224
by a pair of BHB bridge bonds6 The B21H233-ion is analogous to B20H1g3-, and the new B2rH20122should be analogous to B20H1B2-. The transfer of two electrons and liberation of one H+ per B24H21IZ3-support this analogy (eq 1). The equiva-
+
BZaH1,I23+BZ4H~oI22- H +
+ 2e-
(1)
lent weight and analyses for boron and iodine lead to BioHto ION (PHOTOISOMER) the formulation B24H20122-.The lack of a uv absorption maximum (observed for the stable isomer of B20HlB2- and not for the photoisomer) and the presence of an ir band at 2200 cm-' (BHB bridge, absent in the stable isomer of B20Hls2-and present in the photoisoI mer) suggest a structure for B24H20122analogous to that of the photoisomer of BzoHle2(see Figure 1). a; Unfortunately, the nmr spectra do not provide an h q H a o I ~ ION unambiguous indication of the structure of the Figure 1. Circles represent boron atoms. Each boron of B2oH12B24H2012zion. The presence of at least three environexcept 2,6,2',9' is bonded to a terminal H. Each boron in the proments in the 'H nmr spectrum and a poorly resolved posed structure of B24H~~12Zexcept 2,11,2',7' is bonded to a termdoublet in the 'B nmr spectrum indicate either that the inal H or I. oxidized ion is a mixture of isomers with respect to the iodine substituents due to a lack of stereospecificity in A sample of [(CzH6)4N]3Bz4H2112 (2.44 g, 2.64 mmol) the reactions, or that the positions of the iodine atoms was dissolved in 50 ml of acetonitrile (0.1 A4 tetraethyllower the symmetry of the ion to the point that no ammonium perchlorate as supporting electrolyte). unique structure can be assigned on the basis of the The solution was exhaustively electrolyzed under nitrospectral data. Models based on the stable isomer of 1.3 V (sce), using a graphite cloth electrode. gen at BzoHle2-suggest that H-H repulsions between cages The current was monitored and corresponded to the will be more severe for coupled Blz cages than for value n = 2.04 equiv per mol of [(CzHj)4N]3Bz4Hzl12. coupled Blo cages. These steric repulsions can be sigThe electrolyzed solution was added to 400 ml of HzO nificantly reduced by the formation of BHB bridge and titrated with standardized NaOH solution. The tibonds between cages instead of direct BB or BBB bonds. tration indicated that 0.92 mol of Hf per mol of (6) (a) M. F. Hawthorne and R. L. Pilling, ibid,, 88, 3873 (1966); B24H2Jz3-was formed in the oxidation process. The (b) B. G. Deboer, Z. Zalkin, and D. H. Templeton, Inorg. Chem., 7. aqueous solution was heated to boiling to remove the 1085 (1968). acetonitrile. The precipitate which formed as the aceR. J. Wiersema, R. L. Middaugh tonitrile evaporated was collected by filtration of the hot Department of Chemistry, University of Kansas solution and recrystallized from an acetonitrile-water Lawrence, Kansas 66044 mixture. Almost quantitative yields of the product Received Nooember 5. 1969 were obtained by this procedure. Anal. Calcd for [(C2H5)4N]zBz4H2012: B, 32.68; I, 31.97; equiv wt, 398.0. Found: B, 31.0; I, 31.6; equiv wt, 400. Evaluation of Fast Homogeneous Electron-Exchange The ir spectrum included absorptions at 2510 and 2200 Reaction Rates Using Electrochemistry and cm-'. The 32-MHz IIB nmr spectrum of [(CZH&Nl2- Reflection Spectroscopy BZ4H20Iain acetonitrile consisted of a single poorly Sir: resolved doublet 34.1 ppm upfield from external methyl borate. The 'H nmr spectrum of aqueous Na2BZ4We wish to report the evaluation of the kinetic rates HzoIzwith I'B decoupled consisted of at least three of fast homogeneous reactions involving electrogensignals at 1.17, 1.30, and 1.55 ppm downfield from inerated species using the technique of internal reflection ternal sodium 3-(trimethylsilyl)-l-propanesulfonate. spectroscopy (irs) at optically transparent electrodes Only end absorption was observed near 200 nm in the (ote). This technique has not been previously applied uv spectrum. successfully to kinetic measurements. Moreover, we A one-electron oxidation of BloHloZ- yields either wish to demonstrate that fast homogeneous rates apBzoHI93-or BZ0Hls4-,depending on the pH of the meproaching diffusional control are determinable, taking dium. Under slightly more vigorous conditions, a advantage of the repetitiveness of the process for imhigher oxidation state results, B z o H I B ~ - In . ~ B20H~~4-, provement of the signal to noise ratio. there is a direct B-B bond between Blo cages, and that Although this technique can be applied in general to bond is apparently protonated to form a BHB bridge follow any species whose concentration is perturbed bond in B20H193-.4b,cIn the higher oxidation state, through a heterogeneous electrochemical reaction, the BZ0H1s2-,two Blo cages are linked by a pair of BBB reaction scheme's2 three-center Uv irradiation of B20H~s2A +B =LC eEla (1) gives a photoisomer in which two Blo cages are linked
+
(4) (a) A. Kaczmarczyk, R. D. Dobrott, and W. N. Lipscomb, Proc. Nut. Acad. Sci. U.S., 48, 729 (1962); (b) B. L. Chamberland and E. L. Muetterties, Inorg. Chem., 3, 1450 (1964); (c) M. F. Hawthorne, R. L. Pilling, and P. F. Stokely, J . Amer. Chem., SOC.,87, 1893 (1965); (d) R.L. Middaugh and F. Farha, Jr., ibid., 88,4147 (1966). (5) C. H. Schwalbe and W. N. Lipscomb, ibid., 91, 194 (1969).
Journal of the American Chemical Society
B --f C
=LC
e-
Ezo
(2)
(1) J. W. Strojek, T. Kuwana, and S. W. Feldberg, J . Amer. Chem. SOC.,90, 1353 (1968). (2) T. Kuwana and J. W. Strojek, Discuss. Faraday Soc., 45, 134 (1968).
/ 92:I / January 14, I970
225 490
with the equilibrium reaction
800
kl
AfC-2B
(3)
kb
where kf and kb are the rate constants of the forward and reverse reactions in the equilibrium, will be used as an example. The redox potentials, Elo and Ez”, and the equilibrium constant for reaction 3, Keq = kf/kb, are thermodynamically related 2 as previously discussed for the normal transmission spectral studies. Experimentally, the approach is to follow the irs absorbance of either reactant or products during the usual potential step (chronoamperometric) experiment where the concentration at the electrode surface is zero. Thus, mechanistically, species A is transformed at the electrode at a diffusion-controlled rate; as C moves to the bulk solution by diffusion, it encounters A diffusing toward the electrode and reaction occurs to attain equilibrium. Theoretical analysis of this problem is similar t o previously discussed work3 with the exception that the “optical cell” is now restricted and defined by the penetration depth 6 of the light beam into the solution phase. The initial requirement is to mathematically analyze and experimentally verify the light absorbance A as a function of time for the simplest case, i.e., reaction 1, for a chronoamperometric experiment where the rate of electrochemical transformation is controlled by diffusion. The relation to solve is A ( t ) = NeffrlmC(x,t)exp( -x/d)dx
(4)
where Neffis a sensitivity factor characteristic of the ote and number of reflections, r is the molar extinction coefficient of the monitored species, and C(x,t) is the concentration of the same species. The solution of eq 4 for species B through Laplace analysis gives ~ ~ (= tNeffaB6CAo{ ) 1 - exp(a2t) erfc(l/i.))
(5)
wherea = 4316. The calculated and experimental A us. t curves are shown in Figure 1, curves a and c, for the one-electron reduction of methyl viologen to the monocation radical C H 3 - t N M N + - C H 3
--+e-
MV2+
300
600
1
900
TI M E b sec
Figure 1. A us. t behavior for MV2+reductionand TaeBrlP oxidation. Curve a, potential stepped beyond Elo’ and curve b, beyond E$‘ for MV. + at h 605 mp,e = 2 X lo4M-’ cm-I in 0.5 M tetraethylammonium perchlorate in acetonitrile. Solid line drawn for CO = 2.5 X lod4M, K,, = lo’, %@IS = 98.6 s~c-’/’, kr 3 X 109 M-1 sec-1. Curve c, potential stepped beyond EP’, and curve for conditions given in Table I. Solid d, beyond E#‘ for TasBrIz2+ line drawn for CAO = 2.7 X M , do/S = 58 sec-1’2,kr = 6.9 X 107 M-1. All scans were obtained by applying 2-msec potential pulse with repetition rate of 50 Hz for 60 sec; t = 25 I 1 ’.
>
alters the A us. t curve from the diffusion-controlled case, as long as this reaction is sufficiently fast such that it occurs within the optical cell defined by 6. For the disproportionation reaction 3 as dictated by the experimental conditions, a working curve for the irs absorbance can be computed by a digital simulation of the concentration profile.5 Experimentally MV2+ undergoes two consecutive one-electron transfers with Elo’ and Ezo‘ of -0.50 and -0.92 V us. sce in acetonitrile.68’ To evaluate the kf of this reaction, the potential is stepped about 200 mV more negative than EzO’ so that the transformation of MV2+to MVO occurs at the electrode at a diffusion-controlled rate. The absorbance of MV .+ is monitored during the potential step as shown in Figure 1, curve b, and kf is evaluated from matching the observed absorbance to the working curve. Similarly, in the case of Ta6Br12*+ two oxidation waves of one electron each are observed with Elo’ and E&” of 0.35 and 0.65 V us. sce.8 The Ta6BrlZ2+ was monitored during potential step, Figure 1, curve d, and the kf for the reaction was evaluated from the workkr e 2Ta6BrlzS+
TasBriz2+4-TaaBnz4+
MVf
and the oxidation of TaBBrlZ2+cluster to the 3+ species, respectively. Agreement with predicted results is within 5 % at 50 psec. The experimental curve is an output from a PAR Waveform Eductor which signal averages the repetitive A us. t curve. The repetition is possible in irs since the optical cell can be “filled” and “dumped” at will by controlling the value and width of the potential pulses. This advantage allows the resolution of small absorbance changes on a time scale heretofore unattainable. Any homogeneous chemical reaction, which perturbs the concentration of the optically monitored species, (3) N. Winograd and T. Kuwana,J. Elecfroanal. Chem., 23,333 (1969). (4) J. H. Espenson, Inorg. Chem., 7,631 (1968).
(7)
ing curve. The data for Ta6BlZ2+are summarized in Table I for two different concentrations. The value of 3 X IO9 M-I sec-1 for the methyl viologen electron exchange between the neutral and divalent cation represents the upper limit attainable with the present experimental ote cell and instrumentation. Limitations are due to the finite resistance of the ote and large power level required for the potentiostat at times less than 10 psec. Irs at ote taking advantage of the repetition of the process allows monitoring of ( 5 ) S. W. Feldberg in “Electroanalytical Chemistry,” Vol. 3, A. J. Bard, Ed., Marcel Dekker Inc., New York, N. Y., 1969. (6) J. W. Strojek, G. A. Gruver, and T. Kuwana, Anal. Chem., 41. 481 (1969). (7) The E”‘ values are reported as formal potentials for the defined solution conditions. (8) N. E. Cooke, J. H. Espenson, and T. Kuwana, to be published.
Communications to the Editor
226 Table I.
Kinetic Data for TaeBr122f
d/zjjSc= 77.2 SW-'/Z; Neif
(A) CAO = 1.0 X lO-'M; 168 378 67 1 1048 1306
(B)
cAo= 2 . 7 x
297 669 1189 1858
7.40 4.82 2.86 1.73 1.35
0.93 1.95 5.10 9.5 10.0
10-4 M ; 5.46 3.08 1.62 1.08
dBjS
= 98
= 9.7 7.71 7.72 7.88 7.94 7.88 AV = 7 . 8 3 + 0 . 1 0
sec-'/l;
6.6 12.5 24.6 27.0
N,tr
= 7.0 7.91 7.84 7.87 7.76 AV = 7.84 0.06
*
aTa8BrlnZ+ monitored at 620 mp, where e = 6000 M-1 cm-I, Keq = 1 X 105. Kinetic data evaluated from potential step beyond Eta', 0.04 M HClO4 used as supporting electrolyte. c fi/S evaluated from potential step beyond Elo'.
any light-absorbing species within the confines of the irs optical cell during an electrochemical perturbation. High sensitivity, due to N e f f>> 1, and signal averaging give absorbance sensitivity better than one part in l o 5 with time resolution in the microsecond range. Detailed mechanism and kinetic studies of fast homogeneous electron exchange reactions following the heterogeneous electron-transfer step at ote are now in progress. Acknowledgment. The authors gratefully acknowledge the financial support provided from No. G M 14036 from the Research Grant Branch of the National Institutes of Medical Science, National Institute of Health, and by the National Science Foundation Grant No. GP9306. One of us (N. W.) gratefully acknowledges support by the National Institutes of Health for a Predoctoral Fellowship, 1968-1969. The authors also appreciate helpful discussions with Stephen W. Feldberg at Brookhaven National Laboratory. The sample of Ta6Brlz2+was generously provided by J. H. Espenson at Iowa State University. Nicholas Winograd, Theodore Kuwana
Department of Chemistry, Case Western Reserce Unioersity Cleoeland, Ohio 44I06 Receioed November 18,I969
Modification of the Wittig Reaction to Permit the Stereospecific Synthesis of Certain Trisubstituted Olefins. Stereospecific Synthesis of a-Santalol Sir: It has been shown'that the reaction between an aldehyde and an alkylidenetriphenylphosphorane to form a 1,2-disubstituted ethylene can be directed efficiently to the trans isomer by interposing a process which modifies the stereochemistry of the intermediate phosphorus betaine. The process which was described involves sequential treatment of the primary betaine (resulting from addition of the Wittig reagent to the aldehyde) with phenyllithium (at -70") and potassium t-butoxide-t-butyl alcohol. Experiments performed in these laboratories indicated that, at least in the cases studied, (1) M. Schlosser and K. F. Christmann, Angew. Chem. Intern. Ed. Engl., 5, 126(1966); Ann. Chem., 708, l(1967).
Journal of the American Chemical Society
92:l
the potassium t-butoxide-t-butyl alcohol reagent is not required for the Wittig-Schlosser synthesis. This finding implied that the intermediate P-oxido phosphonium ylide (the obvious product of the reaction of a phosphorus betaine with phenyllithium) is capable of stereospecific protonation to form the threo betaine which is the precursor of the trans-l,2-disubstituted ethylene finally formed. It seemed logical to expect that only one of the isomeric betaines might be formed specifically by the reaction of electrophiles other than proton donors with P-oxido phosphonium ylides and that a useful stereospecific route to trisubstituted olefins might thereby become available. This communication describes the realization of such a synthetic process. Publication of the results which have been obtained thus far is prompted by the appearance of a note describing a study of reactions of P-oxide phosphonium ylides4 which partially overlaps the present work. Methods which have been described previously permit the stereospecific synthesis of trisubstituted olefins of types la,3351b,315and also 2a,3a,6but not of type 2b. We shall first describe a new approach to the synthesis of R1\
,c=c
/X
H '
R2
la, R1 CH,; R2= alkyl E
lb,R1=alkyl;R2=CH3
(X = CH,OH or COOCH,) R3\
H'
c=c
,R1
\R2
-
2a,R, = CH,; R3 = alkyl; R; = CH,OH or COOCH3
2 b,R1 CHzOHor COOCH3;R2 = CH3;R, = alkyl
olefins of type 2b, via P-oxido phosphonium ylides as key intermediates, which closes this methodological gap. Reaction of a solution of ethylidenetriphenylphosphorane in tetrahydrofuran at - 78 " with heptanal for 5 min produced the betaine 3 which was treated with 1 equiv of n-butyllithium (in hexane) at -78" t o form the deep red P-oxido phosphonium ylide 4. The solution of the relatively stable ylide 4 was allowed to warm to 0" and then treated with 2 equiv of dry paraformaldehyde to generate the colorless P,P'-dioxido phosphonium derivative 5. After a reaction time of 0.5 hr at 0" and 12 hr at 25", the mixture was worked up (6) in the usual way to give 2-methyl-cis-2-nonen-1-01 in 73 yield. The stereochemistry of 6 is clearly indicated by of the nmr spectrum (in CCL) which (2) We suggest this name for the modified' Wittig reactions which afford trans-1,2-disubstituted olefins. (3) For other recent contributions from this laboratory on new stereospecific routes to trisubstituted olefins, see (a) E. J. Corey, J. A. Katzenellenbogen, and G.H. Posner, J . Amer. Chem. Soc., 89, 4245 (1967); (b) E. J. Corey, J. A. Katzenellenbogen, N. W. Gilrnan, S. A. Roman, and B. W. Erickson, ibid., 90, 5618 (1968); (c) E. J. Corey and J. A. Katzenellenbogen, ibid., 91, 1851 (1969). (4) M. Schlosser and K. F. Christmann, Synthesis, 1, 38 (1969), describe the reaction of certain P-oxido ylides with methyl iodide, perchloryl fluoride, iodobenzene dichloride, and bromine. ( 5 ) E. J. Dorey, N. W. Gilman, and B. C. Ganern, J . Amer. Chem. Soc., 90, 5616 (1968). (6) E. J. Corey, K. Achiwa, and J. A. Katzenellenbogen, ibid., 91, 4318 (1969). (7) K. C. Chan, R. A. Jewell, W. H. Nutting, and H. Rapoport, J . Org. Chem., 33,3382(1968).
1 January 14, 1970
