Absorption spectra of isomeric OH adducts of 1,3,7-trimethylxanthine

Article abstract of DOI:10.1021/jp963272q

The reactions of OH^?, O^?-, and SO₄^?- with 1,3,7-trimethylxanthine (caffeine) were studied by pulse radiolysis with optical and conductance detection techniques. The absorption spectra of transients formed in OH? reaction in neutral solutions exhibited peaks at 310 and 335 nm, as well as a broad absorption maximum at 500 nm, which decayed by first-order kinetics. The rate (k = (4.0 ± 0.5) × 10⁴ s^-1) of this decay is independent of pH in the range 4-9 and is in agreement with that determined from the conductance detection (k = 4 × 10⁴ s^-1). The spectrum in acidic solutions has only a broad peak around 330 nm with no absorption in the higher wavelength region. The intermediates formed in reaction of O^?- absorb around 310 and at 500 nm, and the first-order decay at the latter wavelength was not seen. The OH radical adds to C-4 (X-4OH^?) and C-8 (X-80H^?) positions of caffeine in the ratio 1:2 as determined from the redox titration and conductivity measurements. H abstraction from the methyl group is an additional reaction channel in O^?- reaction. Dehydroxylation of the X-4OH^? adduct occurs, whereas the X-8OH^? adduct does not undergo ring opening. The rate constant for addition of O₂ to X-4OH^? is estimated to be ～1 x 10⁹ M^-1 s^-1, whereas it is unreactive toward X-8OH^?. The spectrum obtained for OH^? reaction in oxygenated solutions is similar to that observed in SO₄^?- reaction in basic solutions. The radical cation of caffeine formed from its reaction with SO₄^?- (λ_max = 320-340 nm) is hydrolyzed in basic solutions to yield the X-8OH^? adduct. The molar absorptivities of the X-8OH^? and the X-4OH^? adducts at 300 and 335 nm are 6500 and 5300 M^-1 cm^-1, respectively. The yield of 1,3,7-trimethyluric acid in OH^? reaction in the absence of O₂ (28%) is reduced by more than 50% in its presence. The differences in the mechanism of OH^? reaction with caffeine and its isomer 1,3,9-trimethylxanthine (isocaffeine) are discussed.

Full text of DOI:10.1021/jp963272q

J. Phys. Chem. A 1997, 101, 2953-2959
2953
Absorption Spectra of Isomeric OH Adducts of 1,3,7-Trimethylxanthine1
M. S. Vinchurkar and B. S. M. Rao*
Department of Chemistry, UniVersity of Pune, Pune 411 007, India
H. Mohan and J. P. Mittal*
Chemistry DiVision, Bhabha Atomic Research Centre, Mumbai 400 085, India
K. H. Schmidt and C. D. Jonah*
Chemistry DiVision, Argonne National Laboratory, Argonne, Illinois 60439
X
ReceiVed: October 21, 1996; In Final Form: January 3, 1997
•
•-
•-
The reactions of OH , O , and SO
4
with 1,3,7-trimethylxanthine (caffeine) were studied by pulse radiolysis
•
with optical and conductance detection techniques. The absorption spectra of transients formed in OH reaction
in neutral solutions exhibited peaks at 310 and 335 nm, as well as a broad absorption maximum at 500 nm,
4
-1
which decayed by first-order kinetics. The rate (k ) (4.0 ( 0.5) × 10 s ) of this decay is independent of
4
pH in the range 4-9 and is in agreement with that determined from the conductance detection (k ) 4 × 10
-1
s ). The spectrum in acidic solutions has only a broad peak around 330 nm with no absorption in the higher
•
-
wavelength region. The intermediates formed in reaction of O absorb around 310 and at 500 nm, and the
•
first-order decay at the latter wavelength was not seen. The OH radical adds to C-4 (X-4OH ) and C-8
•
(
X-8OH ) positions of caffeine in the ratio 1:2 as determined from the redox titration and conductivity
•-
measurements. H abstraction from the methyl group is an additional reaction channel in O reaction.
•
•
Dehydroxylation of the X-4OH adduct occurs, whereas the X-8OH adduct does not undergo ring opening.
The rate constant for addition of O
•
9
-1 -1
2
to X-4OH is estimated to be ∼1 × 10 M s , whereas it is unreactive
•
•
toward X-8OH . The spectrum obtained for OH reaction in oxygenated solutions is similar to that observed
•
-
•-
in SO
4
reaction in basic solutions. The radical cation of caffeine formed from its reaction with SO
4
(λmax
•
)
320-340 nm) is hydrolyzed in basic solutions to yield the X-8OH adduct. The molar absorptivities of
•
•
-1
-1
the X-8OH and the X-4OH adducts at 300 and 335 nm are 6500 and 5300 M cm , respectively. The
•
yield of 1,3,7-trimethyluric acid in OH reaction in the absence of O
2
(28%) is reduced by more than 50%
in its presence. The differences in the mechanism of OH reaction with caffeine and its isomer
,3,9-trimethylxanthine (isocaffeine) are discussed.
•
1
-
Introduction
Radiation chemical studies on the reactions of primary
the rate constant values for the OH elimination and the ring
6
5 -1
opening reactions are 2.2 × 10 and 2.3 × 10 s . Furthermore,
the rate of the dehydration process as compared to the ring
opening reaction has been found to be strongly influenced by
the substituent at C-6.
-
•
•
radicals of water (e aq, OH , and H ) and secondary radicals
•-
•-
•-
(
e.g., SO4 ,Br2 , and O ) derived from them with pyrimidines
and purines are of current interest due to their importance in
Xanthine and its methyl derivatives are structurally similar
to purines, but the six-membered ring closely resembles that of
uracil. Thus, the methylated xanthines form an interesting class
of compounds for examining the effect of the nature and position
of the substituents on the rates of transformation processes of
the adducts formed in the reaction with primary radicals of water
radiolysis. We have undertaken a comprehensive study of reac-
tions of these radicals with xanthine derivatives and have pre-
the understanding of DNA radiation chemistry (see refs 2-4
-
•
for recent reviews). Both e aq and OH show high reactivity
9
10
-1 -1
4,5
(
k ) 10 - 10
M
s ) toward these compounds.
Of
particular importance are the studies carried out on the reactions
6
-8
of oxidizing radicals with the derivatives of guanine
and
adenine.9-11
10
It has been reported that at least two different
•
isomeric OH adducts are formed in the reaction of OH with
adenine derivatives. The additives have been identified as the
1
5
-
•
•
viously reported the reactions of e
ylxanthines.
aq
with di- and trimeth-
radicals formed by OH addition to C-4 (A-4OH ) and C-8
•
(
A-8OH ) positions of these compounds.
In contrast to pyrimidines,1
2-14
the A-4OH and A-8OH
•
•
•
•-
This paper deals with the study of reactions of OH , O ,
9-11
•-
radical adducts of purines are shown
to undergo unimo-
and SO4 with caffeine (structure 1, Scheme 1). The double
lecular transformation reactions involving dehydration of the
former and ring opening of the latter. The opening of the
imidazole ring of the A-8OH adduct of purines was manifested
bond in the imidazole ring of this compound is between C-8
and N-9, unlike other purine derivatives. This structural
difference, combined with the possibility to employ conductance
detection in addition to optical absorption techniques to
•
in the first order increase of absorption around 350 nm, while
the decrease in the higher wavelength region was attributed to
investigate the nature of the transformation reactions as in the
-
•
9,10
the elimination of OH from the A-4OH adduct. In the case
case of other fully substituted adenines,
makes the study
6
6
10
of fully alkylated adenines such as N ,N ,9-trimethyladenine,
interesting. While our study has been in progress, results from
the oxidation of methyl derivatives of xanthine were re-
X
16,17
Abstract published in AdVance ACS Abstracts, March 1, 1997.
ported.
In their work on isocaffeine, an isomer of caffeine
S1089-5639(96)03272-0 CCC: $14.00 © 1997 American Chemical Society

2954 J. Phys. Chem. A, Vol. 101, No. 16, 1997
Vinchurkar et al.
SCHEME 1
(
structure 2, Scheme 1), Viera and Steenken17 have shown that
saturated solutions were used to measure the OH adduct spectra
in the presence of O2.
•
the X-4OH adduct, even if it is formed, did not undergo
•
•-
dehydroxylation and that the X-8OH adduct underwent a
The reaction of SO4 was studied in N2 saturated aqueous
-
3
-2
unimolecular transformation due to the ring opening in agree-
solutions of caffeine (10 M) containing K2S2O8 (1.5 × 10
9
ment with the results obtained with DNA purine derivatives.
In contrast, it will be shown that the OH attack in caffeine
leads to the formation of both X-4OH and X-8OH adducts
and that while OH elimination takes place from the former,
M) and 0.2 M tert-butyl alcohol.
•
The OH radicals are scavenged by tert-butyl alcohol, while
•
•
e-aq react quantitatively with S2O8 to produce the SO4
2-
•-
-
radical anion (reactions 3 and 4),
the latter does not undergo a ring-opening reaction. This study
demonstrates the usefulness of radiation chemical methods in
obtaining the distinct absorption spectra of these two adducts.
•
•
OH + (CH ) COH f H O + CH (CH ) COH (3)
3
3
2
2
3 2
S2O8 + e (H) f SO4•- + SO4 (HSO )
2
-
-
2-
-
4
(4)
aq
Experimental Section
Pulse radiolysis experiments for optical absorption measure-
ments were carried out using 7 MeV electron pulses (pulse width
Caffeine (SD Fine) and other chemicals (Qualigens) used in
this study were of high purity and were used as received. The
solutions were prepared in water obtained from the Millipore
50 ns) from a linear accelerator at the Bhabha Atomic Research
•
Centre, Mumbai. KSCN dosimeter was used in the optical pulse
Milli-Q purification system. The reaction of OH was studied
-
3
radiolysis. The dose per pulse was between 10-15 Gy. The
in N2O saturated aqueous solutions of caffeine (10 M) where
18
-
•
details of this facility have been described elsewhere. The
e
aq is converted into OH (reactions 1 and 2),
1
9,20
1
6 MeV electron LINAC facility
(pulse width 4-20 ns) at
the Argonne National Laboratory was employed for the pulse
conductivity experiments.
H O Df e- , OH , H , H , H O , H
•
•
+
(1)
(2)
2
aq
2
2
2
Separation of the products formed in the reactions of OH•
e- + N O f OH + OH + N
•
-
•-
aq
2
2
and SO4 with caffeine under steady-state conditions was done
by reverse phase HPLC technique (Perkin Elmer Series 10
Liquid Chromatograph, Nucleosil-5-C18 column, 40% acetoni-
trile in water as an eluent under isocratic conditions). A diode
array detector was used for optical detection.
The reaction of O•- was studied at pH 13 (OH h O + H+,
pKa ) 11.9) where >90% OH is converted and thus the radical
anion is essentially the reacting species. N2O:O2 (4:1 v/v)
•
•-

OH Adducts of 1,3,7-Trimethylxanthine
J. Phys. Chem. A, Vol. 101, No. 16, 1997 2955
k ) 1.3 × 10 M s-1 for N ,N ,9-trimethyladeninium to 8.4
8
-1
6
6
9
-1 -1
6
6
×
10 M
s
for N ,N ,9-trimethyladenine. The second-order
rate constant value obtained in the case of caffeine is comparable
6
6
to that found for N ,N ,9-trimethyladenine, and this increase in
reactivity can be attributed to larger electron-density due to the
electron donating methyl groups. However, the rate of reaction
•
-
of O with caffeine was reduced by nearly an order of
9
-1 -1
•
magnitude (k = 1 × 10 M s ) relative to the OH reaction
and represents the diffusion-controlled reaction. Such a decrease
•
-
in the reactivity of O with the substrate was also seen in the
case of uridine and adenosine systems. For instance, the rate
2
2
•
•-
9
constants of OH and O with uridine are 5 × 10 and 0.7 ×
9
-1 -1
1
0 M s , respectively. Since there is no ionizable H in
•
•-
caffeine, the variation in the reactivity between OH and O
must be attributed to differences in the reaction pathways
2
Figure 1. Transient absorption spectra measured in N O saturated
(
section f, below).
The rate of reaction of SO4•- with caffeine was measured by
-
3
neutral solutions of caffeine (1 × 10 M) at 2 (O) and 20 µs (b) after
the pulse. (0) This plot is drawn from the difference in the fully grown
•
-
spectra recorded in the absence and the presence of O
2
at 7 µs. Dose/
following the decay of the SO4 absorption at 460 nm, and
the bimolecular rate constant from this decay was found to be
4
pulse ) 15 Gy Inset: Traces showing the (a) buildup at 335 nm
-
4
[
caffeine] ) 4 × 10 M and (b) first-order decay at 500 nm and the
9
-1 -1
9
-1
.5 × 10 M
s
at neutral pH. A value of 4.7 × 10 M
•+
corresponding growth of TMPD at 610 nm on pulsing N
neutral solutions [caffeine] ) 1 × 10 M. Dose:pulse was 12 Gy.
2
O-saturated
-1
s
was obtained from the measurement of the growth of
-
3
•-
absorption of the intermediates at 320 nm. The rates for SO4
reaction are found to be same in both acidic (pH 3) and basic
(pH 10) solutions.
TABLE 1: Various Rate Constants and Yields Obtained in
•
•-
Reactions of OH and SO
4
with Caffeine
optical
work
b. Absorption Spectra of OH Adducts. i. Deoxygenated
and Oxygenated Solutions. The transient absorption spectra for
reaction
conductance detection
•
the reaction of OH with caffeine in neutral solutions (pH ∼ 7)
caffeine (X)
•
9 a
9
OH + X
8.5 × 10
were recorded in the range 280-650 nm. The spectrum
exhibited peaks at 310 and 335 nm with a broad maximum
around 500 nm. Figure 1 shows the fully grown absorption
spectrum at 2 µs after the pulse. The time-resolved spectrum
recorded at 20 µs revealed a significant decrease in absorbance
around 500 nm. The rate of this decay was found to be first
-
9 a
(
X-4)•+ + OH 3 × 10
•
-
9
O
+ X
1 × 10
•
-
SO
4
+ X
4.5 × 10
•
9
X-4OH + O
2
1 × 10
•
-
•-
2
X-8OH + OH h (X-8O) + H O
•
+
9
(
(
X-4) + TMPD 1.5 × 10
•
•+
9
X(-H)) +MV
2 × 10
b
9 a
4 c
dimethyladenosine (A)
k
f
) 6.0 × 10
k
b
) 3.0 × 10
4
-1
order, kobs ) (4.0 ( 0.5) × 10 s , and nearly constant in the
•
8
A-4OH + O
2
4 × 10
•
10
pH range 4-9.
A-8OH + O
2
1.6 × 10
The observed behavior is in contrast to that observed11 with
its isomer isocaffeine where only the bimolecular decay of
absorption in the higher wavelength region (400-450 nm) and
a first-order increase in absorbance at 330 nm with k ) 1.7 ×
4
c,d
trimethyladenine (A)
decay at 500 nm (4 ( 0.5) × 10
-
4 c,d
OH elimination 4 × 10
•
8
A-4OH + O
2
5 × 10
product yields (%OH•)
4
-1
1
0 s were seen. A typical trace depicting the rate of the
X-4OH•
8-OH-X
decay at 500 nm in neutral solutions of caffeine is shown in
the inset of Figure 1. However, this rate was found to decrease
above pH ) 9 and this variation is given in Figure 2A. For
3
5f
33e
28g
12h
a
k (M s-1). Reference 10. k (s-1). d pH 4-9. Yields of
-1
b
c
4
-1
example, kobs values are e1 × 10 s beyond pH 10. Our k
•
e
f
X-4OH from the conductivity measurements and buildup of
value for the decay of absorption at 500 nm is much lower than
•
+
g
TMPD at 610 nm. Yields of 1,3,7 trimethyluric acid from N
2
O and
9
5
6 -1
those observed for methyl derivatives of adenine (10 -10 s );
h
N
2
O:O
2
saturated solutions.
9
but it is comparable to that found for hypoxanthine where the
electron-withdrawing carbonyl group is at the C-6 position.
Thus, this is in accord with the observed influence of the C-6
Results and Discussion
10
a. Determination of Rate Constants. The reaction of OH•
with caffeine was studied by measuring the transient absorption
from the pulse radiolysis of N2O-saturated neutral solutions of
substituent on the rate of this reaction.
•
The absorption spectrum obtained for OH reaction with
caffeine in N2O:O2 (4:1 v/v) saturated solutions is different from
that obtained in deoxygenated solutions. Nor is it similar to
-
3
caffeine (10 M). The rate of the reaction was monitored by
the buildup in the absorption of the transients at 335 nm (inset
Figure 1) and the bimolecular rate constant obtained from the
•
-
that obtained in SO4 reaction in neutral solutions of caffeine.
The fully grown spectrum recorded in the presence of O2 at 7
µs after the pulse exhibited a single peak at 300 nm, while the
absorption above 335 nm was scavenged to an extent of more
than 70% by O2 compared to that observed in deoxygenated
solutions (Figures 1 and 3). This spectrum remained unchanged
even up to 50 µs after the pulse. This behavior is in contrast
to that observed with its isomer (isocaffeine) where the inhibition
of the delayed increase in absorbance at 335 nm due to the ring
-3
plot of kobs Versus [solute] in the range (0.2-1) × 10 M (data
9
-1 -1
not shown) at neutral pH is 8.5 × 10 M s . This is in
9
-1 -1
reasonable agreement with the value (6.9 × 10 M s )
reported21 earlier. The rate constants obtained (accurate to
within (10%), for this and other reactions in this work, are
listed in Table 1.
The rate of the reaction of OH radical with purines was
reported6 to be influenced by the nature and position of the
substituents. For example, the second-order rate constants have
been shown10 to vary by more than an order of magnitude with
-8
•
opening of the X-8OH adduct in the presence of O2, was
17
•
noticed. Thus, the oxidation of the X-8OH adduct predomi-
nates over the ring opening reaction.

2956 J. Phys. Chem. A, Vol. 101, No. 16, 1997
Vinchurkar et al.
Figure 4. Absorption spectra for O•- reaction recorded in N
O-
2
-
3
saturated basic solutions (pH g 13) of caffeine (1 × 10 M) at 2 (O)
and 18 µs (b) after the pulse. Dose:pulse ) 15 Gy. Inset: (a) Buildup
•
+
at 300 nm and (b) the rate of formation of MV at 605 nm. Dose:
pulse ) 12 Gy.
Figure 2. (A) Dependence of the rate constants for the decay at 500
nm (b) and OH elimination (9) from conductance measurements on
-
pH. (B) Dependence of the rate (kobs) of increase in absorption at 610
-
3
nm (b) on [TMPD]. For both (A) and (B), [caffeine] ) 1 × 10 M.
Figure 3. Transient absorption spectra recorded in N
2
-saturated
(1.5
10 M) and 0.2 M tert-butyl alcohol at 18 µs (0) after the pulse
and in N O-O (4:1 v/v) saturated neutral solutions at 7 µs (9) after
the pulse. This spectrum is normalized to a G value of 3.3. Inset:
-
3
solutions (pH ) 10) of caffiene (1 × 10 M) containing K
2 2 8
S O
-
2
×
2
2
•-
Transient absorption spectra recorded for the reaction of SO
4
at neutral
pH at 2 (0) and 18 µs (9) after the pulse. Dose:pulse ) 15 Gy.
ii. Acidic and Basic Solutions. The transient absorption
spectrum recorded in the reaction of OH with caffeine in acidic
2
Figure 5. Conductance changes recorded in N O-saturated solutions
•
-3
of caffeine (1 × 10 M) at (A) 1 ) pH 8.9 and 2 ) pH 4.9, (B) pH
solutions showed a broad maximum around 335 nm with no
absorption in the higher wavelength region. This spectrum is
9.6, and (C) pH 10.3. The figures are all normalized to a dose of ∼1
-
Gy. Inset (B): pH 9.3. Inset (C): Dependence of kobs on [OH ].
•
-
similar to that measured in the SO4 reaction at neutral pH.
The nature of transient absorption spectrum obtained in the
-
[OH ] between pH 4 and 9; but decreases at higher pH. The
•
-
-
1
reaction of O with caffeine was similar to that observed in
yield of OH was estimated to be about /3 OH from the growth
of conductivity signal at pH 8.9.
•
the OH reaction with maxima at 300, 340, and 500 nm, though
•
-
the intensities of absorption were reduced by nearly 40%
d. Reaction of SO4 . The transient absorption spectrum
recorded for reaction of SO4 with caffeine at neutral pH
•
•-
compared to those observed for OH reaction. Furthermore,
the first-order decay of absorption at 500 nm as observed in
neutral solutions was not seen (Figure 4).
exhibited a broad peak with maximum around 320-330 nm
(inset of Figure 3). This spectrum is different from that obtained
with isocaffeine where the peak was centered at 370 nm. This
blue shift indicates that the radical cation of the former is
relatively more stable and does not get hydrolyzed in neutral
solutions. Though the initial absorption spectrum measured in
basic solutions (pH =10) immediately after the pulse is similar
to that recorded at neutral pH, it was finally transformed to a
spectrum having a peak at 300 nm (Figure 3). However, the
spectral nature of the transients formed in acidic solutions (pH
= 3) is similar to that observed in neutral solutions as well as
c. Conductivity Measurements. Pulse conductivity experi-
ments were carried out in both basic and acidic media. As
shown in Figure 5A, the conductivity traces clearly indicate the
-
expulsion of OH in solutions at pH, near neutral. In acid
solutions, there is a clear loss of conductivity arising from the
-
+
reaction of the ejected OH with H in the bulk of the solution
while there is an increase in conductivity in basic solutions.
The rate of this increase was evaluated to be approximately 4
4
-1
×
10 s . Further, this rate was found to be independent of

OH Adducts of 1,3,7-Trimethylxanthine
J. Phys. Chem. A, Vol. 101, No. 16, 1997 2957
•
•+
in OH reaction in acidic medium. This finding suggests that
The radical cation, (X-4) , is stabilized by the +I effect of
the same transient species was formed under both these
condition.
the methyl group that is expected to be oxidizing in nature. Thus,
about 35% OH corresponding to the yields of TMPD in
optical and OH in conductivity experiments (reaction 10) seems
to attack at C-4 and the remainder adds to the C-8 position of
caffeine. The yield of the X-8OH adduct is higher than that
of the X-4OH adduct owing to the electrophilic nature of the
OH radical. Furthermore, being a nitrogen-centered radical, it
is relatively more stable and does not undergo ring opening as
is observed with other purine derivatives and with its isomer
isocaffeine. The lack of attack at C-4 In the case of isocaffeine
is possibly due to the steric hindrance from the adjacent methyl
groups at the N-3 and N-9 positions.
•
•+
-
e. Redox Reactions of Transient Adducts. The oxidant
2+
methylviologen (MV ) and the reductant N,N,N′,N′-tetramethyl-
p-phenylenediamine (TMPD) were used to differentiate between
the reducing and oxidizing radicals produced in reactions of
OH and O with caffeine. The reduced form (MV ) of the
oxidant has two well-defined peaks at 395 and 605 nm. The
oxidized form of TMPD (TMPD ) has the absorption maxima
at 335, 565, and 610 nm. However, the rate for the oxidation
of the transient adducts was monitored at 605 nm (ꢀ ) 12 800
•
•
•-
•+
•
+
17
-
1
1
-1
2+
M
M
cm ) in the case of MV and at 610 nm (ꢀ ) 12 000
cm ) for TMPD as the other maxima are not suitable
-
-1
The OH adduct spectrum recorded in the presence of O2 at
owing to the absorption of other transients formed at these
wavelengths.
7 µs after the electron pulse (Figure 3) matches very well with
the absorption spectrum measured in SO4 reaction in basic
•
-
solutions (pH 10) at 18 µs after the pulse (i.e., after the reaction
The redox experiments, were done over the range (2 - 8) ×
-
-5
-3
of OH with the radical cation is complete). The radical cation
10
M [TMPD] with solutions of 1 × 10 M caffeine saturated
2-
formed after SO4 elimination (reaction 7) gets converted much
with N2O, where all the OH radicals will essentially react with
•
9
-1 -1
faster in basic solutions to give the more stable X-8OH adduct
caffeine (k ) 8.5 × 10 M s ). The dose per pulse was
(
reaction 13).
kept at 12 Gy to minimize the effect of bimolecular radical
recombination reactions. The formation of TMPD was
observed showing that oxidizing radicals were indeed formed.
•
+
The spectrum obtained in SO4•- reaction in basic solutions
is assigned to this adduct. Since the absorption at 500 nm is
The dependence of the rate of formation of TMPD•+ on
completely scavanged by O in N O:O (4:1 v/v) ([O ] ) 0.25
2
2
2
2
mM) solutions within 7 µs, the bimolecular rate constant for
[
TMPD] is shown in Figure 2B. The kobs reaches a plateau value
•
4
-1
-5
the reaction of O with X-4OH is estimated to be about 1 ×
of 3 × 10 s at 4 ×10 M TMPD, which matches reasonably
2
9
-1 -1.
1
0 M
s
This rate constant is slightly higher than those
well with that obtained for the rate of decay at 500 nm. This
behavior indicates that the radical cation formed by OH
elimination from the X-4OH is only responsible for the
-
observed for the corresponding adducts of dimethyladenosine
8
-1 -1 10
•
and trimethyladenine (4 - 5) × 10 M s ). However, a
reverse trend was seen in the reactivity of O2 with C-8OH•
adducts of these compounds. For example, the k value for O2
oxidation and that the neutral radical itself is unable to oxidize
TMPD. If it were so, the kobs Versus [TMPD] plot should have
yielded two lines with different slopes corresponding to the rates
of these two reactions. The bimolecular rate constant estimated
from the linear part of the plot of kobs Versus [TMPD] is ∼1.5
•
reaction with A-8OH adduct of dimethyladenosine was found
1
0
-1 -1
to be 1.6 × 10
M
s
(Table 1). The value for isocaffeine
7
-1 -1
is at least g7 × 10 M
s
based on the observed inhibition
4
-1
9
-1 -1
5
-1
of the ring-opening reaction (k ) 1.7 × 10 s ) in N O:O
×
10 M s . Similar behavior with kplateau ∼ 1.2 × 10 s
2
2
10b
(4:1) solutions. The corresponding rates in the case of caffeine
must be much lower. This is due to the delocalization of the
was noticed in the case of adenine, which was explained on
the basis of the reaction of TMPD with the neutral radical
•
•
unpaired spin density more on a heteroatom in the X-8OH
formed from dehydration of A-4OH and its reactivity with
•
•
than in X-4OH adduct of caffeine. Similar structures of
A-4OH adduct itself was seen to be low. However, the larger
•
10
•
A-4OH adducts of methyl derivatives of adenine were shown
oxidizing strength observed for the A-4OH adduct of adenine-
1
0
to be responsible for their low reactivity with O . The resulting
5
′-triphosphate and 2′-deoxyadenosine was attributed to the
2
spectrum obtained from the difference between the OH adduct
electron-withdrawing ribose phosphate group at N-9.
spectra recorded in deoxygenated and oxygenated solutions must
The electron-donating methyl groups in caffeine make it
relatively less reactive toward the reductant. The yield of the
•
then represent that of X-4OH .
The molar absorptivities of X-8OH and X-4OH adducts
•
•
oxidizing radicals calculated from the maximum absorbance at
-
1
-1
•
are found to be 6500 and 5300 M cm , respectively. These
are estimated at 300 and 335 nm from the corresponding yields
6
10 nm was about 35% OH . Complementary experiments
2+
using MV as a scavenger were done to estimate the yield of
•
•
•
•
+
of OH attack (60% and 35% OH ). The X-8OH adduct
the reducing radicals. The yield of MV was marginal (e 5%
•
spectrum of caffeine matches reasonably well with the spectra
OH ) indicating that the formation of the reducing adduct
•
measured for the C-8H adduct spectra (λmax ) 305 nm) of
radicals is negligible. In contrast, a reverse trend was observed
1
5
23
-
•
-
•+
caffeine and guanosine obtained in the reaction of e with
in the case of O reaction, where the yields of MV and
aq
•
+
•
these compounds.
TMPD were about 40% and ∼5% OH , respectively.
•
In acidic solutions, the X-8OH adduct undergoes acid
f. Reaction Mechanism. In caffeine, the three possible sites
•
catalyzed water elimination resulting in the formation of its
radical cation (reverse of reaction 13). This is in accord with
the observed similarity of the absorption spectra induced by
of OH addition are at C-4 (reaction 5, Scheme 1), C-5, and
C-8 positions (reaction 6). The attack at C-5 position is not
•
likely since the X-5OH adduct will be structuraly unstable.
•
•-
OH in acidic solution and SO4 at neutral pH.
The unimolecular decay observed at 500 nm in neutral solutions
is due to the OH- elimination from the X-4OH adduct
•
O•- reacts by H-abstraction, besides, addition to the double
bonds and abstraction from the CH3 group at N-1 is more likely
because of the electron-withdrawing nature of the neighboring
(
reaction 9a) leading to the formation of the radical cation in
9
,10
analogy with the behavior reported
in adenine derivatives.
-
•
Confirming this are the conductivity data (the yield of OH
and its rate of formation, Table 1), which are in excellent
agreement with the corresponding values obtained in our optical
-CdO. This carbon-centered radical ( X(-H)) formed after
H-abstraction (reaction 8) will be reducing in nature and reacts
with the oxidant methylviologen, as can be seen from the trace
-
2+
-5
•+
absorption measurements supporting the proposed OH elimina-
recorded ([MV ] ) 5 × 10 M) for the absorption of MV
•
tion from X-4OH .
(inset Figure 2) at 605 nm. The second-order rate constant and

2958 J. Phys. Chem. A, Vol. 101, No. 16, 1997
Vinchurkar et al.
SCHEME 2
•
+
the yield of MV formed were estimated from this trace are 2
millisecond scale at pH 10.6 (Figure 5C) that decreases the
conductivity and the increased size of this total signal suggests
9
-1 -1
•-
×
10 M
s
and 40% O , respectively. The remaining 60%
•-
•
•
O
is assigned to the formation of the radical anion from the
that both X-4OH and X-8OH adducts are reacting and
•
-
•
-
•
-
-
addition of O at C-4 (X -4-O ) and C-8 (X -8-O )
destroying OH . The mechanism for this reaction is not clear
•
-
positions. Such an addition of O to halotoluenes besides
at present
H-abstraction from the CH3 was reported24 by us.
g. Product Studies. Aqueous solutions containing 1 mM
60
The conductivity data obtained in basic solutions (8.9 e pH
e 10.5) are shown in Figures 5A-C. In mildly basic solutions
pH e 9), the initial growth of conductivity corresponding to
caffeine were Co γ-irradiated to obtain the product distribution
•
•-
in reactions of OH with and without O2 and in SO4 reaction.
The doses were such that the conversion was below 15%
(
-
•
OH formation decayed to the base line by a second-order
assumming G(OH ) ) 6 (G is the yield/100 eV energy
9
-1 -1
kinetics (k ) (3.0 ( 0.2) × 10 M s ). This may be due to
absorbed). The HPLC analysis has shown the formation of only
one major product with a retention time of 3.6 min in all of the
-
the reaction of OH either with the radical cation formed from
•
•-
the X-4OH adduct to form a new product. The other
cases except in the SO4 reaction in basic solutions (pH∼10).
-
•
25
possibility of OH reaction with the X-8OH adduct is not
Recently, it has been shown that the Fenton type oxidation of
•
likely as its yield is about 70% OH , and one would expect a
caffeine leads to the formation of products arising from
hydroxylation, demethylation, and C(8)-N(9) bond scission.
The product formed in caffeine is attributed to 1,3,7-trimethy-
-
net loss of OH in that case. It is suggested that the radical
•+
-
cation (X-4) formed (reaction 9a, Scheme 1) reacts with OH
•
10
to give X-8OH (reaction 9b, Scheme 1).
luric acid (8-OH-X) in accord with the published work on
The results in more basic solutions (9.5 e pH e 10.5) show
the formation of 8-hydroxyadenines from the oxidation of the
-
•
a net loss of OH immediately after the pulse, suggesting a
corresponding A-8OH adduct and based on the similarity of
new reaction that consumes OH- compared to the growth
followed by decay in conductivity observed in less basic
solutions. The limiting value for the latter process is pH 9, as
the trace recorded at pH 9.3 (inset Figure 5B) showed a mixed
behavior with the signal eventually decaying to a negative value.
the chromatogram obtained in our work to that reported with
12
isocaffeine.
•
The yields of 8-OH-X in reaction of OH with and without
O2 are 12% and 28% OH , respectively (Table 1). The
•
formation of the oxidized product even in the absence O2 must
be due to the reaction of the radical cation formed after the
The rate for the former reaction was seen to depend on the
OH ] (inset Figure 5C) and the second-order rate constant for
-
-
•
[
elimination of OH from X-4OH (reaction 9a, Scheme 1) with
9
-1 -1
•
this reaction was evaluated to be 6 × 10 M
s
and an
X-8OH (reaction 16, Scheme 2). Such an explanation was
also given for the formation of the oxidized product in TMA.
The observed yield of the oxidized product (28% OH ) is in
5
-1
10
intercept of 3 × 10 s . We interpret this as an equilibrium
•
-
•
between X-8OH and OH (reaction 15). If the reaction is an
approach to equillibrium, one obtains a pKa value of ap-
proximately 9.7. This pKa value is consistent with the magni-
•
+
•
reasonable agreement with that of (X-4) (30-35% OH ).
The formation of the oxidized product in the presence of O2
can be interpreted in terms of H-abstraction from X-8OH•
-
tude of the conductivity signal as a function of [OH ].
(
reaction 17, Scheme 2), and by the peroxyl radical formed from
•
the addition of O2 to X-4OH (reaction 11, Scheme 1). Since
the yield of the oxidized product is reduced nearly by 50%, the
competing hydrolysis reaction (reaction 18, Scheme 2) leading
to the elimination of O2 must be equally effective. No
formation of the oxidized product of trimethyladenine was
noticed in the presence of O2, which was attributed to the
complete elimination of O2 from the peroxyl radical. The
H-abstraction from A-8OH is not likely to occur due to its
•
-
(15)
1
0
•
-
•
•
It seems unlikely that it is the reaction with X-4OH adduct
because of the kinetic shape of the conductivity traces obtained
ring-opening because the competing ring opening process is too
fast.
•
below pH 9.4 Also, if the equillibrium were with X-4OH ,
-
the adduct would eventually react Via OH elimination mech-
The formation of the oxidized product of caffeine in reaction
•-
anism. Finally, there is a slower reaction ocurring on a
of SO4 in neutral solutions must be the result of the formation

OH Adducts of 1,3,7-Trimethylxanthine
J. Phys. Chem. A, Vol. 101, No. 16, 1997 2959
of X-8OH (and its subsequent oxidation by S2O82-) from the
•
(2) (a) von Sonntag, C. The Chemical Basis of Radiation Biology;
Taylor and Francis: London, 1987. (b) Indem. Radiat. Phys. Chem. 1987,
hydrolysis of the radical cation (occurring under steady-state
conditions) formed from the initial attack of SO4 . The lack
of formation of 1,3,7-trimethyluric acid in basic solutions is
because of the deprotonation of the X-8OH adduct and its
subsequent ring opening. The estimated pKa value (9.7) for
3
0, 313. (c) Physical and Chenical Mechanisms in Molecular Radiation
•
-
Biology, Glass, W. A, Varma, M. N., Eds.; Plenum: New York, 1991.
(3) von Sonntag, C. Angew. Chem., Int. Ed. Engl. 1991, 30, 1229.
(4) Steenken, S. Chem. ReV. 1989, 89, 503.
5) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J.
Phys. Chem. Ref. Data 1988, 17, 513.
6) O’Neill, P. (a) Idem. Radiat. Res. 1983, 96, 198. (b) Indem. In
•
(
•
X-8OH from the conductivity is in support of this observation.
(
OxidatiVe Damage and Related Enzymes; Rotilio, G., Banniste, J. V., Eds.;
Harwood Academic: Chur, NY 1984; p 337.
Conclusions
(
(
(
7) O’Neill, P.; Chapman, P. W. Int. J. Radiat. Biol. 1985, 47, 71.
8) O’Neill, P.; Davies, S. E. Int. J. Radiat. Biol. 1986, 49, 937.
9) Vieira, A. J. S. C.; Steenken, S. J. Phys. Chem. 1987, 91, 4138.
Reaction of OH radicals with caffeine yields the OH adducts
at C-4 and C-8 positions, which absorb between 300-330 nm.
•
Ring opening of the X-8OH has not been observed; but the
(10) Vieira, A. J. S. C.; Steenken, S. (a) Indem. J. Am. Chem. Soc. 1987,
109, 7441. (b) Indem. J. Am.Chem. Soc. 1990, 112, 6986. (c) Indem. J.
Phys. Chem. 1991, 95, 9340.
11) Vieira, A. J. S. C.; Candeias, L. P.; Steenken, S. J. Chim. Phys.
Phys.-Chem. Biol. 1993, 90, 881.
(12) Fujita, S.; Steenken, S. J. Am.Chem. Soc. 1981, 103, 2540.
(13) Hazra, D. K.; Steenken, S. J. Am. Chem. Soc. 1983, 105, 4380.
14) Jovanovic, S. V.; Simic, M. G. J. Am. Chem. Soc., 1986, 108, 5968.
15) (a) Rao, R. R.; Arvindakumar, C. T.; Rao, B. S. M.; Mittal J. P.;
Mohan, H. J. Chem. Soc., Faraday Trans. 1995, 99, 615. (b) Rao, R. R.
Ph. D. Thesis, University of Pune, 1996.
•
-
X-4OH adduct undergoes OH elimination. The yield of the
latter adduct estimated from both optical and conductivity
(
•
•
measurements is 35% OH , and the rest is assigned to X-8OH .
The position of the CH3 group at N-9 or N-7 of the imidazole
ring does seem to affect not only the yields of the isomeric OH
adducts but also their transformation rates. The resolved spectra
(
(
•
•
of X-8OH and X-4OH have λmax at 300 and 335 nm,
respectively, and the corresponding ꢀ values are 6500 and 5300
-
1
-1
M
cm .
(16) (a) Langman, S. R.; Novais, H. M.; Shohoji, M. C. B.; Telo, J. P.;
Viera, A. J. S. C. presented at the 19th Miller Conference Radiation
Chemistry, Cervia/Milano Marittima (RA), Italy, September 16-21, 1995.
Acknowledgment. The authors thank Mr. R. R. Rao for his
(
b) Idem. J. Chem. Soc., Perkin Trans. 2 1996, 1461.
help in carrying out the experiments in the initial stages of the
work. They are also thankful to Prof. M. S. Wadia for useful
discussions and one of us (B.S.M.R.) is thankful to Dr. A. D.
Trifunac for supporting his visit to Argonne National Laboratory.
The financial support from the Department of Science and
Technology, India, for carrying out this project is acknowledged
and M.V. is thankful to the Council of Scientific and Industrial
Research, India, for the award of a Senior Reasearch Fellowship.
The work at Argonne was performed under the auspices of the
Office of Basic Energy Sciences, Division of Chemical Science,
U.S. DOE under Contract W-31-109-ENG-38.
(
17) Vieira, A. J. S. C.; Steenken, S. J. Chim. Phys. 1996, 93, 235.
(18) Guha, S. N.; Moorthy, P. N.; Kishore, K.; Naik, D. B.; Rao, K. N.
Proc. Indian Acad. Sci., Chem. Sci. 1987, 99, 261.
(19) Schmidt, K. H. Int. J. Radiat. Phys. Chem. 1972, 4, 439.
(20) (a) Schmidt, K. H.; Gordon, S.; Thompson, R. C.; Sullivan, J. C.;
J. Inorg. Nucl. Chem. 1980, 21, 321. (b) Schmidt, K. H.; Gordon, S.;
Thompson, R. C.; Sullivan, J. C.; Mulac, A. Radiat. Phys. Chem. 1983,
21, 321.
(
21) Kesavan, P. C.; Powers, E. L. Int. J. Radiat. Biol. Relat. Stud. Phys.
Chem. Med. 1985, 223.
22) Scholes, M. L.; Schuchmann, M. N.; von Sonntag, C. Int. J. Radiat
(
Biol. 1992, 61, 443.
(23) Candeias, L. P.; Wolf, P.; O’Neill, P.; Steenken, S. J. Phys. Chem.
992, 96, 10302.
1
(
24) (a) Mohan, H.; Mudaliar, M.; Arvindakumar, C. T.; Rao, B. S.
References and Notes
M.; Mittal, J. P. J. Chem. Soc., Perkin Trans. 2 1991, 1387. (b) Mudaliar,
M. Ph. D. Thesis, University of Pune, 1993.
(1) Dedicated to Professor M. S. Wadia on his 60th birthday. This work
forms a part of the Ph.D. thesis of M.V. to be submited to the University
of Pune.
(25) Stadler, R. H.; Janique, R.; Turseky, R. J.; Welti, D. H.; Fay, L. B.
Free. Radical Res. 1996, 24, 225.