9 V. Kertesz, J. Q. Chambers and A. N. Mullenix, Electrochim. Acta,
16 J. M. Estrela, A. Ortega and E. Obrador, Crit. Rev. Clin. Lab. Sci.,
2006, 43, 143.
1999, 45, 1095.
10 H. S. Kim, T. D. Chung and H. Kim, J. Electroanal. Chem., 2001,
17 W. Droge, Curr. Opin. Clin. Nut. Metab. Care, 2006, 9,
190.
498, 209.
11 M. Salas, M. Gomez, F. J. Gonzalez and B. Gordillo, J. Electro-
18 C. A. Weber and M. E. Erns, Ann. Pharmacother., 2006, 40,
935.
anal. Chem., 2003, 543, 73.
12 G. Pastor Moreno and D. J. Riley, Electrochem. Commun., 2002, 4,
218.
19 M. Valko, C. J. Rhodes, J. Moncol, M. Izakovis and M. Mazur,
Chem.-Biol. Interact., 2006, 160, 1.
20 J. C. Kern and J. P. Kehrer, Front. Biosci., 2005, 10, 1727.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
1724 | New J. Chem., 2006, 30, 1718–1724