Electrochemistry of pteridine

Electroanalytical Chemistry and Interfacial Electrochemistry, 59 (1975) 75-98 ~) ElsevierSequoia S.A.,Lausanne- Printed in The Netherlands

75

ELECTROCHEMISTRY OF PTERIDINE

DAVID L. McALLISTERand GLENN DRYHURST* Department of Chemistry, University of Oklahoma, Norman, Okla. 73069 (U.S.A.) (Received 21st January 1974;in finalform 7th October 1974)

Several years ago Komenda and Laskafeld I reported that pteridine is electrochemically reducible at the dropping mercury electrode (DME). The latter work, however, was limited to a study of the d.c. polarography of pteridine and consequently little information was obtained concerning products, electrochemical and related chemical mechanisms and appropriate rate constants. The electrochemical oxidation of pteridine has not been previously reported. As part of a systematic investigation of the electron-transfer reactions of pteridines reported here is a study of the electrochemistry of pteridine. EXPERIMENTAL Chemicals Chemicals were obtained from the following sources: quinoxaline, 4,5diaminopyrimidine,4,5-diamino-6-hydroxy-2-mercaptopyrimidine (Aldrich); glyoxal (Mann Research Laboratories); phosphorus pentasulfide (Eastman); Raney nickel (Sargent-Welch). 4,5-Diaminopyrimidine was synthesized from 4,5-diamino-2,6-dimercaptopyrimidine according to Beaman et al. 2, 4,5-diamino-2,6-dimercaptopyrimidine from 4,5-diamino-6-hydroxy-2-mercaptopyrimidine by the method of Beaman and Robins 3. Raney nickel catalyst was prepared according to Brown 4. Pteridine was synthesized from 4,5-diaminopyrimidine and glyoxal by the method of Albert and Yamomoto 5, except that crystalline glyoxal hydrate was used instead of polymeric glyoxal monohydrate. Buffer solutions were prepared with an ionic strength of 1.0 M, giving a 0.5 M ionic strength upon 1:1 dilution. Argon and nitrogen used for deoxygenation were equilibrated with water in a bubbling chamber. Thin-layer chromatography was carried out on Eastman Chromagram sheets of silica gel with fluorescent indicator. Although several developing solvents were tried, the best separations were obtained with absolute methanol. Visualization was with ultraviolet light. Apparatus The equipment employed for polarography, voltammetry, coulometry and preparative electrolysis has been described extensively elsewhere 6. Double potential

* To whomfurthercorrespondenceand reprint requests should be directed.

76

D . L . McALLISTER, G. DRYHURST

step chronoamperometry utilized a programmable square wave pulse generator constructed in this laboratory, All potentials are referred to the saturated calomel reference electrode at 25°C. A Buchler Fractomat fraction collector was utilized in column chromatography. I.r. spectra were recorded on a Beckman i.r.-10 spectrophotometer, u.v. spectra on a Hitachi Model 124 spectrophotometer, n.m.r, spectra on a Varian T-60 spectrometer and mass spectra on a Hitachi RMU-6E spectrometer. Isolation and characterization of electrolysis products. After completion of the reduction of 200-500 mg pteridine on the plateau of wave I (le) in pH 7 phosphate buffer, the solution was lyophilized, and then extracted with methanol. Thin-layer chromatography of the electrolysis solution and the methanol extract confirmed that the principal reduction products were contained in the extract. The methanol was evaporated and the residue was dissolved in a small amount of water and lyophilized, leaving a brown pi~wder. As described later, repeated attempts at purification were unsuccessful. For n.m.r, spectra the brown powder was dissolved in deuterated methanol (100 mg in 0.4 ml), and for i.r. a KBr pellet was prepared (1 mg in 100 mg KBr). Reduction of pteridine at pH 2 was carried out in 0.01 M HC1 in order to eliminate problems of separating organic product from mixtures of the organic and inorganic materials normally used as buffer constituents at this pH. The pH of the solution was monitored continuously during the electrolysis and adjusted to pH 2 by addition of 2 M HC1. Lyophilization of the solution following reduction of pteridine at pH 2 at -0.80 V vs. SCE (on the plateau of wave IV, 2e) yielded a brown residue which gave a very acidic solution when dissolved in water. The residue was dissolved in a small amount (3 5 ml) of water, applied to the top of a 2 x 50 cm column of ion retardation resin (Bio-Rad 11A8, 50-100 mesh), and eluted with water at a flow rate of 5 ml min- 1. Fractions of 100 drops were collected, and the fractions which contained reduction product (detected by u.v.) but no chloride (absence of precipitate with AgNO3) were combined and lyophilized. The column was regenerated by washing with 1 M HC1 (400 ml), 1 M NH4OH made 0.5 M in NH4C1 (800 ml), 1 M NH4C1 (400 ml), and water (3000 ml or until column effluent is chloride-free), in that order. To prevent formation of bubbles in the column, the water used in the final regeneration was purged with nitrogen. From the electrolysis of 200 mg pteridine, 75-100 mg of chloride-free product was usually obtained. For n.m.r. spectra, 100 mg of product was dissolved in 0.4 ml 0.5 M DC1 in DzO (solubility of the product in neutral solutions was not sufficient for n.m.r.), and for i.r. spectra, a KBr pellet was prepared (1 mg in 100 mg of KBr). RESULTS AND DISCUSSION

Chemical equilibria of pteridine In order to understand the electrochemistry of pteridine, it is first necessary to review the rather complex equilibria of this compound in aqueous solution. These equilibria have been studied extensively by Perrin v and Albert et al. 8 using u.v. and n.m.r, spectroscopy. When pteridine is initially dissolved in neutral aqueous solution, it exists virtually entirely as the non-hydrated neutral molecule (I, Fig. 1),

77

ELECTROCHEMISTRY OF PTERIDINE H

0

o

"H÷ I I N

H HN

~--~ N~j

I

(~T)

OH N'-K

H~,O N

slow

. H÷

IH30,

)

H'l l''''

(i)

OH

1

A

o.

H÷

ili-

H

(m)

(~1

Fig. 1. Equilibrium between the neutral, non-hydrated form of pteridine (I), neutral 3,4-hydrate (II), the cation of the 3,4-hydrate (III), the cation of the 5,6,7,8-dihydrate (IV) and the anion of the 3,4hydrate (V).

I

I

I

I

I

I " I

[

I

I

I

t

I

I

1

I

I

[

I

[

I

I

1

I

I

I

I

I

1

I

I

I

I

I

I

I

I

I

I

I 3

[ 4

I 5

I 6

I 7

I 8

J 9

I 10

] I1

L 12

I 13

-1.6 - 1.4

U

: -1.o 2 .oe o

> -.0.6 "~- 0.4 -0.2

0

B

i2°-~ I 0

0.0 ! 1

( 2

pH

Fig. 2. (A) Variation of E~ and (B) polarographic diffusion current constant, I, with pH for pteridine wave I (O--(D), wave II ( A - - / x ) and wave III ( D - - D ) .

78

D.L. McALLISTER, G. DRYHURST

but over the course of several minutes attains equilibrium with the neutral 3,4hydrate (II, Fig. 1). At equi!ibrium in neutral solution the mixture consists of 79% of I and 21~o of II 8. In acidic solutions (pH < 3), pteridine exists initially as the cation of the 3,4-hydrate (III, Fig. 1) which slowly equilibrates with the cation of the 5,6,7,8-dihydrate (IV, Fig. 1). In basic solution of pH > 11 the anion of the 3,4-hydrate (V, Fig. 1) is formed.

D.c. polarography Between pH 1-13 pteridine exhibits three polarographic reduction waves at the DME, the E~ of which shift linearly more negative with increasing pH (Fig. 2A) according to the equations:

I

I

I

l

J

I

I

II

g

,( 4 =L

o 2 ia

-+ I

I

I

I

I

-0.2

-0.4

-0.6

-0.2

-0,4

Potential/Volts I

I

1 -0.6

vs

]

[

--0.8

-1.0

SCE

I

I

I

I

B

(a)

(b)

IT

4

n

c e

,32

0

I 0.0

I -0.2

I 0.0

I -0.2

Potential/Volts

I -0.4

I -0.6

I -0.8

v$ S C E

Fig. 3. Polarograms of 1 mM ptcndine (A) in pH 7.0 Mc|lvaine buffer and (B) in 1 M acetic acid pH 2.3. (a) Immediately after dissolution, (b) after 15 min.

ELECTROCHEMISTRY OF PTERIDINE

79

wave I (pH 3-13), E}= 0.04-0.064 pH wave II (pH 1-13), E ~ = - 0 . 3 6 - 0 . 0 7 0 pH wave III (pH 11-13), E}= -0.54-0.080 pH The variation of the diffusion current constant ( I = i l / C m } t ~) is shown in Fig. 2B. Owing to the slow chemical equilibria involved, values of I at a given pH change as a function of time. The values of I shown in Fig. 2B are those measured 5 min after preparation of pteridine test solutions. The pteridine species responsible for each polarographic wave is readily established by comparison of the changes in polarography of pteridine at various pH values as a function of time to the slow equilibria processes. Thus, a freshly prepared solution of pteridine at pH 7.0 shows a well-defined wave at E+ = -0.44 V (wave I) and a very small wave at E}= -0.82 V (wave II) (Fig. 3A). After about 15 rain, however, wave I has decreased in height while wave II is larger. This behavior implies that wave I is due to the reduction of non-hydrated pteridine, and wave II to the reduction of the 3,4-monohydrate. Similarly, in 1 M acetic acid (pH 2.3), pteridine shows a very small wave I and a much larger wave II (Fig. 3B). After a few minutes, wave II has decreased in height owing to formation of the non-electroactive 5,6,7,8-dihydrate. Wave III, observed only at high pH, was ascribed to reduction of the anionic form of the 3,4-hydrate. That the decrease in wave height for wave II at low pH with time is due to equilibria and not to decomposition can be shown, for example, by neutralization of a solution of pteridine in 1 M acetic acid to pH 7. After several minutes, a polarogram of the equilibrium mixture ofpteridine and its 3,4-hydrate is observed (i.e., a polarogram similar to that shown in Fig. 3A,B). Owing to the fact that at least two pteridine species are in equilibrium at any pH, values of the diffusion current constant (I) for waves I, II or III did not allow any definite conclusions regarding the number of electrons involved in the electrode process. However, over the range of 2-12, the sum of I for all waves appearing at any pH had values ranging from 2 to 3, suggesting an overall 2 electron polarographic process. For reasons described later, the n-value for wave I was of particular interest. This was first estimated by the well-known log-plot method 9 where the electrode potential, expressed as E + E~, is plotted as a function of log10 i/il-i over the rising portion of the polarographic wave, where /=current at potential E and il=limiting current of the polarographic wave. For a reversible process such plots should be linear for values of the log term from -1.5 to 1.5 and the slope should be 59/n mV at 25c'C. Experimental plots at representative pH values for wave I were linear over the latter range and the slope values at pH 7.1 (31.9 mV) and 8.1 (36.6 mV) were indicative of a reversible 2e reaction, while the value for pH 10.8 (46.7 mV) is intermediate between that expected for n = 1 and n =2. The corrected mercury column height dependence of the limiting current was indicative of diffusion control for waves I, II and III at all pH values, i.e., il/hcorr was independent of the corrected mercury column height 1o. However, the value of q/h .... varied greatly with pH. For example, for wave ! the magnitude of the latter value paralleled the variation of the diffusion current constant, I, with pH (see Fig. 2B). Thus, it was small at pH 2, reached a maximal value at ca. pH 5.6 and •

80

D. L. McALLISTER,G. DRYHURST

then decreased with further increase of pH. The constancy of ij/h~,,rr at any specific pH along with the variation in its value with change of pH implies, at low pH, that there is no appreciable kinetic contribution to the wave I or wave II processes by dehydration of the 3,4-hydrate of pteridine to pteridine or vice versa, i.e., the waves show behavior typical of diffusion control. The decrease of ij/h~o~ for wave I above about pH 8 will be discussed subsequer~tly. Values of the temperature coefficient of the waves confirmed diffusion control, falling within the range 12~o per ~C.

Cyclic and linear sweep voltammetry At both the pyrolytic graphite electrode (PGE) and the hanging mercury drop electrode (HMDE), pteridine exhibits three reduction peaks designated Ic, IIc and III~ and corresponding to polarographic waves I, II and III. No oxidation peaks are observed at a clean PGE on the initial positive-going sweep, i.e., pteridine is not electrooxidizable at the PGE. However, once having scanned peak lc, and then sweeping towards more positive potentials, two anodic peaks are observed. The first, peak I~, occurs at potentials slightly positive of cathodic peak Ic, and the other, peak IIa, is observed at more positive potentials at the PGE (Fig. 4). The variation of peak potentials, Ep, at the PGE with pH along with the pH range over which the peaks appear is shown in the following equations: peak peak peak peak peak

I c (pH II~ (pH IIIc (pH I, (pH II, (pH

2-13) 1-3 ) I0-11) 1-13) 2-13)

Ep=0.005-0.068 pH Ep= - 0 . 3 3 - 0 . 0 7 9 pH Ev= -0.38-0.108 pH Ep=0.12-0.071 pH Ev= 1.08-0.080 pH

At low pH (e.9., pH 2.8 McIlvaine buffer), if the initial sweep at the PGE I

1

I

I

Ic 50

Dc

"2o~ c ®

-t; 0

1

t

I

I

O. 5

0

-0,5

-1.0

Potential/Volts

vs

SCE

Fig. 4. Cyclic v o l t a m m o g r a m of 1 m M pteridine at p H 7.0 at the PGE. Potential sweep pattern:

0.00 V~ - 1.30V~0.90 V~0.00 V. Sweeprate 200 mV s 1.

ELECTROCHEMISTRY OF PTERIDINE

81

or H M D E is in a negative direction, a very small peak Ic is observed. If, after sweeping past peak I,, the sweep direction is reversed, a small peak I a appears (Fig. 5A). However, should the negative going sweep be continued after scanning peak I a a large peak II~, corresponding to the reduction of the 3,4-hydrate of pteridine, is observed. If the sweep direction is reversed after sweeping beyond peak II~, a large peak I a is observed. After sweeping past this peak and reversing the sweep direction again, peak Ic becomes much larger (Fig. 5B). This behavior implies that the initial electrode product of peaks I~ and II~ is the same, and that this product can be oxidized at peak I a producing pteridine. I 20

A

0 -20

I

t

I

I

I

<

.r c

40

20 m

0 U

m

-20 -40 -60

I

~

0.2

0

I

Potential/Volts

L -0.4

1

L -0.8

vs SCE

Fig. 5. Cyclic voltammogram of 1 mM pteridine at the PGE in pH 2.8 McIlvaine buffer. Potential sweep pattern (A) 0.00 V ~ - 0 . 2 3 V~0.05 V ~ - 0 . 2 3 V; (B) 0.00 V ~ - 0 . 9 0 V-~0.05 V-~0.23 V. Sweep rate 100 mV s 1.

D.c. polarography indicates that the reaction responsible for polarographic wave I is an electrochemically reversible 2-electron process. Cyclic voltammetry confirms this finding since at slow sweep rates ( 10-50 mV s - 1) at both the HMDE and PGE the potential increment between the pteridine peaks Ic and I a peak potentials is close to the theoretical 29 mV (between pH 3.8 and 11) expected lz'12 for a reversible 2e reaction. At faster sweep rates peak separation increases, the effect becoming noticeable at lower sweep rates at the PGE, i.e., the electron transfer reaction is somewhat less reversible at graphite than at mercury (vide infra). Further confirmation of the 2e nature of peak Ic was obtained by comparison of the experimental and theoretical reversible voltammetric peak current functions (ip/ACv~), the latter being calculated from eqn. (1) 13, all terms having their usual electrochemical significance. The experimental peak

82

D.L. McALLISTER, G. DRYHURST I

I

I

I

(a)

<

c e

u

,

1

__

-0.4

-0.2

I

-d

I -0.4

I

-0.6 -0.2

Potential/Volts

I

l

vs

!

SCE

'1

-

A

(a)

1 ~0.6

1

I

(b)

40

20 < c w

~O u

-20

I

I

t

I

I

I

-0.1

-0.3

-0.5

-0.1

-0.3

-0.5

Potential/

Volts

vs

SCE

Fig. 6. Cyclic voltammograms of 1 mM pteridine in (A) pH 9.0 borax buffer, sweep rate (a) 200 mV s 1, (b) 10 mV s - 1 (B) 1 mM pteridine in (a) pH 10.0 borax buffer and (b) pH 7.0 McIlvaine buffer. Sweep rate 100 mV s -1. Both (A) and (B) obtained at a PGE.

ELECTROCHEMISTRY OF PTERIDINE

83

iv = 2.69 x 1 0 5 A D } v 4 n ~ C

(1)

current functions at p H 7 and 8 at, for example, the H M D E were 1651_+60 and 1801 + 6 0 t~A cm 2 (mmol 1 - 1 ) - i ( V s - 1 ) - 4 respectively. The theoretical reversible peak current function for le and 2e reactions are 551 and 1560 pA cm -2 (mmol 1-1)- 1 ( V s - 1)- ~ respectively. The value of the diffusion coefficient of pteridine used in the latter calculations, 4.203 x 10- 6 cm 2 s - x, was determined by the potentiostatic method of Shain and Martin 14. Studies of the cyclic voltammetry of pteridine also revealed variations in the relative magnitude of currents for peak Ic and peak I a with variation of p H and voltage sweep rate. That is, at low sweep rates the value of the ratio (ip)c/(ip) a where (ip)c = I~ peak current and (ip)a -~ Ia peak current was much larger than at faster sweep rates (Fig. 6A). Also, the ratio (ip)c/(ip) a was larger at high p H than at lower p H at a constant sweep rate (see, for example Fig. 6B). This behavior indicates that during the time of the cyclic voltammetric experiment, a portion of the initial electrode reduction product reacts to form an electroinactive species which cannot be electrochemically oxidized on the reverse sweep. The effect is clearly more pronounced at low sweep rates and high pH. The reaction following the initial electron transfer process could be one of several types:(1) a chemical reaction followed by a second electrochemical reaction (i:e., an e.c.e, reaction), (2) a simple chemical decomposition of the initial electrode product, (3) reaction of the electrode product with solvent or supporting electrolyte, (4) dimerization of the electrode product or (5) ,reaction of the electrode product with starting material. The possibility of an e.c.e, mechanism was eliminated by a study of the variation of the peak current for peak Ic with voltage sweep rate. According to eqn. (1) ip/1)½ should remain constant with changes in v. However, in the case of an e.c.e. m e c h a n i s m ip/t: ½ should decrease with increasing voltage sweep rate a5. F o r pteridine peak Ic, ip/V ~ w a s constant at p H 7 and 8 indicating the absence of any secondary reaction (Table 1). At p H 9 and 10 the value of iv/v ~ is somewhat lower at slow TABLE 1 VARIATION OF ip/v+- AS A FUNCTION OF v FOR VOLTAMMETRIC PEAK Ic OF PTERIDINE HMDE, area 0.022 cmz, pteridine concentration 1 raM. v/V ~-i

iov-~/pA V }s ~ pH 7

0.01 0.02 0.05 0A0 0.20 0.50 1.0 2.0 5.0 10.0

36.0 36.2 36.5 36.5 37.4 36.8 35.0 34.0 33.9 34.8

pH 8

pH 9

pH 10

38.4 . 38.7 39.5 40.2 40.8 39.6 40.0 36.9 35.7 38.0

31.7 32.1 34.2 35.7 36.0 36.8 37.0 36.8 37.5 38.0

23.4 24.1 27.6 29.8 32.6 33.9 35.0 35.4 38.0 39.6

84

D.L. McALLISTER, G. DRYHURST

sweep rates. This is opposite of the effect expected for an e.c.e, process. A discussion of the latter effect will be presented later. A study of the effect of concentration on the cyclic voltammograms of pteridine revealed that the reaction following the initial electron transfer occurred much more rapidly at high pteridine concentration. Thus, at a constant sweep rate, the height of peak I~ relative to that of peak Ic was much less at a pteridine concentration of 2 mM than at 0.2 mM. This suggests that the follow-up chemical reaction might be second order in pteridine, and tends to eliminate possibilities (2) and (3) which would be first order and pseudo first order, respectively. Coupled with coulometric evidence, (vide infra), it was concluded that the reaction is best represented by (5), i.e., reaction of the primary electrode product with starting material.

Controlled potential electrolysis and coulometry Because of the many possible forms of pteridine in aqueous solution (Fig. 1), it was necessary to choose carefully the conditions for controlled potential electrolysis to ensure that essentially a single species was available to react at the electrode surface at the desired potential. For example, electrolysis on the plateau of wave II at pH 7 (see Fig. 3) would result in the reduction of both non-hydrated pteridine and its 3,4-hydrate. For this reason, electrolysis on wave II was carried out at low pH (2-3) where essentially all of the pteridine is in the form of the 3,4-hydrate. On the other hand, electrolysis of wave I could be performed at almost any pH, although at low pH the concentration of non-hydrated pteridine is very low. Since

I

I

I

I

I

IT

I 3/u A I, <~

B

-÷

J

+c

I o

-0.2

I

I

I

I

-0.4

-0.6

-0.8

-1,0

Potential/Volts vs

1, -1.2

SCE

Fig, 7. Polarograms of a 1 mM solution of pteridine at pH 4.7 (A) before electrochemical reduction, (B) after reduction at -0.45 V and (C) after reduction at - 1.00 V.

ELECTROCHEMISTRY OF PTERIDIN. E

85

equilibrium between the non-hydrated form and the 3,4-hydrate is slow electrolysis at. wave I potentials proceeded very slowly. Electrolysis of pteridine on the plateau of wave I at a mercury pool electrode at the 1 m M concentration level between pH 3.6-11.9 results in the transfer of 1.0 + 0.03 electrons per molecule of pteridine. With decreased pteridine concentration the n-value increased, thus a 0.078 m M solution of pteridine gave an n-value of 1.28 at pH 4.7 and 1.13 at pH 7. The only logical interpretation of these facts is that the initial electrochemical reaction product resulting from transfer of 2 electrons reacts with starting material (pteridine) on a 1:1 basis forming an electroinactive (at potentials corresponding to the plateau of wave I) compound. This product will be shown subsequently to be a pteridine dihydro dimer. Upon completion of the electrolysis the solution exhibited two polarographic reduction waves, one (wave IV) at a potential slightly negative of pteridine wave II, and another very poorly defined wave (wave V) at more negative potentials (Fig. 7). The E~ values for both waves IV and V shift linearly more negative with increasing pH over the pH ranges indicated according to the equations: / wave IV (pH 2.3-12), E~ = - 0 . 4 8 - 0 . 0 6 8 pH wave V (pH 2.3-8), E~ = - 0 . 8 0 - 0 . 0 5 5 pH Wave V is not observed if the electrolysis is carried out above pH 8. Also, after electrolysis at wave I or wave II potentials, an oxidation peak is observed at the PGE the peak potential of which is described by the equation Ep = 0 . 7 6 - 0 . 0 3 4 pH

The u.v. spectrum of the electrolysis product (,~max= 307 nm at pH 4.7, 308 nm and 263 nm at pH 7.0) was quite different to that of pteridine (e.g., 2 .... = 309 nm and 298 nm at pH 7.0). At all pH values studied, electrolysis at potentials corresponding to wave I resulted in the disappearance of wave II. Electrolysis ofpteridine at potentials corresponding to polarographic wave II always resulted in identical electron numbers and products as electrolysis at wave I. However, owing to the proximity of wave II to the wave remaining after electrolysis (wave IV, Fig. 7), it was necessary to select the electrolysis potential very carefully to correspond approximately to the E~ of wave II rather than on the plateau. After electrolysis at potentials corresponding to pteridine waves I or II, electrolysis on the plateau of wave IV resulted in the transfer of 1.0_+0.1 electrons based on the amount ofpteridine initially present. However, the diffusion current for wave IV, when compared to that for wave I (previously shown to be a 2e process under polarographic conditions) is about one half as large, indicating a l e process for wave IV. Analysis of wave IV by the log plot method indicated that the electrode process is irreversible (i.e., the plot was not linear over the range of the log term from - 1 . 5 to 1.5 vide supra). In addition, a reversible couple was not observed by cyclic voltammetry at the HMDE. As in the case of a reversible system, the number of electrons involved in an irreversible electrochemical process may be estimated from the slope of the log plot, although a different approach is required. For an irreversible electrode reaction the slope of the log plot is given by eqn. (2) 16.

86

D.L. McALLISTER,G. DRYHURST d(E+E~)

_

- 54.2 mV

(2)

d log where ~ = electron transfer coefficient, na = number of electrons involved in the rate controlling step, all other terms have been described earlier. At pH 7.0, for example, ~na for pteridine wave IV, calculated from the slope of the log plot and eqn. (2) is 0.98. Assuming 16 a value of a of approximately 0.5, the value of na must be 2. Similar data were obtained at other pH values. The apparent contradiction between the coulometric value for wave IV (1) and the n-value obtained from the log plot method (2) is readily explained if the compound responsible for wave IV is a form of pteridine dimer. In this case, if the number of electrons transferred per molecule of dimer is 2, as the log plot indicates, the n-value calculated on the basis of the original pteridine present will be 1. Also, since the molar concentration of a dimer would be one half the original pteridine concentration, the polarographic diffusion current would be approximately one half the value of the diffusion current observed for the 2e reduction of the original pteridine. Following reduction at potentials corresponding to the plateau of wave IV, only polarographic wave V is observed (Fig. 7C). A single oxidation peak is observed at the PGE having the same peak potential as that observed after reduction at wave I or wave II potentials. The u.v. spectrum of the wave IV product was identical to that of the wave I-II product, but the molar absorptivity had decreased by approximately

25°4. Coulometry on the plateau of wave V, after having first electrolyzed at potentials sufficient to remove waves I, II and IV gave fractional n-values (0.2-0.3). The u.v. spectrum after removal of wave V showed no appreciable difference to that observed after electrolysis at wave IV potentials. Coulometry of pteridine on the plateau of wave III at pH 12 gave n-values of 2+0.1, and a product was formed that had an identical u.v. spectrum to that observed following reduction of pteridine at wave I or wave II followed by reduction at wave IV. After coulometry at a potential corresponding to polarographic wave III, a voltammetric oxidation peak was observed at the PGE that occurred at the same peak potential as that observed following reduction of wave IV. However, no polarographic (DME) reduction waves were observed.

Mass electrolysis, product isolation and characterization In order to properly interpret the polarography, cyclic voltammetry and coulometric data on pteridine it was necessary to characterize the various reduction products. In order to obtain sufficient quantities of material for characterization, solutions 10 m M to 40 m M in pteridine were electrolyzed. Based upon voltammetric and polarographic studies, the electrochemical behavior of pteridine in solutions of pH 2.0 (0.01 M HC1) and pH 7.0 (McIlvaine or phosphate buffers) was judged to be representative and mass electrolyses were conducted almost exclusively in these media. Pteridine wave I reduction product Lyophilization of a solution after reduction of 500 mg of pteridine at pH 7

87

ELECTROCHEMISTRY O F PTERIDINE

at a potential corresponding to the plateau of wave I gave a pale brown residue. The reduction product was separated from buffer components by extraction with methanol. Evaporation of the methanol resulted in formation of a brown powder. Thin-layer chromatography of this product (usually on silica gel developed with methanol) showed three spots (u.v. visualization). The major product had an Rf value of 0.55 with minor products having Rf values of 0.40 and 0.00. Voltammetry and the u.v. spectrum of the brown product were identical to that of the electrolyzed solution. Studies of the u.v. spectrum of the product in water and ethanol suggested that above a pH of 6.5 the predominant species in the solution is a free base, while below this pH a protonated species is formed. Attempts to purify the major product (Rf=0.55) by recrystallization from polar solvents (water, methanol, acetonitrile) in which it was moderately soluble resulted in decomposition, particularly upon heating the sample to affect dissolution. Several chromatographic purifications were attempted (e.9., with silica gel, alumina, Sephadex and ion-exchange columns). Again, partial decomposition of the sample made it impossible to effect a complete purification of the major product.

= 80 o "o

g 60 m ~: 20

60

80

100

120 m /e

B

140

4

160

5

260

5'

4'

I

1'

I

4,4'

6 6.6,

1

'2

I

7,7,

H

H

8

8,

,M.OH

0

l

[

I

l

10

9

8

7

I

6

'"1"

i

I

I

i

i

1

I

i

i

I

I

/

I

I

5

4

2

(~)

ppm I

I

I

I

I

i

I

I

I

I

I

I

I

90 80

~ 70 ~ 60 E 50

~ 40 ~ 30 N 2O 10

4000

3000

2000

1600 ¢ m -1

1200

800

I

400

Fig. 8. Mass (A), 60 mHz n.m.r. (B) and i.r. (C) spectra of pteridine wave I reduction product. N.m.r. solvent deuterated methanol. Mass spectrum taken at 175°C and 75 eV ionizing voltage. I.r. is of KBr disc,

88

D.L. McALLISTER,G. DRYHURST I

I

t

A

~= 8 0

e0

I

-= 4 0 20

I ' 60

5O

7'0

80

90

I

I

i

I

100

110

120

130

m/e I

I

i

3N~N~G 4

10

5

I

I

I

[

I

I

I

9

8

7

6

5

4

3

ppm I c

i

I

I

I

(~) I

I

I

I

1

I

6O 4O 2O I 4000

I 3000

2000 ¢m

1600

1200

800

.1

Fig. 9. Mass (A), 60 mHz n.m.r. (B) and i.r. (C) spectra of pteridine. Mass spectrum taken at l(~)°C

and 75 eV ionizingvoltage.N.m.r. solventdeuteratedmethanol.I.r. is of KBr disc. Even though the reduction product could not be purified, n.m.r, and mass spectrometry of a partially purified material gave considerable information. The mass spectrum of the reduction product is shown in Fig. 8A. Of particular significance is the group of peaks at role of 264, 265, 266 which indicate the presence of a pteridine dimer. Very small isotope peaks at m/e 267, 268 are not shown in Fig. 8A. The mass spectrum of pteridine is shown in Fig. 9A. The n.m.r, spectrum of the wave I reduction product in deuterated methanol shows three groups of peaks centered around 6 values of 5.5,7.3 and 7.8 (Fig. 8B). The n.m.r, spectrum of pteridine is shown in Figl 9B. The most significant difference in the two spectra is the presence of the peaks at ca. 5.5 6 in the reduced product. This peak probably indicates the presence of methylene or methine protons, probably located at C 6 or C7. Proposed assignments are also shown in Fig. 8B. The upfield shift of the signals for the C2, C¢ and C6 protons as compared to those of pteridine (Fig. 9B) is caused by increased saturation of the pyrazine ring. A similar shift has been noted by Albert et al. 8 for 5,6,7,8-tetrahydro-6,7-dihydroxy pteridine. Because of the position of the water peak and the poor separation of the signals at 6 = 7.5 and 7.8 integration of the spectrum was of little value. However,

89

ELECTROCHEMISTRY OF PTERIDINE

visual inspection indicates that the approximate area ratio is consistent with the proposed assignments. Infrared spectra of pteridine and the wave I reduction product are shown in Fig. 9C and Fig. 8C respectively. In the case of pteridine, Mason iv assigned the bands between 1500-1600 cm-1 to ring stretching vibrations (C=N and C=C). The band at 3000 cm -1 is a C - H stretch, and the broad band at 3500 cm -1 is probably caused by small amounts of water in the KBr disc. Bands from 700-1050 cm- 1 were assigned to out-of-plane C - H bending vibrations. In particular, the band at 800 c m - ~ was proposed to be associated with the C - H group at position 4. The wave I reduction product exhibits a broad band at 3500-3000 cm- 1 indicative of N - H stretches and, perhaps, some contamination by water. The shoulder at 3000 cmmay be caused by the presence of additional C - H bonds. Ring stretching vibrations remain in the region 1500-1600 era-1, and the band at 800 cm 1 indicative of the C H group at C4 is present. The n.m.r., mass and i.r. spectral evidence is therefore consistent with a dimeric species of molecular weight 266 that contains a methine group at the pteridine C6 or C7 positions and an unchanged C - H bond at the C4 position. These data are in accord with 7,7',8,8'-tetrahydro-7,7'-dipteridyl (VI), although the possibility of isomeric forms cannot be entirely eliminated. Other possible structures (VII and VIII) are shown below. 5

I H 8

5'

H 7

H 7'

I H 8'

(9].)

H I

6 H

6' H

(~m)

H I

I H

H H 7 6" OD]I)

I H

It would be extremely difficult to distinguish between structures (VI) and (VII). Assignment of structure (VI) to the pteridine wave I reduction product is based largely on the fact that the reduction of other pteridines appears to occur at the ~27=N8-group rather than the - N s = C 6 group 18 20. In addition, the asymmetric dimer (VIII) should show two additional n.m.r, signals since protons 6 and 6' or 7 and 7' are not equivalent as they are in structures (VI) and (VII). Upon further reduction (VIII) should yield an equimolar mixture of 7,8-dihydropteridine and 5,6dihydropteridine. It wilt be shown shortly that u.v. and n.m.r, spectroscopy on the product of electrochemical reduction of the dimer indicates formation of a single product. It would be expected that structures formed by dimerization between the

90

D.L. McALLISTER, G. DRYHURST

pyrazine and pyrimidine rings (e.g., IX or X) could undergo further electrochemical reduction in the pyrazine ring. No such reduction is observed. In addition, i.r. spectrometry supports the presence of the C 4 H group, eliminating the possibility of structure (X). N

N I H

H

H

(IX)

I H

f H

H

(X)

N

//N

The major peaks observed on mass spectrometry of the pteridine wave I reduction product shown in Fig. 8A can be readily rationalized on the basis of structure (VI). Details are presented elsewhere 21. It will be recalled that the n-value for reduction of the 3,4-hydrate of pteridine (which gives rise to polarograpic wave II) was identical to that for reduction of non-hydrated pteridine at wave I potentials. In addition identical products were obtained as evidenced by the u.v. spectrum, polarography and voltammetry of the solution and the identical thin-layer chromatography of the product. In other words the dimeric pteridine having structure (VI) is the ultimate product of electrochemical reduction of non-hydrated pteridine at wave I potentials and the 3,4hydrate of pteridine at wave II potentials.

Pteridine wave IV reduction product Following reduction of 200-500 mg of pteridine at pH 2 at a potential more negative than polarographic wave IV, lyophilization of the solution left a brown acidic solid. To effect separation of the reduction product from potassium chloride and residual hydrochloric acid, the solid product was dissolved in a small volume of water and passed through a column of ion retardation resin (see Experimental). The chloride-free effluent was lyophilized leaving a yellow to buff neutral residue. This material did not melt but charred and decomposed at 200-225°C. This product was sparingly soluble in neutral or alkaline aqueous solutions and methanol, but readily dissolved in acidic solutions. It was concluded from these observations that the initial isolated product was a hydrochloride which was converted to the free base by passage through the ion retardation resin. Attempts to completely purify the product by recrystallization, reprecipitation or chromatographic methods were unsuccessful because of partial decomposition of the material. Decomposition was particularly noticeable when solutions of the product were heated, especially acidic or basic solutions. The highest purity product appeared to be obtained by passing a solution of the initially lyophilized product through an ion retardation resin. The mass spectrum of the base obtained in the above manner (Fig. 10A) showed a molecular ion at m/e = 134 which is that expected for a dihydropteridine. The n.m.r, spectrum with probable peak assignments is shown in Fig. 10B. Because of the insolubility of the free base in most solvents, the n.m.r, spectra were obtained in DzO containing deuterated hydrochloric acid. The feature of major significance in the n.m.r, spectrum is the chemical shift of the C7 protons relative to

91

E L E C T R O C H E M I S T R Y O F PTERIDINE t o~

!

I

I

I

I

I

I

l

A

8om+

.~ 4 0 m 2Q

II,

j ,Jj,J, ,,, ,1

I

60

.5O

I

I

70

80

I

I

I

I

I

00

100

110

120

130

m/e I

I

I

B

I

I

I 4

I

I

I

I

10

9

8

7

I

I

I

I

I

t

t

I

I

5

I 7

,

I..

6 ppm(G)

I

]

I

i

I

5 1

l

I

~

I

I

I

4

3

2

|

I

I

I

t

I

I

I

I

~_4o N 2O

4000

3000

2000

l

1600 on1 - I

1200

800

I

400

Fig. 10. Mass (A), 60 mHz n.m.r. (B) and i.r. (C) spectra of the pteridine wave IV reduction product. Mass spectrum obtained at 150°C and an ionizing voltage of 75 eV. N.m.r. solvent 0.5 M DC1 in D20. I.r. is of KBr disc.

their chemical shift in the wave I product (Fig. 8B). In the dimer (VI) the Cv protons are deshielded by the attachment of the second pyrazine ring, and thus appear downfield of the C7 protons in the dihydro compound. As expected the C2, C4 and C6 protons have very similar chemical shifts in the dimer and dihydro compound particularly when the different solvent systems employed for the spectra are taken into account. The broad appearance of the n.m.r, peaks may be caused by the chemical equilibria between the protonated form of the dihydro compound and the free base. The i.r. spectrum of the wave IV reduction product (Fig. 10C) is very similar to that of the wave I product (Fig. 8C), and the same interpretation in general applies. It should be noted that the band at 800 cm-1 (C4-H) appears in Fig. 10C. Although the dihydropteridine wave IV product is probably 7,8-dihydropteridine (XI), the 5,6-dihydro derivative (XII) would be expected to have almost identical physical and chemical properties. A decision between these two isomers could probably be made only on the basis of X-ray crystallographic data or b"

92

D . L . M c A L L I S T E R , G. D R Y H U R S T H I

N,,,'~xH I H

comparison with authentic samples. Unfortunately, sufficient purification then crystallization could not be satisfactorily effected, and neither 5,6- nor 7,8-dihydropteridine has been previously reported. Available data, however, allow other dihydro structures to be eliminated as possibilities. The preparation and properties of 3,4dihydropteridine have been published 22 and differ considerably from those of the electrochemical reduction product. Also, electrochemical reduction of the pyrimidine ring has never been observed with any pteridine. In addition, if the final reduction product was 5,8-dihydropteridine, the C6 and C7 vinyl protons would be expected to appear further downfield than the methylene protons of 7,8- and 5,6-dihydropteridine, and no signal for C6 (or C7) would appear at 6=7.8 (Fig. 10B). An additional argument for assignment of the wave IV product structure as 7,8-dihydropteridine rather than 5,6-dihydropteridine is based on the electrochemical behavior of other pteridines, in which the 7,8-dihydro species is always formed, if the structure of the starting compound so permits 18-2°. Reasonable explanations of the major peaks observed on mass spectrometry of the wave IV product on the basis of structure XI are possible and are presented elsewhere 2a. KINETIC STUDIES AND MECHANISM

Wave I

The evidence thus far presented indicates that pteridine is electrochemically reduced in a 2e, pH-dependent process. Analysis of the d.c. polarographic wave and slow sweep cyclic voltammetry indicate that the electron-transfer reaction is reversible. The heterogeneous rate constant for electron transfer, k~, was measured TABLE 2 HETEROGENEOUS

R A T E C O N S T A N T (k~) F O R T H E P E A K I c R E D U C T I O N

V o l t a g e sweep rate 1 V s-1, p t e r i d i n e c o n c e n t r a t i o n 1 m M . pH

3.8 5.6 7.0 9.0 11.0

kjcm s- i HMDE a

PGE b

3.3 x 2.7x 2.5× 3.6 x 2.0 x

l.Tx 1.2x 1.0x 1.3 x 1.1 x

10-2 10 - 2 10 -2 10- 2 10 -2

10 2 10 - 2 10 - 2 10- 2 10 -2

" H a n g i n g m e r c u r y d r o p electrode, area = 0.022 c m 1. b P y r o l y t i c g r a p h i t e electrode, area = 0.12 c m z.

OF PTERIDINE

ELECTROCHEMISTRY OF PTERIDINE

93

using the separation of anodic and cathodic peak potentials (AEp) under cyclic voltammetric conditions as described by Nicholson 23. The AEp values were measured at various pH values at sweep rates where the system showed peak separations in excess of reversible behavior (i.e., AEp > 30 mV). A sweep rate of 1 V s-1 was employed for most studies since then z~Ep at all pH values is greater than 30 mV and complications caused by the follow-up chemical reaction were not noticeable. Typical values of ks are presented in Table 2, where it is clear that the value of the rate constant is, as noted qualitatively earlier, approximately twice as large at the H M D E as at the PGE. Cyclic voltammetry, parUcularly at slow sweep rates and high pH, coulometry and product identification clearly support the view that the product of the 2e electrode reaction reacts with pteridine to give a dihydro dimer. The basic reaction scheme for this process is shown in eqn. (3A,3B), where the electron-transfer is reversible and ox+ne Red+Ox

ks Red

(3A)

~Z

(3B)

kz

Ox, Red and Z are soluble in the solution phase. The homogeneous reaction (eqn. 3B) is irreversible and characterized by rate constant k2. The product Z (dihydro dimer) is electroinactive over the potential range of interest with the Ox/Red couple. As indicated earlier by cyclic voltammetry the follow-up chemical reaction (eqn. 3B) occurs more rapidly at higher pH. Thus at pH 9 and 10, the reaction of electrode product with pteridine is quite fast and, at slow sweep rates, decreases the amount ofpteridine available for reduction at and near the electrode surface. This effect causes the quantity ip/V ½ (Table 1) to be appreciably lower than at pH 7 or 8. At fast sweep rates, however, the voltammetric experiment is complete before significant amounts of starting material can react, resulting in the larger values of ip/v ~. The fact that polarographic wave I at pH > 8 was smaller than at pH 5-7 (Fig. 2B) but was still under diffusion control as evidenced by constancy of il/h~orr as a function of h .... (vide supra) is still in accord with this scheme since all pteridine reaching the electrode at high pH would still arrive by pure diffusion. However, irreversible reaction of some pteridine diffusing to the electrode with product diffusing from the electrode would reduce the supply of pteridine for reduction, and hence decrease the limiting current. The decreased limiting current however, should still exhibit all the characteristics of a diffusion controlled process. In view of the relatively slow electron transfer reaction (Table 2) the solution kinetics were measured by the double potential step chronoamperometric method of Bard et al. 24. Plots of the normalized experimental current ratio R~, vs. t~ the real time (s) at which R~= 0.5 (see reference 24 for definition of terms) between pH 7 and 11 were in close agreement with that expected for a second order e.c. mechanism, i.e., that shown in equations 3A and 3B. For example, at TF=0.5 t~, R~ was 0.68 (theoretical for the latter mechanism is 0.69), at Tv= 1.5 t~, R~ was 0.30 (thcoretical 0.375), at Tv = 2.5 t~, RI was 0.235 (theoretical 0.22). Typical results for k2 obtained at PGE are presented in Table 3. Similar results were obtained at the H M D E but the latter electrode was not used extensively because of the difficulty of maintaining linear diffusion conditions at times greater than ca. 1-2 s.

94

D . L . M c A L L I S T E R , G. D R Y H U R S T

TABLE 3 VALUES OF k2 FOR PTERIDINE DETERMINED AMPEROMETRY AT THE PGE"

BY D O U B L E P O T E N T I A L

pH

Pteridine concentration ~tool 1 1

k~2"c/1 tool 1 s 1

7.0 8.0 8,0 9.0 9.9 11.0

5.30 x 6.11 × 3.99 x 1.17 x 5.19 x 1.99 x

1.1 1.7 1.7 4.4 2.2 4.1

10 3 10 - 4 10 -3 10 -3 10 ~ 10 4

x x x x x x

STEP CHRONO-

102 102 102 102 103 103

" E l e c t r o d e area 0.126 cm z. b C a l c u l a t e d b y the m e t h o d of B a r d et al. 24. c R e p r o d u c i b i l i t y of rate c o n s t a n t s was a l w a y s better t h a n + 10°/£.

Clearly, voltammetric and polarographic data indicate that electrochemical reduction ofpteridine at potentials corresponding to polarographic wave I proceeds by an electrochemically reversible process involving 2e and two protons. Studies of the electrochemical reduction of 2-amino-4-hydroxypteridines 18.19, folic acid 2° and quinoxalines 25-27 have shown that these compounds are also reduced in an initial 2e-2H + reversible process to the 5,8-dihydro derivative which then rearranges to the more stable 7,8-dihydro compound. It is likely that the initial reversible reduction ofpteridine proceeds in a similar fashion. That the initial reduction of pteridine occurs in the pyrazine ring rather than in the pyrimidine ring is supported by the fact that neither 5,6,7,8-tetrahydropteridine nor 5,6,7,8-tetrahydro-6,7-dihydroxypteridine is electrochemically reducible 28, i.e., the pyrimidine ring is not reducible. Further evidence favoring the initial formation of 5,8-dihydropteridine was obtained from studies of the electrochemical reduction of 6-hydroxypteridine (XIII) H

,

½kN I H

and 7-hydroxypteridine ( X I V ) 6. The latter compounds are electrochemically reduced in irreversible 2e-2H + processes to the 7,8- and 5,6-dihydro compounds, respectively. It is apparent from the structures of (XIII) and (XIV) that the 5,8-dihydro compound cannot be formed on reduction and hence a reversible reduction is not observed, i.e., the reversibility of the initial 2e-2H + reductions appears to depend on the ability to form the 5,8-dihydro derivative. Cyclic voltammetry of pteridine (I, Fig. 11) reveals that, following the initial 2e - 2H + reduction to 5,8-dihydropteridine (XV, Fig. 11) characterized by the rate constant ks (Table 2), the latter compound reacts with pteridine producing a dihydro dimer, 7,7',8,8'-tetrahydro-7,7'-dipteridyl (VI, Fig. 11) in a base catalyzed second order reaction characterized by the rate constant k 2 (Table 3). The nature of

" 95

ELECTROCHEMISTRY OF PTERIDINE

WaveI N~N~

H

+ 2H + +2e (IINr ) H

(I)

(I)

N~/'N~

~N~*~N

l-;q?

WaveIX

._ N ~ I " ~

+ 2H + +2e ~N~*~N

WaveIn'

~

(~1H

(rB

H

,O-N

~.~.N.~ .

+ 2H ÷ +2e ('~')

-

~..~N.,,,,,I~,~N..,j, + OH ('~Z)] L

N~N~ Wave]~r

"

%'¢'N)."

Fig. 11. P r o p o s e d electrochemical reduction scheme for pteridine.

the dimer has not been unambiguously established, but evidence from n.m.r., i.r. and mass spectrometry and electrochemical studies (vide supra) is consistent with the formation of the species shown in Fig. 11. The chemical reaction resulting in the formation of the dimer (VI, Fig. 11) is a base catalyzed Michael addition, similar to the reaction between 7,8-dihydro-6-hydroxypteridine and 6-hydroxypteridine 29, and to the reaction of 6-hydroxy-7-methyl-7,8-dihydropteridine with 6-methyl-7hydroxypteridine 30. A probable mechanism for the dimerization reaction is shown in Fig. 12. The voltammetric oxidation peak observed after reduction of pteridine at wave i or wave II potentials (e.9., peak IIa, Fig. 4) is clearly due to oxidation of the

96

D . L . McALLISTER, G. D R Y H U R S T H

I H

-

"

I H

- - -

H

<'aNJ

I H

H

N~ , , , , ~

N ~--.~ ¢ ~ N "-,,,¢.,."I~N

I

H

H

H

I

H

Fig. 12. Proposed mechanism for the reaction of 5,8-dihydropteridine with pteridine to form 7,7',8,8'tetrahydro-7,7'-dipteridyl.

dihydro dimer. Cyclic voltammetry revealed that the oxidation reaction resultecl in reformation of pteridine. Wave I I

Cyclic voitammetric and coulometric evidence suggests that reduction of pteridine on the plateau of wave II in acidic solution is an irreversible process whereby the 3,4-hydrate of pteridine is reduced to give products and n-values identical to those obtained by reduction on the plateau of wave I at higher pH. The reaction hence must involve reduction of the 3,4-hydrate (II, Fig. 11) with loss of water to 5,8-dihydropteridine (XV, Fig. 11), which then reacts with pteridine as before. Because of the fact that some non-hydrated pteridine must be reduced at wave II potentials and the test solutions contain both hydrated and non-hydrated pteridine which may both react with 5,8-dihydropteridine to give the dimer, rate constants for the follow-up reaction in acid solution where wave II is relatively large were not measured. Wave I I I and wave I V

It will be recalled that direct reduction of pteridine on the plateau of wave III at pH 12 is a 2e process which gives the same product as is observed when pteridine is first reduced at wave I or wave II potentials followed by further reduction at wave IV. Wave IV is clearly due to a 2e reduction of the dihydro dimer of pteridine (VI, Fig. 11) to give two molecules of 7,8-dihydropteridine (XI, Fig. 11). Wave III therefore must involve reduction of the anion of the 3,4-hydrate of pteridine (V, Fig. 11) in a 2e reaction directly to a dihydro derivative. On the basis of the evidence available it is not possible to specify whether the reaction proceeds through the 5,8-dihydro derivative and dimer to 7,8-dihydropteridine or whether a direct reduction to the latter species occurs. It will be recalled, however, that the rate of the dimerization reaction increases with increasing pH and accordingly it is probable that

ELECTROCHEMISTRY OF PTERIDINE

97

5,8-dihydropteridine is formed by reduction of the anion of the 3,4-hydrate of pteridine and that this rapidly dimerizes and undergoes a further reduction to 7,8-dihydropteridine as shown in Fig. 11. The voltammetric oxidation peak observed after reduction of pteridine at wave III or wave IV potentials is due to oxidation of 7,8-dihydropteridine. This was readily shown by cyclic voltammetric experiments when having scanned the oxidation peak the characteristic pteridine peaks were observed on the subsequent sweep to negative potentials. The fact that electrochemical oxidation of 7,8-dihydropteridine and 7,7',8,8'-tetrahydro-7,7'-dipteridyl (dimer) occurs at exactly the same potential and particularly that the 2max of the u.v. spectrum of both compounds are identical is not surprising. Thus, the single pyrazine ~-C=N- double bond remaining in both the dimer (VI, Fig. 11) and 7,8-dihydropteridine (XI, Fig. 11 ) is conjugated in exactly the same way for both compounds with the unsaturated pyrimidine ring which accounts for the 2 .... values. Identical 2n~x values for such structurally related dimers and dihydro compounds have been reported for various dimeric and dihydro pyrimidines31. Wave V

Because reduction on the plateau of wave V results in the transfer of ca. 0.2 to 0.3 electrons with no change in the u.v. spectrum of the solution, it was concluded that this wave is not associated with the principal reduction schemes, but instead represents reduction of perhaps a very minor wave I product. ACKNOWLEDGEMENTS

The authors would like to thank the National Science Foundation and the Faculty Research Committee of the University of Oklahoma for financial support of this work. SUMMARY

Over the pH range 1-12, pteridine is reduced at the dropping mercury electrode (DME) and pyrolytic graphite electrode (PGE) by way of three polarographic or voltammetric processes. The first, least negative process, is a reversible 2 e ~ H + reduction ofpteridine to 5,8-dihydropteridine. The latter species reacts with pteridine in a base-catalyzed Michael reaction producing a dihydro dimer, probably 7,7',8,8'-tetrahydro-7,7'-dipteridyl. The second process involves reduction of the monohydrated form of pteridine (3,4-dihydro-4-hydroxypteridine) which exists in major amounts at low pH which is also reduced in a 2e-2H + reaction to 5,8dihydropteridine but in an irreversible process. Again the latter compound reacts with pteridine to give a dimer. The third process is reduction of the anion of 3,4dihydro-4-hydroxypteridine in the same overall process as is observed for the two previous reactions forming, ultimately, the same dihydropteridine dimer. The latter dimer is also reducible electrochemically in a 2 e ~ H + irreversible process to give 7,8-dihydropteridine. Rate constants have been obtained for the reversible electron transfer and for the dimerization reaction of 5,8-dihydropteridine with pteridine.

98

D.L. McALLISTER, G. DRYHURST

REFERENCES 1 J. Komenda and D. Laskafeld, Collect, Czech. Chem. Commun., 27 (1962) 199. 2 A. J, Beaman, J. F. Gerster and R. K. Robins, J. Org. Chem., 27 (1962) 986. 3 A. J, Beaman and R. K. Robins, J. Amer. Chem. Soc., 83 (1961) 4038. 4 D. J. Brown, Chem. Ind., 69 (1950) 353. 5 A. Albert and H. Yamomoto, J. Chem. Sot., C (1968) 2289. 6 D. L. McAllister and G. Dryhurst, J. Electroanal. Chem., 47 (1973) 479. 7 D. D. Perrin, J. Chem. Soc., (1962) 645. 8 A. Albert, T. J. Batterham and J. J. MacCormack, J. Chem. Soc., B (1966) 1105. 9 L. Meites, Polarographic Techniques, Wiley, New York, 2nd edn., 1965, p. 219. 10 Reference 9, pp. 132, 133, 140. 11 P. Delahay, New Instrumental Methods in Electrochemistry, Wiley, New York, 1954, Chap. 3. 12 H. Matsuda and Y. Ayahe, Z. Elektrochem., 59 (1955) 494. 13 R. S. Nicholson and 1. Shain, Anal. Chem., 36 (1964) 706. 14 I. Shain and K. Martin, J. Phys. Chem., 65 (1961) 254. 15 R. S. Nicholson and 1. Shain, Anal. Chem., 37 (1965) 178. 16 L. Meites, Polarographic Techniques, Wiley, New York, 2nd edn., 1965, p. 240. 17 S. F. Mason, J. Chem. Soc., (1955) 2336. 18 H. Braun and W. Pfleiderer, Ann., (1973) 1802. 19 S. Kwee and H. Lurid, Biochim. Biophys. Acta, 297 (1972) 285. 20 K. Kretzschmer and W. Jaenicke, Z. Naturforsch., 26B (1971) 999. 2l D. L. McAllister, Ph.D. Dissertation, Oklahoma, 1973. 22 A. Albert and K. Ohta, J. Chem. Soc., (1970) -1540. 23 R. S. Nicholson, Anal. Chem., 37 (1965) 1351. 24 W. V. Childs, J. T. Maloy, C. P. Keszthelyi and A. J. Bard, J. Electrochem. Soc., 118 (1971) 874. 25 T. Goto, A. Tatematsu and S. Matsuura, J. Org. Chem., 30 (1965) 1844. 26 M. Strier and J. Cavagnol, J. Amer. Chem. Soc., 79 (1957) 433. 27 J. Pinson and J. Armand, Collect. Czech. Chem. Commun., 36 (1971) 385. 28 E. C. Taylor and W. R. Sherman, J, Amer. Chem. Soc., 81 (1959) 2464. 29 A. Albert, J. Chem. Soc., (1955) 2690. 30 A. Albert and E. D. Sergeant, J. Chem. Soc., (1964) 3357. 31 D. L. Smith and P. J. Elving, J. Amer, Chem. Soc., 84 (1962) 2741.