October, 1989]
`
`© 1989 The Chemical Society of Japan
`
`Bull. Chem. Soc. Jpn., 62, 3187—3194 (1989)
`
`3187
`
`Perkin—Markovnikov Type Reaction Initiated with
`Electrogenerated Superoxide Ion
`
`Fumihiro OJIMA and Tetsuo OSA*
`Pharmaceutical Institute, Tohoku University, Aobayama, Aoba-ku, Sendai 980
`(Received May 12, 1989)
`
`The cyclic condensation of active methylene compounds such as diethyl malonate, dimethyl malonate,
`ethyl acetoacetate, or acetylacetone and dibromoalkanes such as 1,2-dibromoethane, 1,3-dibromopropane, 1,4.
`dibromobutane,
`l,5~dibromopentane, 1,6-dibromohexane, 1,3-dibromobutane, or 1,4-dibromopentane with
`electrogenerated superoxide ion was studied electrochemically in N,N-dimethylformamide (DMF) using cyclic
`voltammetry (CV) and controlled potential macro-electrolysis. The CV shows that electrogenerated superox-
`ide ion reacts with both active methylene compounds and dibromoalkanes in the dissolved oxygen medium.
`Controlled potential macro-electrolysis of the above components generally yielded cycloalkanes as the main
`In comparison, the chemical method using sodium ethoxide was also carried out. Two reaction
`products.
`mechanisms via the proton abstraction of active methylene compounds with electrogenerated superoxide ion
`and via the nucleophilic attack of the superoxide ion on dibromoalkanes are presented
`
`Superoxide ion (02') is not only one of the most
`important activated forms of molecular oxygen in
`biological systems but also novel activating reagent
`for organic synthesis,1'3l and has recently attracted a
`great deal of attention of biological and organic chem-
`ists. Superoxide ion is produced by the electrochemi-
`cal reduction of oxygen dissolved in non-aqueous
`solvent
`such as pyridine,
`acetonitrile, or N,N—
`dimethylformamide (DMF) at —0.85 V vs. silver—silver
`chloride (Ag/AgCl) as shown in Eq. 1.
`
`02+e‘—>O'2_
`
`(1)
`
`reactivities,
`various
`possesses
`ion
`Superoxide
`namely characteristics as a strong base, a nucleophile,
`an oxidant, a reductant, a free radical, and an electron-
`transfer agent.“
`In aprotic solvents, superoxide ion
`acts as an electrogenerated base (EGB),5) and abstracts
`a proton of an active methylene group to form a
`carbanion
`(Ewe—EH—EWG,
`EWG:
`electron-
`withdrawing group). Electrogenerated superoxide
`ion has another feature, as a nucleophile;5) it reacts
`with alkyl halides to form alkylperoxy radicals (R—
`OO-), which might be readily reduced further by
`electrogenerated superoxide ion or by an electron from
`the electrode to form alkylperoxide ions (R—OO‘).
`Alkali metal superoxides of potassium and sodium
`salts are well-known, but
`their insolubility in the
`usual organic solvents makes them of little prepara~
`tive use.“ Therefore,
`the reaction of alkali metal
`superoxides with organic compounds in benzene has
`been reported by the use of crown ethers in order to
`solve the problem of
`their insolubility." On the
`' other hand, the electrochemical method is experimen-
`
`tally more convenient because superoxide ion is con-
`tinuously generated on the electrode and its solubility
`is sufficiently high in the presence of the tetraethylam-
`monium cation of supporting electrolyte.“
`'
`We have investigated the characteristics of the reac-
`tivity of electrogenerated superoxide ion toward active
`methylene compounds and dibromoalkanes, and have
`reported preliminary results on the reaction of the
`active methylene compound, such as diethyl malo-
`nate, dimethyl malonate, or ethyl acetoacetate and
`dibromoalkane with
`electrogenerated superoxide
`ion.9l
`In this paper, we describe in detail the reaction
`of dibromoalkanes and active methylene compound of
`acetylacetone as well as malonates and ethyl acetoace-
`tate with electrogenerated superoxide ion in DMF
`solutions, using cyclic voltammetry and controlled
`potential macroelectrolysis.
`The reaction of this type is well known as Perkin—
`Markovnikov reaction (Scheme 1).““11 The reaction
`using sodium ethoxide as a base to form an aliphatic
`ring compound is widely used in organic syntheses.
`However, the reports on the reaction of this type using
`electrogenerated superoxide ion as a base, have not
`previously been published.
`’
`After detailed investigation using cyclic voltamme-
`try and macro-electrolysis in DMF, we have concluded
`that the present reaction system should be considered
`in terms of the nucleophilic reaction of electrogener-
`ated superoxide ion with dibromoalkanes as frequent
`as in terms of the proton abstraction of electrogener-
`ated superoxide ion from active methylene com-
`pounds. We report here the reaction of active methyl-
`ene compounds and dibromoalkanes using cyclic
`
`3;)
`_
`\C
`/CH2
`t1 OEt
`/ 82
`cr-12\CH CH(C02Et)2—+CH2\
`2
`8;
`8;;
`Scheme 1.
`
`C52
`
`f1
`/CH2\
`H OEt
`C(COZEt)2 ——> ca2\
`/C(C02Et)2
`CH2
`
`Page 1 of8
`
`Incyte Exhibit 1102
`
`Incyte V. Concert
`
`IPR2017-01256
`
`

`

`3188
`
`Fumihiro OJIMA and Tetsuo OSA
`
`[Vol. 62, No. 10
`
`voltammetry and macro-electrolysis in DMF, and dis
`cuss the results in terms of the basic properties and the
`nucleophilicity of electrogenerated superoxide ion.
`
`Experimental
`
`Materials. DMF (Guaranteed reagent, Tokyo Kasei) was
`stored over Molecular Sieves (4A 1/16) for 24 h and then
`distilled under reduced pressure. Further purification was
`achieved by passage through a column of activated alumina
`(ICN Alumina N-Akt. I). Tetraethylammonium perchlo-
`rate (TEAP) from Tokyo Kasei (Guaranteed reagent) was
`dried in vacuo and used as a supporting electrolyte in the
`electrochemical experiments. The substrates, diethyl mal-
`onate (Guaranteed reagent, Nakarai Chemicals), dimethyl
`malonate (Extra pure reagent, Tokyo Kasei), ethyl acetoace-
`tate (Extra pure reagent, Wako Pure Chemical Industries),
`and acetylacetone (Guaranteed reagent, Nakarai Chemicals)
`were purchased commercially and were redistilled.
`1,2-
`Dibromoethane (Extra pure reagent), 1,3-dibromopropane
`(Extra pure reagent), 1,4—dibromobutane (Guaranteed re~
`agent),
`1,5—dibromopentane
`(Guaranteed reagent),
`1,6-
`dibromohexane (Extra pure reagent), 1,3-dibromobutane
`(Extra pure reagent), and 1,4-dibromopentane (Extra pure
`reagent) were also obtained commercially from Tokyo Kasei
`and were distilled under reduced pressure before use.
`Cyclic Voltammety. Cyclic voltammetry was carried out
`in a DMF solution containing 0.2 M (1 M21 mol dmra)
`TEAP as a supporting electrolyte. A glassy carbon (GC)
`disk (area ca. 0.71 cm?) was employed as a working electrode,
`and a platinum wire was employed as a counter electrode for
`cyclic voltammetry. The working electrode was polished
`with a 0.05 pm alumina/water slurry on a felt surface,
`sonicated in distilled water, throughly washed with acetone,
`and dried before
`electrochemical measurements. The
`cathode potentials were referred to Ag/Ag+ (0.05 M AgNOa
`and 0.1 M TEAP in acetonitrile). Cyclic potential sweeps
`were generated by a self-made function generator in con-
`junction with a Hokuto Denko Model HA-305 potentiostat/
`galvanostat. Cyclic voltammograms were recorded on a
`Graphtec Model WX1200 X-Y recorder. All electrochemi-
`cal measurements were carried out at room temperature
`under a nitrogen atmosphere.
`Controlled Potential Macro-Electrolysis. For controlled
`potential macroeelectrolysis, a three compartment cell (1—1»
`type cell),
`in which cathodic and anodic chambers were
`separated by two fine-porosity sintered-glass
`frits, was
`employed with a magnetic stirrer and a reference electrode
`was put near the cathode. A glassy carbon plate with a
`surface area of ca. 12.5 cm2 was used as the cathode, and a
`platinum plate was used as the anode. An electrolyte solu-
`tion (02 M TEAP) of 60 cm3 containing two substrates (0.1
`M activated methylene compound and 0.] M dibromoal-
`kane) was placed in the cathodic chamber (25 cma),
`the
`anodic chamber (25 cm3), and the middle chamber (10 cm3).
`Dry oxygen was bubbled through the cathode. The
`cathode potential was controlled with a Hokuto Denko
`1 Model PIA-305 potentiostat/galvanostat. The quantity of
`electricity passed was measured with a Hokuto Denko
`Model HF-201 coulomb/amperehour meter. These electro-
`lyses were carried out at room temperature under a nitrogen
`atmosphere.
`Identification of Products and Determination of
`
`the
`
`Yields. The products of the electrolyses were separated
`from the catholyte and identified by the following proce-
`dure. The catholyte was evaporated in vacuo to remove
`DMF. The residue was dissolved in water and extracted
`with diethyl ether. The ether layer was dried over anhy-
`drous sodium sulfate and the ether was removed by distilla—
`tion under reduced pressure. The resulting liquid was
`tried to be separated into its components by silica-gel
`column chromatography using several organic mixed elu-
`ents such as hexane—diethyl ether (20:1), but could not be
`well separated.
`Authentic samples of the cycloalkanes were prepared by
`the ordinary chemical preparative method”) The proce-
`dure was as follows.
`In a lOO-ml three-necked round bot—
`tomed flask, equipped with a reflux condenser capped with
`a calcium chloride tube, a magnetic stirrer, and a 50—ml
`pressure—equalizing dropping funnel for addition of pre-
`pared 0.02 g-atom sodium ethoxide-ethanol solution, were
`mixed 0.01 mol each of active methylene compound and
`dibromoalkane. The mixture was heated to 80°C and
`vigorously stirred while the sodium ethoxide—ethanol solu-
`tion was slowly added into the flask over a 1.5 h period.
`After the addition was finished, the mixture was refluxed,
`with continued stirring, for an additional 45 min, and then
`the ethanol was removed by distillation. The reaction mix-
`ture was cooled, and 5 ml of cold water was added; After
`the sodium halides were completely dissolved, the organic
`layer was separated and the aqueous layer was extracted with
`three 10 ml portions of diethyl ether. The combined ether
`extracts were shaken with 5 ml of saturated NaCl solution,
`dried over anhydrous sodium sulfate, filtered, and concen-
`trated on a rotary evaporator. The identity of the electro—
`chemically prepared cycloalkanes with the
`respective
`authentic samples was established by gas chromatography
`and GC-MS spectroscopy. The main products of electro-
`lyses were detected by using a Shimadzu Model GC-4CM gas
`chromatograph equipped with 2 m><3 mmq‘) column packed
`with Carbowax 20M, and identified by comparing their
`retention times with those of authentic samples. Yields of
`the products were also determined by gas chromatographic
`analysis. The by-products of electrolyses were identified by
`using a JEOL Model JMS-DX300 GC-MS spectrograph
`equipped with 2 m X 3 mmd) column packed with Silicone
`OV—lOl.
`
`Results and Discussion
`
`Cyclic Voltammetry of Active Methylene Com-
`pounds and Dibromoalkanes. The cyclic voltamme-
`try of diethyl malonate (l), dimethyl malonate (2),
`ethyl acetoacetate (3), acetylacetone (4), 1,2-dibro-
`moethane (a), 1,3-dibromopropane (b),
`1,4-dibro—
`mobutane (c), 1,5-dibromopentane (d), 1,6-dibro-
`mohexane
`(e),
`1,3—dibromobutane
`(f),
`and
`1,4-
`dibromopentane (g) was carried out in a 0.2 M TEAP/
`DMF solution under a nitrogen atmosphere. All the
`substrates were reduced at more negative potentials
`than *25 V vs. Ag/Ag+ as shown in Table l. The
`compounds 4, c, d, e, and g exhibited prepeaks (Eé) at
`more positive potentials than the main peaks (E3).
`The cyclic voltammograms of 1, a, and c are demon-
`strated in Fig. l. The reduction peak of 1 could not
`
`Page 2 of 8
`
`

`

`October, 1989]
`
`Perkin—Markovnikov Type Reaction Using Superoxide Ion
`
`3189
`
`Table 1. Reduction Peak Potentials of Substrates
`——————._.__.______________________
`
`Compound
`E;
`Ef,
`Diethyl malonate (1)
`—3.36
`Dimethyl malonate (2)
`“3.33
`Ethyl acetoacetate (3)
`—2.96
`Acetylacetone (4)
`—2.97
`1,2-Dibromoethane (a)
`—2.58
`1,3~Dibromopropane (b)
`—2.87
`1,4-Dibromobutane (c)
`—3.29
`1,5-Dibromopentane (d)
`—3.31
`1,6-Dibromohexane (e)
`~3.39
`1,3»Dibromobutane (f)
`“-2.97
`
`l/l-Dibromopentane (g) —3.24 (~2.80)
`E/V vs. Ag/Ag+, solvent: DMF, supporting electro-
`lyte: TEAP, working electrode: GC, reference elec-
`trode: Ag/Ag+, counter electrode: Pt.
`
`(~2.80)
`(~2.80)
`(“2.80)
`
`
`
`“2.54
`
`ISODA
`
`:9 ;
`
`;
`
`;
`
`a.
`
`C
`
`
`
`-l.0
`
`0
`
`+1.0
`
`E / v
`
`vs.
`
`Ag / Ag+
`
`Fig. 2. Cyclic voltammograms of Oz in the absence
`(—-) and presence ( ----- ) of 2 mM 1.
`Sweep rate: 0.1 Vs‘l.
`
` 0
`
`0.2
`
`OJ»
`
`(Vs)%
`v /
`1
`~———-——o—-——-—+——-—-———o——-
`—l.0
`0
`+1.0
`
`E / V
`
`vs.
`
`Ag / Ag
`
`+
`
`-—-+——————+———¢-—-——-—~c——————¢_
`~3.0
`-2.0
`-1.0
`0
`+1.0
`
`E/V vs. Ag/Ag+
`
`Fig. 1. Cyclic voltammograms of 16 mM 1, a, and c
`in 0.2 M TEAP/DMF at a GC disk electrode.
`Sweep rate: 0.2 V s“.
`
`be measured directly because of the deformed CV
`curve, but
`the peak potential
`is estimated to be
`-3.36 V.
`
`Cyclic Voltammetry of Oxygen in the Presence of
`Active Methylene Compounds (1—4). As the first
`reduction potential of oxygen in a 0.2 M TEAP/DMF
`solution on glassy carbon electrode is —1.28 V,
`the
`dissolved oxygen is generally reduced more easily than
`active methylene compounds. Figure 2 indicates the
`cyclic voltammograms of saturated oxygen in the
`absence and presence of 1.
`In the absence of l, the
`ratio of the anodic peak current (if) to the cathodic
`peak current (i?) is almost 1.0; that is, the superoxide
`ion formed in this system is stable.
`In the presence of
`2 mM 1, the reductive peak current of oxygen at —1.28
`V increased and the reoxidation peak current of the
`
`Fig. 3. Cyclic voltammograms of 02 in the presence
`of 3 mM 1 at different sweep rates.
`—~: 0.02 V 3*, -———: 0.05 Vs‘l, — ‘ —-—: 0.1 Vs—l,
`----- -: 0.2 V s“.
`Inset. Sweep rate dependence of the cathodic peak
`current for the reduction of 02 under 3 mM 1.
`
`superoxide ion at —1.16 V decreased, in comparison
`with those in the absence of 1. These results indicate
`
`that some of the superoxide ion are consumed by the
`reaction of 1. The cathodic peak current of oxygen is
`linear to the square root of scan speed in the presence
`of 1 (Inset in Fig. 3). This means that the electrode
`reaction of oxygen is apparently diffusion-controlled,
`even if 1 is present in the system. The oxygen reduc-
`tion behavior in the presence of other active methylene
`compounds 2, 3, and 4 gave similar to those of 1. But
`the cyclic voltammograms of oxygen reduction in the
`case of 3 and 4 gave prepeaks. The prepeaks in the
`case of 3 is shown in Fig. 4.
`The cyclic voltammetry of oxygen in the differrent
`concentrations of 1 is demonstrated in Fig. 5. Increas-
`ing the concentration of 1 resulted in a corresponding
`increase in the reduction peak current of oxygen.
`This increase of catalytic current did not continue
`
`Page 3 of 8
`
`.
`
`_
`
`'
`
`'
`
`

`

`3190
`
`Fumihiro OJIMA and Tetsuo OSA
`
`[Vol. 62, No. 10
`
`
`
`-l.0
`
`O
`
`+1.0
`
`
`
`0
`
`2
`
`A
`
`1.6
`1.4
`12
`10
`8
`6
`Concentration of 1 / mM
`
`18
`
`20
`
`30
`
`E/V vs. Ag/Ag+
`
`Fig. 4. Cyclic voltammograms of 02 in the absence
`(—) and presence ( ----- ) of 2 mM 3.
`Sweep rate: 0.1 V 5‘1.
`
`Fig. 6. Dependences of the peak current ratio for
`the concentration of 1.
`Sweep rate: 0.1 Vs'l, O: ic/if, O: z'a/z'f.
`
`02 + e‘ -> O'z‘
`R1
`>an + or _>
`R2
`
`R1
`
`R2
`
`>EH + no.3
`
`(1)
`
`(2)
`
`HOé + e‘
`R1, R2:
`
`(3)
`——+ H02"
`electron-withdrawing group (EWG)
`
`The new anodic waves agreed with the oxidation
`peaks observed in the cyclic voltammogram of l by the
`addition of sodium ethoxide and in the cyclic voltam—
`mogram of H202 by the addition of tetraethylammo-
`nium hydroxide, respectively.
`increased with the
`This oxidation peak current
`increasing concentrations of l. The increase did not
`continue infinitely, but reached a saturation in the
`presence of l of more than 18 mM,
`In the acetylace—
`tone system, three new anodic peaks beside the oxida-
`tion peak of the superoxide ion formed appeared as
`shown in Fig. 7. The first anodic peak at —0.05 V
`and the second anodic peak at +0.10 V were assigned
`
`I
`
`,~-...
`.= «x -
`s/‘°;$Eb'i’:m--
`
`‘4
`
`—1.0
`
`o
`
`+1.0
`
`E/V vs. Ag/Ag+
`
`Fig. 7. Cyclic voltammograms of 02 in the presence
`of different concentrations of 4.
`
`Sweep rate: 0.1 V5”,
`:0 mM, —-——: 4 mM,
`—-——:8rnM, ------ :16me
`
`1....
`
`l
`
`—-—-—-—o—-—-————o—-—-—-—-——o—
`—1.0
`0
`+1.0
`
`E/V vs. Ag/Ag+
`
`Fig. 5. Cyclic voltammograms of Oz in the presence
`of different concentrations (0, 2, 4, 8, and 16 mM)
`of 1.
`Sweep rate: 0.1 V 5'1.
`
`infinitely but reached a saturation at 18 mM 1. The
`maximum value of the ratio, ic/z'é’, is ca. 1.65, where to
`and if are the reduction peak currents of oxygen in the
`presence and absence of 1, respectively. These behav-
`ior is shown in Fig. 6.
`Increasing the concentration
`of 1 also resulted in a corresponding decrease in the
`reoxidation peak current of the superoxide ion. The
`peak current ratios,
`23/2}? are getting smaller with
`increasing concentrations of l, where I}, and is? are the
`reoxidation peak currents of the superoxide ion in the
`presence and absence of 1, respectively. These results
`also support that the electrogenerated superoxide ion
`is consumed by subsequent chemical reactions during
`cyclic voltammetry. The new anodic wave at —0.08 V
`(Fig. 5) is considered to be the oxidation peak charac-
`teristics of carbanion (EH(C02Et)2) at glassy carbon
`electrode, overlapped with the reoxidation wave of
`hydroperoxide ion (HOE) formed by the following
`_ elementary reactions.
`
`Page 4 of 8
`
`

`

`October, 1989]
`
`Perkin»Markovnikov Type Reaction Using Superoxide Ion
`
`3191
`
`Table 2.
`Observed Peak Potentials in the CV of 02 under 1—4W
`Reduction peak potential/V
`Reoxidation peak potential/V
`
` Prepeak Oxygen Superoxide ion HOE Carbanion Unknown
`
`
`
`
`1
`—
`—1.28
`—l.l6
`—0.08
`-0.08
`—
`2
`——
`—1.28
`—1.16
`—0.06
`-0.06
`——
`3
`“1.20
`—1.28
`—-1.17
`-0.07
`+0.00
`+0.86
`
`—l.18 —l.28 -1.16 -0.05 +0.104 +0.57
`
`
`
`
`
`
`I SOuA
`
`———-——-t—————-o—————————+———
`-l.0
`0
`+1.0
`
`E/V vs. Ag/Ag+
`
`Fig. 9. Cyclic voltammograms of 02 in the presence
`of different concentrations (0, 1, 2, 4, 8, and 16 mM)
`of a.
`Sweep rate: 0.1 Vs’l.
`
`‘
`
`
`
`16
`14
`12
`10
`'ozasz
`Concentration of a / mM
`
`18
`
`20
`
`30
`
`Fig. 10. Dependences of the peak current ratio for
`the concentration of a.
`Sweep rate: 0.1 Vs‘l, .1 ic/ié’, O: ia/if.
`
`When dichloroalkanes were used in place of dibro—
`moalkanes, no change of cyclic voltammogram in the
`absence and presence of dichloroalkanes was observed.
`This means that the superoxide ion is hard to react
`with dichloroalkanes. On the other hand, when di-
`iodoalkanes were used in place of dibromoalkanes, the
`superoxide ion reacted with diiodoalkanes to reach a
`saturation at
`lower concentrations than dibromo-
`alkanes.
`
`The cyclic voltammetry of oxygen was also carried
`out in the different concentrations of dibromoalkanes.
`
`A typical result containing a is shown in Fig. 9.
`
`
`
`-l.0
`
`0
`
`+1.0
`
`E / V vs.
`
`Ag / Ag+
`
`Fig. 8. Cyclic voltammograms of 02 in the absence
`(—) and presence ( ------ ) of 2 mM at.
`Sweep rate: 0.1 V5“.
`
`to the oxidation peaks of H03 and Emcocnm,
`respectively. The second peak was confirmed to
`appear by the addition of sodium ethoxide. The
`third anodic peak was not reasonably assigned. The
`cyclic voltammograms of oxygen in the presence of 2
`were very similar to those of 1 and those of 3 were
`similar to those of 4. These results are summarized
`in Table 2.
`
`By comparing the ratio of anodic peak current to
`cathodic peak current for the oxygen redox system in
`the presence of 8 mM active methylene compounds,
`the superoxide ion is found to react with active meth-
`ylene compounds
`in the order of 4(pKa=9.0)>3
`(10.2)>2=1(l3.5). This order is the same as that of
`acidity (pKt) of active methylene compounds as car-
`bon acids.
`
`Cyclic Voltammetry of Oxygen in the Presence of
`Dibromoalkanes (a—g). As a typical result, the cy-
`clic voltammetry of oygen in the presence of a is shown
`in Fig. 8. This cyclic voltammogram is very similar
`to that in the presence of 1.
`In the presence of 2 mM
`2, the first reduction peak current of oxygen at —l .30 V
`increased, and the reoxidation peak current of the
`_ superoxide ion at —l.18 V decreased, in comparison
`with those of
`the saturated oxygen alone. These
`results indicate a progress of the reaction of superox-
`ide ion with a. The cyclic voltammetry of oxygen in
`the presence of other dibromoalkanes (b—g) gave sim-
`ilar results to that of afl‘
`
`'1 The electrode reaction of oxygen is also apparently
`diffusion-controlled in the presence of a—g.
`
`Page 5 of 8
`
`

`

`3192
`
`Fumihiro OJIMA and Tetsuo OSA
`
`[Vol. 62, No. 10
`
`Table 3.
`-——-————-—-—-——-——————_________—__—______
`Observed Peak Potentials in the CV of 02 under a—g
`Reduction peak potential/V
`Reoxidation peak potential/V
`
`
` Prepeak Oxygen Superoxide ion Bromide ion Unknown
`
`
`a
`—
`—l.30
`—-1.18
`+0.38
`+0.86
`b
`—
`+1.30
`—l.18
`+0.33
`+0.86
`c
`——
`—l.29
`+1.18
`+0.33
`+0.83
`d
`-—
`—l.29
`—1.18
`+0.35
`+0.85
`e
`———
`—1.29
`-l.18
`+0.40
`+0.90
`f
`—
`+1.30
`—1.19
`+0.32
`+0.82
`
`——g +0.82 -1.30 —1.19 +0.32
`
`
`
`
`
`‘
`
`Increasing the concentration of a resulted in a corre—
`sponding increase in the reduction peak current and
`reached a saturation in the presence of 11 mM a. The
`maximum value of the ratio, ic/ié’, is ca. 1.65 and the
`same as in the presence of active methylene com-
`pounds. This behavior is demonstrated in Fig. 10.
`Two new anodic peaks at +0.38 V and +0.86 V
`observed in the presence of a can be ascribed to the
`oxidation of bromide ion which was produced by
`the nucleophilic substitution of the superoxide ion
`
`Br(CH2)nBr + 02; —» Br(CH2),,OO- + Br-
`
`(4)
`
`- with the dibromoalkane (Fig. 8). These oxidation
`peak currents also increased with the increasing con-
`centrations of a and reached a saturation at the con-
`
`centration of ll mM. The cyclic voltammograms of
`oxygen in the presence of other dibromoalkanes (b—
`g) gave similar behavior to that of a and are summar-
`ized in terms of the potentials in Table 3.
`In order
`Determination of Simple Rate Constants.
`to clarify each reaction rate of the electrogenerated
`superoxide ion with a variety of substrates, rate con-
`- stants have been estimated based on a simple ECE
`reaction mechanism and the values determined by use
`of a digital simulation method are listed in Table 4.
`This is the method which simulates the ratio of
`
`enhancement of the cathodic peak current of oxygen
`reduction in the presence of substrate to that in the
`absence of substrate. The results indicate that the
`
`superoxide ion is more reactive with dibromoalkanes
`than with active methylene compounds. The rate
`constants for 3 and 4 could not be obtained accurately
`in the present simulation method, since the cyclic
`voltammograms of oxygen showed the prepeaks
`
`‘
`
`Table 4. Reaction of Electrogenerated 0‘2‘
`with Substrates
`
`
`
` Substrate Rate constant lam/M45"1
`
`l><102
`Diethyl malonate (l)
`l><102
`Dimethyl malonate (2)
`l><102
`Ethyl acetoacetate (3)
`1X102
`Acetylacetone (4)
`6X102
`1,2-Dibromoethane (a)
`6X102
`1,3—Dibromopropane (b)
`5X102
`1,4‘Dibromobutane (c)
`6X102
`1,5-Dibromopentane ((1)
`6X102
`1,6—Dibromohexane (e)
`6X102
`1,3—Dibromobutane (f)
`
`1,4-Dibromopentane (g) 5X102
`a) Determined by digital simulation based on' the
`following mechanisms.
`(1) 02+e-—->O'2'
`0'2-+ RHz —> RH- + HOé
`o‘;+ e- —» HO;
`R = (EWG)2C
`(2) 02 + e“ —» 0';
`03+ RBr ——+ ROO’ + Br—
`ROO‘ + e- -+ R00—
`R : imam).
`
`which might be a very rapid reaction between 02' and
`3 or 4.
`
`Controlled Potential Macro-Electrolysis. The con-
`trolled potential'macro-electrolyses of oxygen in the
`presence of two kinds of substrates, active methylene
`compounds (1—4) and dibromoalkanes (a—d, f, and
`g) were carried out at —1.5 V in a 0.2 M TEAP/DMF
`solution. At +1.5 V, only oxygen was reduced to
`superoxide ion by one-electron transfer, while both
`
`R}
`
`\
`/CH2
`R2
`
`»
`+ BarBr(or BartT‘HBr)
`CH3
`
`—-——->
`
`R1
`
`\
`/C
`R2
`
`Rn(or
`
`R1
`
`\
`/Rn
`/C\
`R2
`
`) )
`(IZH
`CH3
`
`1: R1=R2=C02Et
`2: R1=R2=C02Mt
`3: R1=COCH3, R2=C02Et
`4: R1=R2=COCH3
`
`a: Rn=(CH2)2
`b2 Rn=(CH2)3
`c: Rn=(CH2)4
`d: Rn=(CH2)5
`e: Rn=(CH2)s
`f: R’n=(CH2)2
`g: R’n=(CHa)3
`Scheme 2.
`
`5a—d
`Ga—d
`7a-—d
`Ba—d
`
`5f, g
`6f, g
`7f, g
`8f, g
`
`Page 6 of 8
`
`

`

`October, 1989]
`
`Perkin—Markovnikov Type Reaction Using Superoxide Ion
`
`3193
`
`chemical method was in the order of C5>C4QC6>C3.
`The use of branched dibromoalkanes such as 1,3-
`dibromobutane
`(f)
`and
`1,4-dibromopentane
`(g)
`decreased the yields of the annulated products in com—
`parison to those of the same membered ring products
`formed from a,w-dibromoalkanes; the yields of 5—8f,
`and 7g, 8g are lower than those of 5—8c and 7d, 8d,
`respectively (excluding the cases of 5f=5d and 6f>6d).
`This trend may be caused by steric hindrance of
`methyl branch at the carbon substituted.
`From the above mentioned results,
`the following
`electrochemical reaction mechanism for the system of
`malonic acid ester and a,w-dibromoalkane can be
`presented:
`Mechanism 1: Superoxide ion as an EGB
`
`02 + e- —» 0'2-
`
`R1
`\CH2 + 0'2-
`/
`R2
`
`—»
`
`R1 _
`\c H + HOé
`/
`R2
`
`Hot + e‘(O'2‘) —» H05
`
`1
`
`Mechanism 2: Superoxide ion as a nucleophile
`
`02 + e' _. 0‘2-
`Br(CH2)nBr + 02‘ —> Br(CH2),,OO‘ + Br“
`Br(CH2),,OO' + e’(O'2“) —+ Br(CH2),.OO‘
`Br(CH2),,0-
`
`R1
`
`>CH2 + Br(CH2),oo— ~—>
`
`R2
`
`R1
`
`\EH+Br(CH2)nOOH
`/
`R2
`
`R1
`
`R2
`
`(
`
`>CH2+ Br(CH2),,O- —+
`R1
`\6H+Br(CH2).0H)
`R/2
`
`(1)
`
`(2)
`
`(3)
`
`(1)
`(4)
`(5)
`(6)
`
`(7)
`
`(8)
`
`(9)
`
`the substrates, active methylene compounds and
`dibromoalkanes, were not reduced.
`The main products were aliphatic cyclic or annu-
`lated compounds of one carbon extension comparing
`to the starting dibromoalkanes (Scheme 2).
`Diethyl cycloalkanedicarboxylates (5a—d, f, and g)
`were produced from diethyl malonate and dibromoal-
`kanes in the presence of electrogenerated superoxide
`ion, dimethyl cycloalkanedicarboxylates (Ga—d, f, and
`g) from dimethyl malonate and dibromoalkanes, ethyl
`l-acetylcycloalkanecarboxylates (7a—d, f, and g) from
`ethyl acetoacetate and dibromoalkanes, and 1,1-
`diacetylcycloalkanes (8a—d, f, and g) from acetylace-
`tone and dibromoalkanes, respectively.
`The electrolysis results are summarized in Table 5.
`Numbers in parentheses show the yields of the corre-
`sponding products by the ordinary chemical prepara-
`tive method using sodium ethoxide as a base. The
`total amount of charge passed was ca. 482.5 C
`(2 Fmol‘l). The yields by the electrochemical
`method and the ordinary chemical method are based
`on the consumed amount of active methylene com—
`pounds. The highest yield of the annulated com—
`pounds by the electrochemical method was 90% of 6c.
`The yields of 5, 6, and 8 by the electrochemical method
`were higher than those by the chemical method except
`the case of St. However, the yields of 7 were of the
`inverse tendency. The yields of the annulated prod
`ucts depended markedly on the methylene length of
`a,w-dibromoalkanes, and the yields of stable ring size
`such as five or six membered ring were generally high.
`The reason can be explained by the degree of strain,
`when the ring closure occurs.
`In the case of 1 or 2 as active methylene compound,
`the yields of the annulated products by the electro-
`chemical method decreased in the order of C5>C5>
`C4>C3, similarly to those observed in the ordinary
`chemical reaction.
`In the case of 3, a somewhat dif-
`ferent
`tendency was observed. For
`instance,
`the
`yields by the electrochemical method decreased in the
`order of C3>C5>C4~C3, whereas the yields by the
`
`Table 5. Yields of Annulated Products by
`Preparative Electrochemical Method
`Substrate
`Product
`Dibromoalkane
`
`No.
`No.
`a
`b
`c
`d
`f
`g
`l
`5
`49% 51% 87% 82% 20% 83%
`(28%) (43%) (66%) (50%) (48%)(58%)"
`52% 55% 90% 81% 17% 89%
`( 4%) (19%) (20%) (21%) (14%)(36%)
`28% 30% 35% 56% 27% 29%
`(20%) (58%) (58%) (67%) (45%) (61%)
`19% 68% 72% 58% 36% 37%
`8
`4
`
`( 5%) (16%) (23%) (25%) (21%) (22%)
`
`2
`
`3
`
`6
`
`7
`
`3) Numbers in parentheses show the yields of the
`corresponding products by the chemical preparative
`method in which 0.01 mol each of active methylene
`compound and dibromoalkane was reacted with 0.02
`g atom sodium ethoxide in ethanol at 80 °C.
`
`Page 7 of 8
`
`R1\_
`/CH + Br(CI—I2)nBr _+
`R2
`
`/H
`R1\
`/c \
`
`(CH2)nBr
`
`_
`
`+Br
`
`R2
`R
`
`4—.
`
`R1
`
`>c—\
`
`—+—>
`
`1\C/\(CH2 )n + Br"
`
`R2
`(CH2)nBr
`R2
`R1, R2: electron-withdrawing group (EWG)
`
`In mechanism 1, superoxide ion, as an electrogener-
`ated base (EGB), deprotonates an active methylene
`compound (Eq. 2), and the resulting carbanion reacts
`with a,w-dibromoalkane to yield the aliphatic annu-
`lated compound (Eq. 9). In mechanism 2, superoxide
`ion, as a nucleophile, reacts with a,w-dibromoalkane,
`and the produced bromoalkylperoxy radical (Eq. 4) is
`further reduced by superoxide ion or by an electron
`from the electrode to form bromoalkylperoxide ion
`
`

`

`Furnihiro OJIMA and Tetsuo OSA
`
`[V0]. 62, No. 10
`
`
`
`50
`
`100
`
`150
`
`SCAN
`
`Fig. 11. Typical gas chromatogram of the reaction
`of 02‘ with 1 and c.
`A: acetone, B: DMF, C: CH2(COzEt)2,
`D: Br(CH2)4Br,
`CHz—CHz
`\ /
`
`COzEt
`
`E:
`
`/C\
`I
`CHz-CHZ
`
`COzEt
`
`(M+=215),
`
`.
`/CH2-Cl12\ /C02E[
`F. CH2\
`/C\
`CHz—O
`COzEt
`
`+2
`(M 231),
`
`/C02Et
`Et02C\
`/CH—CH\
`EtOzC
`COzEt
`EtO2C
`O COzEt
`>c5—‘c<
`EtOzC
`
`COzEt
`
`G:
`
`H:
`
`(M+=333),
`
`(M+:319),
`
`/C02Et
`EtOzC\
`/C(CH2)4C\
`EtOzC
`CO2Et
`
`1:
`
`(M+=375).
`
`'
`
`‘
`
`(Eq. 5) or bromoalkyloxide ion (Eq. 6). The result-
`ing anions deprotonate, as a base, the active methylene
`compound (Eqs. 7 and 8), and the produced carbanion
`reacts with a,w-dibromoalkane to yield the aliphatic
`annulated compound as mechanism 1 (Eq. 9). The
`similar reaction mechanism can be presented in the
`other cases of branched dibromoalkanes.
`
`In addition to the annulated compounds, cyclic
`ethers such as tetrahydrofuran and diols such as 1,4-
`butanediol from c were also detected as minor prod-
`ucts by the careful gas chromatographic analysis at
`room temperature, and identified by comparing their
`retention times with those of authentic samples.
`Other by-products, mainly some dimers, were detected
`and assigned using GC-MS spectroscopy. The gas
`chromatogram in the case of the reaction of 02‘ with l
`and c is demonstrated in Fig. 11, and each peak was
`assigned to the compound shown in the figure, respec-
`tively. Most of these byproducts were not detected in
`the products by the ordinary chemical method.
`These results indicate that the reaction mechanism
`
`the present electrochemical reaction might be
`’ for
`partly different from that for the ordinary chemical
`reaction. The main difference is that electrogener-
`
`ated superoxide ion forms intermediates shown in
`Eqs. 4—6, whereas sodium ethoxide does not form
`such intermediates. Therefore, the carbanion forma-
`tion via Eqs. 7 and 8 cannot proceed by the ordinary
`chemical method.
`
`Conclusion
`
`Perkin-Markovnikov type reaction initiated with
`electrogenerated superoxide ion was investigated elec-
`trochemically using cyclic voltammetry and con-
`trolled potential macro-electrolysis. By electrolyses,
`superoxide ion acted as a base and a nucleophile, and
`reacted with both the substrates, active methylene
`compounds and dibromoalkanes,
`to yield the end
`products, aliphatic cyclic compounds. This electro-
`chemical method can be applied to synthesize one
`carbon-extended cyclic compounds from dibromo-
`alkanes.
`
`Besides these compounds as the main products,
`many by-products, such as dimers and cyclic ethers
`were detected using GC-MS spectroscopy. The for-
`mation reaction of the cyclic ethers is a novel reaction
`which has never been seen in the usual base-catalyzed
`reactions. All of these results are based on the dual
`
`nature of electrogenerated superoxide ion; i.e., basicity
`and nucleophilicity, which vary with substrates.
`
`References
`
`1) G. A. Hamilton, “Chemical Models and Mechanisms
`for Oxygenases in Molecular Mechanisms of Oxygen Activa-
`tion,” ed by O. Hayaishi, Academic Press, New York (1975),
`p. 405.
`2) R. Dietz, M. E. Peover, and P. Rothbaum, Chem. Ing.
`Techn, 42, 185 (1975).
`3) T. Osa and M. Tezuka, Denki Kagaku, 44, 2 (1976).
`4)
`E. Lee-Ruff, Chem. Soc. Rev., 1977, 195.
`5) M. Sugawara, M. M. Baizer, W. T. Monte, R. D.
`Little, and U. Hess, Acta Chem. Scand., Ser. B, 37, 509
`(1983).
`6) R. Dietz, A. E. J. Forno, B. E. Larcombe, and M. E.
`Peover, ]. Chem. Soc. B, 1970, 816.
`7)
`J. S. Valentine and A. B. Curtis, ]. Am. Chem. 800.,
`97, 224 (1975).
`8)
`J. W. Peters and C. S. Foote, ]. Am. Chem. Soc, 98,
`873 (1976).
`9)
`F. Ojima, T. Matsue, and T. Osa, Chem. Lett., 1987,
`2235.
`
`10) W. H. Perkin, Ben, 16, 208 (1883).
`11) W. Markovnikov and Krestovnikov, justus Liebigs
`Ann. Chem., 208, 333 (1881).
`12) R. P. Mariella and R. Raube, Org. Synth., IV, 288
`(1963).
`
`
`
`Page 8 of 8
`
`