Electrochemical oxidation of 2-phenyl-1,2-benzisothiazol-3(2H)-ones and related compounds in acetonitrile. A study using microelectrodes

Pergamon 0013-4686(94)E0152-P

Electrochimica Acta, Vol. 39, No. 13, pp. 1961-1969, 1994 Copyright ~) 1994 Elsevier Science Ltd. Printed in Great Britain. All rights racrvcal 0013-4686/94 $7.00 + 0.00

ELECTROCHEMICAL OXIDATION OF 2-PHENYL-1,2BENZISOTHIAZOL-3(2H)-ONES AND RELATED COMPOUNDS IN ACETONITRILE. A STUDY USING MICROELECTRODES R. MOLLER,* B. DAKOVA,* L. LAMBER'IS*?and M. EVERS:I: *Facult6s Universitaires Notre-Dame de la Paix, Laboratoire d'Electrochimie et de Chimie Analytique, 61, rue de Bruxelles, B-5000 Namur, Belgium ~/Rh6ne-Poulenc Rorer S.A., Centre de Recherche de Vitry, P.O. Box 14, F-94403, Vitry sur Seine, France (Received 4 January 1994; in revisedform 7 February 1994)

Abstract--The electrochemical oxidation of 2-phenyl-l,2-benzisothiazoi-3(2H)-ones and related compounds was studied by conventional and microelectrode techniques. Hammett plots were drawn for a series of N-aryl substituted 2-phenyl-l,2-benzisothiazol-3-(2H)-ones and compared with those of the selenium analogues, 2-phenyl-l,2-benzisoselenazol-3(2H)-ones. Controlled potential electrolysis furnished the corresponding sulphoxides and the reaction pathway was shown to proceed through an ECE mechanism. Key words: 2-phenyl-l,2-benzisothiazol-3(2H)-one, electrochemical oxidation, sulphur compounds, microelectrodes, radical-cations.

INTRODUCTION In a previous p a p e r [ l ] we have shown that Nphenylbenzisothiazol-3-(2H)-one ld studied with a rotating disk electrode gave rise to two signals whose limiting current intensities correspond to a two electron transfer. Interestingly, the ratio of -/1. lira/./2,lira is not equal to unity even for rotation speeds up to 8000 rpm. Furthermore, cyclic voltammetry experiments performed with stationary platinum electrode showed an irreversible signal even for scan-rates up to 25 V s - 1. This particular behaviour of N-phenylbenzisothiazol-one ld prompted us to investigate in more detail the electrochemical behaviour of a set of 21 selected molecules belonging to this class of heterocycles. Steady-state[2-4] and transient[5, 6] microelectrode techniques used in common with classical electrochemical methods, have been combined in order to complete our previous study and to elucidate the electrochemical oxidation mechanism of several derivatives.

EXPERIMENTAL

Anhydrous sodium perchlorate (Janssen Chimica) was recrystallized three times from absolute ethanol[7] and dried under vacuum at 100°C for 12 h. Tetraethylammonium tetrafluoroborate (Fluka, puriss) was directly dried under vacuum at 150°C for 12 h. Ferrocene (Fluka, purum) was purified by sublimation. In order to minimize the influence of residual water, all measurements were done inside a plexiglass dry-box filled with nitrogen[8]. Apparatus

Controlled potential coulometries, large scale electrolysis, chronoamperometries and low speed cyclic voltammetrics (potential scan speed v~<20Vs -1) were performed with a PAR potentiostat model 273 using a three electrode ceil. Higher scan rates in cyclic voltammetry were obtained by using a similar experimental configuration as reported in [9]: a function generator (Newtronics 200TPC) furnished the potential ramp to the PAR 273 and the current was recorded with a digital oscilloscope (Philips PM3350 A, 8 bits). Data treatment was realized on an IBM XT computer. Low-noise steady-state voltammograms at microelectrodes were obtained using the line-synchronization command of the PAR 273; the utilization of a Faraday cage was not necessary.

Solvents and reagents

Electrodes

HPLC grade acetonitrile (Rathburn, grade S) and analytical grade dichioromethane were dried by passing through a column of activated neutral alumina (Aldrich, type Brockman I ) j u s t before use.

Microelectrodes were prepared by sealing fine platinum wires (Goodfellow, 5 and 25pm radius, 99.99% purity) into glass capillaries (o.d.: 6mm, i.d.: I mm; length: 10cm) with a home-built electric furnace. For high speed cyclic voltammetry the microelectrode radius was chosen in accordance with the potential scan rate employed in order to obtain

? Author to whom correspondence should be addressed.

1961

1962

R. MOLLERet al.

Nicholson-Shain type voltammograms with negligible distortion from both ohmic drop and non-linear diffusion[10]. A 3 mm diameter platinum disc electrode (Metrohm 628-10) was used for conventional cyclic voltammetry and rotating disc measurements. Coulometries and electrolysis were carried out at a cylindrical platinum gauze working electrode (diameter: 3.6crn, height: 2era). The reference electrode was Ag/AgC1 KCI 3 M (Ingold); the inner compartment was filled with the same solvent and supporting electrolyte as the electrochemical cell. The auxiliary electrode was a platinum mesh of about 2 em 2, separated from solution by a Vycor frit (EG&G PAR) during bulk electrolysis. Product analysis

After completion of the electrolysis, the solvent was evaporated under vacuum at low temperature (30-40°C). The electrolysis products were extracted from the residual solid with ether, dried over CaCI2 and finally crystallized. Spectroscopic characteristics and melting points (m.p.) of sulphoxides 2a, c, d, f and i are in good accordance with the published values[11]: 2-(4-Mcthoxyphenyl)- 1,2-benzisothiazol-3(2H)-one l-oxide (2a): ir (KBr) Vc-o = 1716em -1, vs-o = 1512cm -1, V c - o - c = 1253crn -1 and Vs-o = 1094cm -1 (lit. [11] 1715, 1515, 1260 and ll00cm -1, resp.), m.p.: 140-142°C (lit. 147-148°C [11]). 2-(4-Methyiphenyl)- 1,2-benzisothiazol-3(2H)-one 1oxide (2e): ir: (KBr) Vc_o = 1717era -1, Vs_o = 1311era -1 and 1094crn -1 (lit. [11] 1710, 1315 and ll00cm -1, resp.), m.p.: 169-171°C (lit. 167.5168.5°C [11]). 2-Phenyl- 1,2-benzisothiazol-3(2H)-one 1-oxide (2d): it: (KBr) v c _ o = 1 6 8 4 c m -1 and Vs_o = 1097cm -1 (lit. [11] 1685 and 1100cm -1, resp.). m.p.: 130-132°C (lit. 136-137°C [11]). 2-(4-Chlorophenyl)- 1,2-benzisothiazol-3(2H)-one 1oxide (2t): it: (KBr) Vc-o = 1711, vs-o = 1495, 1306 and 1096cm -1 0it. [11] 1720, 1495, 1310 and 1095cm -1 resp.), m.p.: 139-140°C (lit. 139-140°C [ll]). 2-(4-Cyanophenyl)- 1,2-benzisothiazol-3(2H)-one 1oxide (2i): ir (KBr) v c . s = 2 2 3 8 c m - l , vc_o = 1712cm -1, v s - o = 1295cm -1 and 1087cm -1 (lit. [ l l ] 2240, 1720, 1300 and 1090cm -1, resp.), m.p.: 145-146°C (lit. 146-147°C [11]). Spectroscopic characteristics and melting points (m.p.) of four unpublished representative sulphoxides ~ , g, h and j synthesized by chemical oxidation according to reference [11] are given below: 2-(4-Fluorophenyi- 1,2-benzisothiazol-3(2H)-one 1oxide (2e): Jr: (KBr) vc_o = 1685cm -1, Vs_o = 1489, 1317 and 1102 cm- 1. m.p.: 147-148°C. 2-(4-Bromophenyl)- 1,2-benzisothiazol-3(2H)-one 1oxide (2g): it: (KBr) Vc_o = 1699cm -1, Vs_o = 1508, 1313 and 1099cm-1. m.p.: 125-127°C. 2-(4-Trifluoromethylphenyl)- 1,2-benzisothiazol3(2H)-one 1-oxide (2h): it: (KBr) vc-o = 1718 era-1, vs-o = 1325 and 1095 era-1, m.p.: 130-132°C. 2-(4-Nitrophenyl)- 1,2-benzisothiazoi-3(2H)-one 1oxide (2j): it: (KBr) v c - o = 1718 crn-1, vs-o = 1302 and 1086cm-1. m.p.: 177-178°C.

RESULTS AND DISCUSSIONS Typical voltammetric experiments were carried out with a 1 x I0 -a M solution of the depolarizer in acetonitrile containing 0.I M of sodium perchlorate. Half-wave and peak potentials were measured for molecules l(a--s) with different electrodes and are summarized in Table I. Investigation of the positive potential side on rotating platinum disc electrode showed one welldefined anodic wave for all the compounds, except Id, lq and Ir which gave rise to a supplementary wave. The limiting current intensity (total Ium for Id, lq and Ir ) agreed with Levich's diffusion criteria (luJca I/2) within the rotation range studied. By increasing the rotation speed from 500 to 3000rpm the half-wave potentials slightly shifted towards more positive potentials. Figure la presents the cyclic voltammetry achieved with a stationary platinum electrode exhibiting two irreversible and diffusion-controlled (Ip/vI/2)peaks within the range of scan rate varying from 0.05 to 0.5 V s-i. Only compounds Id, lq and Ir gave rise to three anodic peaks. Figure 2 presents two steady-state voltammograms recorded under the same experimental conditions illustratingthis difference in behaviour of unsubstituted benzisothiazol-3-(2H)-one Id compared with the 2-(4-chlorophenyl)-l,2-benzisothiazol3(2H)-one If taken as a representative example. The relationship between the limiting current (11i~), obtained for a microdisc electrode[12, 13]

(1)

l l i m = 4nFCo Dr

and the number of electrons exchanged per molecule (n), the microdisc radius (r), the depolarizer's bulk concentration (Co) and the diffusion coefficient (D), allows us to determine the number of electrons exchanged per molecule. Since the limiting currents for Id and If measured under identical experimental conditions are nearly the same (Fig. 2) and because the corresponding diffusion coefficients for those two molecules are very similar, the totalnumber of electrons exchanged per

(a)

T50 ~A

~ ' ~

(b)

I 0.... 0'5.... 0 0 .... 0 5.... I'0.... I'5.... 2 0 E / V vs. Ag/AgCI Fig. I. Cyclic voRammograms taken before (a),during (b, I F mol- i) and afterelectrolysis(c,2 F rnol-t)of 2.0m M of 2-(4-methylphenyl)-l,2-benzisothiazol-3(2H)-oneIc in dry C H 3 C N (0.IM NaCIO,). Pt electrode,r = 1.5mm radius, v = 50mVs -t.

A microelectrode study of electrochemical oxidation

1963

I

o.

2.

i i

e~ @

@

~ H u.l

rl

.= ~

0

Oi

tt , ~2 ~

0

~

~

~

.

.

.

.

.

.

.

=Z

O~

I

@

='.=-6 2=

'-~ ~ 0

~ o , o,

t~

OZo @

0

~uimooZoD~.o~,~

~-]

R. MOLLERet al.

1964 3.0 10 ,2 2.5 10 .2

~ 2.010 -~ E ~1.5

/

lO'"

--. 1.0 10" 5.0 10 .2

j

~

f

ld

/ !' / J

0.0 101. ~ ' ' '1'.3 ' ' '1'.4 ' " 1'.5 '

'1'.6 ' ' '1'.7 " " "1:8 ' ''1.9

E / V vs. Ag/AgCI

Fig. 2. Steady-state voltammograms of 2-phenyl-l,2benzisothiazol-3(2H)-one ld and of an N-aryl substituted compound: 2-(4-chlorophenyl)-l,2-benzisothiazol-3(2H)one If in dry CH3CN (0.1 M NaCIO4). Pt electrode, r = 25pm, v = 5mVs -1. molecule can be assumed to be equal for ld and If, namely two, as already shown for ld in a previous paper[I]. The number of electrons exchanged per molecule at the first oxidation signal was also determined for compounds lc, e--g by controlled potential coulometry and chronoamperometry on disc electrodes in transient and steady-state modes[14]. These two techniques were chosen in order to ascertain that the electron number exchanged per molecule at long characteristic times (eg in coulometry, T~ ~ 15 rain) is

the same at short characteristic time-scales (eg in chronoamperometry, T~ ~ 0.5s), at which voltammetric experiments are realized. Results are quite similar and confirm a total exchange of two electrons per molecule. Treatment of the chronoamperometric data[12] also led to an estimation of the diffusion coefficient: D = 1.2 x 10- 5 cm 2 s - 1 at room temperature (22 _+ 1°C). Controlled potential electrolysis realized at the foot of the oxidation peak was followed by cyclic voltammetry. The general trend for la, e-j (Fig. 1) shows the vanishing of the investigated peak and the evolution to a better defined second peak (at about 1.8 V) which has about the same intensity. The sole signal observed on the reverse scan corresponds to the reduction of released protons as has been confirmed by addition of HCIO 4 (70%, aqueous)• This reduction is reversible, as has been ascertained by separate measurements with a 1 0 - 3 M solution of perchloric acid under identical conditions. It is noteworthy to mention that this reversibility completely disappeared when the experiments were performed outside the dry-box with undried acetonitrile and sodium perchlorate. The electrolysis product could not be reduced at potentials up to - 1 . 0 V . Infrared spectra and melting points are those of the corresponding sulphoxides (see "Product Analysis"). The overall reaction can thus be written as follows (Scheme 2):

a: b: c: d:

O

R= O-CH 3 RffiN(CH3) 2 RffiCH 3 R=H

e: R = F f: R = CI g: R = B r

7

1 1

7

Y

h: R= CF 3 i: R= C N

(a-j)

j:

R4

R= N O 2

k: R2= CI; RI=R3=R4= H 1: Rift CI; R2=R3=R4=H m: RI=R4= CI; R2=R3ffiH n: R3= CI; R2= CF3; RI=R4= H o: RI=R3= O-CH3; R2=R4ffiH P: R2=R3= O-CH3; RI=R4ffi H

NRI~R2R3 1 (k-p)

R1

O N 1

~

R3

q: Rt=CI; R2=R3=H R2=CI; R1=R3=H s: R2=H; RI=R3=CI

(q-s) O

S

3

4

Scheme 1

A microelectrode study of electrochemical oxidation O

1965

O N

R

='

N

R

-2W- 2e II

l a , ¢-j

O

2a, c-j

Scheme 2 Further oxidation of the sulphoxides at the potential of the second anodic peak in Fig. la should result in the formation of the corresponding sulphones. Effectively, the same signal was obtained with the chemically synthesized sulphoxide 2c. Controlled potential electrolysis however does not allow their separation under our experimental conditions, as has already been reported previously[l] in the case of ld.

Correlation of the first oxidation potential with the a + Hammett substituent constant A good relationship between the first oxidation potential (Ev or E1/2) and the Hammett substituent constant values[15-1 is observed for the N-aryl substituted compounds substituted at the para position la, c, e--g, i and j. The a + parameter was chosen for correlation because the reaction centre is directly conjugated with the substituent and since a positive charge is developed in the transition state. Peak and half-wave potentials fitted well the modified Hammett equation AEp or AEt/2=p,,.ra +, and showed a correct ranking in relation to the reaction constant P,.r (Fig. 3). By increasing the electrophilic ability (tr +) of the substituent electron abstraction becomes more difficult and the oxidation potential shifted anodically. The use of the a + parameter instead of a~ gives a better curve fit and permits the inclusion of the methoxy substituted compound. The 2-(4dimethylamino-phenyl)-1,2-benzisothiazol-3(2H)-one

lb was not included in the Hammett plot because its first oxidation potential corresponds to the formation of a nitrogen cation-radical as will be shown below. Since a Hammett plot has already been described for the selenium analogues of the studied compounds[16] it was interesting to establish a comparison with our results (Fig. 3). The sulphur compounds are about 120mV more difficult to oxidize and seem to exhibit a greater substituent influence than their selenium analogues. These two observations can be explained taking into account the atom sizes: selenium is bigger and its lone-pairs are more distant from the nucleus. This results in a decreased influence of the substituents and gives rise to an easier oxidation. The existence of a linear relationship with a slope near unity (0.93 _0.11) between benzisothiazol-3(2H)-ones and their seleno analogues (Fig. 4) confirms that in the two groups of heterocycles and between them, the substituent effects are transmitted by the same kind of mechanism. However, the sulphur atom--probably because of its lower polarizability--feels the substituent effects to a less extent compared to selenium.

Study of the electrochemical reaction mechanism by high speed cyclic voltammetry In the preceding paper of this series[l,1 we have attributed the presence of the second oxidation signal of 2-phenyl-l,2-benzisothiazol-3(2H)-oneld to an unspecified chemical reaction, competing with the

1.60

•

sulfur

p-NO:

- - o- - selenium

1.55

/

-

1.50

~ "~

/

1.45

/

/ >

r PiH

p-NO2 ./c

p'CH3 ~

/

/

~"

t- ~ . . p-CN

1.40

p-OCH3 ~=" 1.35 / 1.30

1.2_I 8

" / p - C H3

/ /

-o6

o'.4

o'.2

olo

0'2

0'.4

0'.6

0.8

G+ Fig. 3. Comparison of Hammett plots for N-aryl substituted 2-phenyl-l,2-benzisothiazol-3(2H)-onesand their selenium analogues: 2-phenyl-l,2-benzisoselenazol-3(2H)-ones.Pt electrode, r= 1.5mm radius, v = 5 0 m V s - 1 , CH3CN (0.1M NaCIO4). Linear regression (expressed as mean + standard deviation) gives: Ep = (1.488 _+0.002) + (0.1349 + 0.0038) × ~+ (r2 = 0.994) for the sulphur compounds (ld excepted) and Ev = (1.361 + 0.002) + (0.1158 + 0.0042) × a+ (r2 = 0.993) for the selenium analogues.

R. MOLTER et al.

1966 0.15 - - y

=-0.033

+ 0.929x

r2=0.929

0.10

•

o.o5

IN

~

o.00

-0.05

-o.,o'

-'0".

....

, ................... 0.00

0

-0.05

0.05

0.10

0.15

A(E 1 x"E1 d)' S Fig. 4. Relationship between the benzisothiazol-3(21t)-ones and the benzisoselenazol-3{2H)-ones.

2.( / (

1.:

I.(

/"

y~

/

/

02 :' 0<

j

j//

i

-02

-i.~.4 .. o'.~ o'.~ " o'.7' " ~.S

0'.9 i'.0 " ;.i

~1.2

E / V vs. AI/AIICI Fig. 5. Cyclic voltammogram of 2.00raM 2-phenyl-l,2benzisothiazol-3(2H)-one Id in dry CH3CN (0.5M NaCIO,). Pt electrode, r = 25~m, v = 1.50kVs -~.

0.4

02

0.~

/

"~

0.~

,~:

,., 0.; "

S

0.1

/. r / ..../

0.0

-0.1 -0,

~.j// "b~9~16"£1

.... i'.2113 .... 114"11S'"1.~ E / V vl, AI/AIICI

Fig. 6. Cyclic voltammogram of 2.0raM 2-(4-methylphenyl)-l,2-benzisothiazol-3(2H)-one I¢ in dry CH3CN (0.5 M NaCIO,). Pt electrode, r = 25 pro, v = 165 V s- x.

formation of the sulphoxide. This was clearly shown through water addition to the sample: the second signal diminished while the first one increased. Since a similar evolution was observed when the potential scan rate in cyclic voltammetry is raised from 0.I to 25Vs -1, we have now studied the bchaviour at potential scan rates up to several kV s - t. By increasing the scan velocity the second oxidation peak of ld diminished in intensity compared to the first one and finally vanished completely. At the same time the cathodic counterpart of the first oxidation peak appeared and the voltammogram reached electrochemical reversibility at about 1.5 kV s - 1 (Fig. 5). So, the present microelectrode study confirms our hypothesis emitted in [1]: the first signal of ld corresponds to the reversible monoelectronic oxidation of the neutral molecule into its radical-cation. Compounds lq and l r which are not substituted at the N-aryl behave in the same manner at even high potential scan rates. High-speed cyclic voltammetry of the N-aryl substituted compounds l a - ¢ and e-p also permitted the detection of the corresponding cation-radicals. If one excludes 2-(4-dimethylaminophenyl)-1,2benzisothiazol-3(2H)-one lb (see below) and the methoxy substituted compounds la, o and p at which surface filming phenomena prohibited a closer study, the scan rate necessary to achieve reversibility varies between 100 V s - 1 for electro-donating groups (le) and 3 kVs-~ for electroattracting substituents (li and i). Since the ratio, faradaic current/double layer charging current, is proportional to v- 1/2['17] a compound which is electrochemically reversible at relatively low scan-rate, 2-(4-methylphenyl)-l,2benzisothiazol-3(2H)-one lc, (Fig. 6) was chosen for a more detailed investigation of the electrochemical reaction mechanism. At low speeds, the normalized current (lp, v- 1/2) is about 2.17 times greater than at high speed (Fig. 7);

A microelectrode study of electrochemical oxidation

1967

2.0 10" 1.8 10 . 4 1.6 10 "4 ,.-= 1.4 1 0 -4

~ •

1.2 1 0 - 4 1.0 1 0 -4 8.0 10 . 5

I

i

,

.

2 0 1 log (v / V . s ' l ) Fig. 7. Evolution of the peak current normalized by the square root of the potential scan rate (v) in function of the logarithm of the latter for the oxidation of a solution containing 1.95 mM of 2-(4-methylphenyl)-l,2-benzisothiazol-3(2H)-one lc and 0.5 M of NaC104 in dry CH3CN. Pt electrode, r = 1.5mm, ohmic drop compensation: 80 fl. 2

-1

this is quite close to the theoretical value of 2.22 expected for an ECE mechanism. The variation can be explained using the following formula[14] which gives the peak current in function of several parameters: Ip = (n2(T¢) +

•,

~,I/ZDI/ZFAC']q r ~RT" v

(2)

"1J"1

In this equation, n I and nacre ) are the number of electrons exchanged per molecule in the first and second electron transfer, respectively, and f is a parameter depending on the reaction mechanism according to

[18] ( f = 0.4463 for an uncomplicated E mechanism and f = 0.4958 for an E C mechanism); all other symbols have their usual meanings. So, at low scan speed in cyclic voltammetry the chemical reaction occurs and is followed by the second electron transfer (nl = n2(Tc)= 1): lp V-1/2 = constant x 2 x 0.4958 while at high speed only the first electron transfer takes place (nl = 1, n2crc)= 0) and lp, v-t/2 = constant x 0.4463. The theoretical current ratio is then 2.22. Because low-speed cyclic voltammetry is realized in the same characteristic time as the chronoamperometric microelectrode measurements which

.+

o N

R _

N

}

E1° (E)

la, c-j

+HzO A

P

_H+ O -¢ h

(E2o < E1°) OH O

O D

.H+

s / N . ~ ~ R II

O 2a, e-j Scheme 3

(c)

R. Mf.)LLERet al.

1968 0.05

0.04 0.03 =3.

(,i

0.02

j,.

0.01 0.00

/

/

f " - ' ~

"-,../ :

-0.01

~

-0"0~'.0"

" '112'

'

"114'

'

'116"

"

118

'

'

" 2.0

E / V vs. Ag/AgCI

Fig. 8. Cyclic voltammetry of 0.93raM of 2-(4-methylphenyl)-l,2-bcnzisothiazol-3(2H)-one le in Et4NBF,, saturated dichloromethane (about 0.08M). Pt electrode, r = 25#m, v = 15.5Vs -t.

have shown an exchange of two electrons per molecule, we can affirm that the reversible signal observed at high speed takes place with an exchange of one electron per molecule and corresponds to the generation of the sulphur cation-radical. This intermediate remains unchanged in the time scale of high scan rates. At low scan speeds however it reacts with residual water present in the solvent and loses a proton. The radical formed undergoes then a further electron transfer, before transformation into the sulphoxide by losing a second proton (Scheme 3). Application of a formula given by Nicholson and Shain[19] to the data from Fig. 7 permitted us to compute a pseudo first-order rate constant of about 330s -1 for the reaction of the cation-radical of le with residual water, giving a half lifetime of 2 ms for 2-(4-methylphenyl)-1,2-benzisothiazol-3(2H)-one intermediate. As dry dichloromethane is known to stabilize cation-radicals[20, 21] several experiments were conducted in this medium. The oxidation of 2-(4-methylphenyl)-l,2-benzisothiazol-3(2H)-one le becomes effectively reversible at a potential scan rate about ten times lower than in acetonitrile (Fig. 8), indicat-

4C

2C

-2

-4.~

ing that the resulting intermediate is ten times more stable in dichloromethane. A systematic use of this solvent however was prohibited by adsorption phenomena and the high solution resistance, resulting in a non-negligible ohmic drop distortion of cyclic voltammograms. In order to investigate the eventually existing stabilizing effects of the carbonyl group on one hand the cyclization and on the other hand we have examined compounds 3 and 4. The study of these molecules indicates that the cation-radical formed on the sulphur atom is stabilized by the presence of the nitrogen lone-pair: 2-(4-chlorophenyl)-1,2benzisothiazol-3(2H)-one I f and 2-(4-chlorophenyl)1,2-benzisothiazol-3(2H)-thione 3 behave reversibly at high scan rates while 4-chloro-2-methylthio-benzanilide 4 stays irreversible even at the highest scan rate employed (10 kV s-1). The relatively high stability of such cation-radicals can be explained by an intramolecular sulphur-nitrogen three-electron bonding, consisting in the coordination of the unpaired p electron of the oxidized sulphur atom with the free electron pair of the nitrogen atom[22, 23]. Selenium compounds behave generally irreversibly in electrochemical oxidations because intermediates seem to react very rapidly with residual water contained in the solvent. In order to verify if oxidation is still irreversible at high scan rates we have studied one selenium compound, which by analogy with the corresponding 2-phenyi- 1,2-benzisothiazol-3(2H)one should produce a relatively stable cation-radical: 2-(4-methylphenyl)- 1,2-benzisoselenazol-3(2H)-one. No sign of reversibility was seen, however, up to 10kVs -1.

"Unusual" behaviour of 2-(4-dimethylaminophenyl)1,2-benzisothiazol-3(2H)-one lb In contrast to the other 2-phenyl- 1,2benzisothiazol-3-(2H)-ones la, ¢-s oxidation of the dimethylamino derivative lb takes place at the nitrogen and not at the sulphur atom. A monoelectronic exchange per molecule has been evidenced by constant potential coulometry and with help of an E vs. log[(llim - I)/I] plot (slope = 60.3 mV decade-1) realized on data from steady-state voltammetric microelectrode experiments. Electrolysis products formed a resinous film after solvent evaporation and could not be identified. The bluish-green coiour of the solution (persisting over days) however and the fact that cyclic voltammetries are already reversible (AEp = 70 mV) at speed as low as 50 mV s-1 (Fig. 9) indicates that the compound formed is similar to the cation-radical of N,N,N',N'-tetramethyl-p-phenylenediamine (Wurster's blue), which remains stable in solution over weeks[-24, 25]. CONCLUSION

- • 010'

' '0'.2'

' o'.4

' 0'.6'

' '018'

" ' 1'.0

' '1.2

E ! V vs. Ag/AgCI

Fig. 9. Cyclic voltammogram of 4.00mM 2-{4-dimetbylaminophenyi)-l,2-bcnzisothiazol-3(2H)-one lb in dry CH3CN (0.5M NaCIO4). Pt electrode, r = 1.5mm, V = 5 0 m V s -1.

Electrochemical oxidation of derivatives of 2phenyl-l,2-benzisothiazol-3(2H)-one in acetonitrile medium into the corresponding sulphoxides occurs by two monoelectronic transfers separated by a chemical reaction with residual water (ECE mechanism). This reaction scheme is complicated by

A microelectrode study of electrochemical oxidation an unidentified chemical reaction in the case of the compounds which are not substituted at the N-aryl (ld, l q and It). The study of 2-(4-methylphenyl)-l,2benzisothiazol-3(2H)-one le by high speed cyclic voltarnmetry permitted us to compute a half lifetime of 2 ms for the cation-radical. Intermediates of the other compounds are less stable. The relatively high half-lifetime of this type of sulphur cation-radicals can be attributed to stabilization effects by the electronic lone-pair of the adjacent nitrogen atom. Comparison of 2-phenyl-l,2-benzisothiazol-3(OH)ones with their selenium analogues has shown that the sulphur compounds oxidize less easily, and that the oxidation potential is more influenced by substituent effects than for the selenium analogues. Furthermore, the stability of reaction intermediates is several orders higher with the sulphur compounds than with the selenium analogues. Acknowledgements---One of us (R. M.) is grateful to the

"Institut pour l'Encouragement de la Recherche Scientifique dans rlndustrie et l'Agriculture" (IRSIA) for financial support in order to accomplish a doctoral thesis.

REFERENCES 1. B. Dakova, R. Miiiler, L. Lamberts and M. Evers, Electrochim. Acta 39, 661 (1994). 2. A. M. Bond, K. B. Oldham and C. G. Zoski, Anal. Chim. Acta 216, 177 (1989). 3. C. G. Zoski, A. M. Bond, C. L. Coyler, J. C. Myland and K. B. Oldham, J. electroanal. Chem. 263, 1 (1989). 4. C.G. Zoski, J. electroanal. Chem. 296, 317 (1990). 5. R. M. Wightman and D. O. Wipf, Acc. Chem. Res. 23, 64 (1990). 6. C. P. Andrieux, P. Hapiot and J-M. Sav6ant, Chem. Rev. 90, 723 (1990).

1969

7. M. D. Ryan, D. D. Swanson, R. S. Glass and G. S. Wilson, J. phys. Chem. 85, 1069 (1981). 8. P. A. F. White and S. E. Smith, Inert Atmospheres in the Chemical, Metallurgical and Atomic Industries. Butterworths, London (1962).

Energy

9. S. Daniele, P. Ugo, G. A. Mazzocchin and G. Bontempelli, J. electroanal. Chem. 267, 129 (1989). 10. C. Amatore and C. Lefrou, J. electroanal. Chem. 324, 33 (1992). 11. N. Kamigata, S. Hashimoto, M. Kobayashi and H. Nakanishi, Bull. chem. soc. Jpn. 158, 3131 (1985). 12. M. Fleischmann, S. Pons, D. R. Robinson and P. P. Schmidt, Ultramicroelectrodes. Datatech Systems, North Carolina (1987). 13. M. I. Montenegro, M. A. Queiros and J. L. Daschbach, Microelectrodes: Theory and Applications. Kluwer, Dordrecht (1991). 14. C. Amatore, M. Azzabi, P. Calas, A. Jutand, C. Lefrou and Y. Rollin, J. electroanal. Chem. 288, 45 (1990). 15. N. B. Shapman and J. Shorter, Correlation Analysis in Chemistry, p. 482. Plenum Press, London (1978). 16. B. Dakova, J.-M. Kauffmann, M. Evers, L. Lamberts and G. J. Patriarche, Electrochim. Acta 35, 1133 (1990). 17. A. J. Bard and L. R. Faulkner, Electrochemical Methods, Fundamentals and Applications, p. 220. Wiley, New York (1980). 18. G. Bontempelli, F. Magno, G. A. Mazzochin and R. Seeber, Annali di Chimica 79, 103 (1989). 19. R. S. Nicholson and I. Shain, Analyt. Chem. 37, 190 (1965). 20. O. Hammerich and V. D. Parker, Electrochim. Acta lg, 537 (1973). 21. M. M. Baizer and H. Lurid, Organic Electrochemistry, 2nd Edition, p. 210. M. Dekker, New York (1983). 22. K.-D. Asmus, Acc. chem. res. 12, 436 (1979). 23. W. K. Musker, A. S. Hirschon and J. T. Doi, J. Am. chem. Soc. 100, 7754 (1978). 24. L. Michaelis, M. P. Schubert and S. Granick, J. Am. chem. Soc. 61, 1981 (1939). 25. L. Michaelis and S. Granick, J. Am. chem. Soc. 65, 1747 (1943).