Table 1. Rates and selectivities of alkali metal cation transport via cryptate complexes [4]. The cation concentrations in the membrane were measured at the end of the experiments and may be somewhat (< 5%) lower than the initial concentrations because of the slight pH change (about 0 .14 .2 pH units).

Carrier Cation log K . Cation conc. Carrier Initial Transport [mol-'1 in membrane saturation transport selectiviy [a1 [~molf l l Ebl [%I [CI rate [pmoI/h] K + : N a + Cs+ :K+

( 1 ) Na K f cs +

(2) Na+ K + c s +

(3) Na +

K + cs +

( 4 ) Na +

K + c s +

7.2 [d] 9.7 [dl 4.4 5.0 7.0 [d] 7.0 [d] 2.7 5.4 5.9 3.5 5.2 2.7

1400 > 1400

130 750

1 loo 3 1450

110 950

1250 140 850

7

95 > 95

20 60 75

295 10 65 85 15 80 0.7

0.6 0.03 2.9 2.4 0.7 0.6 1.5 2.7 3.0 1.6 3.6 0.07

1:20 1:O.Ol

1 :3.5 1 : 1.25

1:0.55 1:0.9

1:0.45 1:SO

[a] K,=stability constant of cryptate complex in methanol [2]. [b] K + . C s + : +5" / ;Na+: +lo"/,. [c] For the weakest complexes, the cation and picrate concentrations in the membrane may differ significantly because of competition of cation extraction with ligdnd protonation. The carrier saturation has been calculated with respect to unprotonated carrier in the membrane. The amount of protonated carrier is equal to the difference between the amounts of picrate (UV) and cation (atomic absorption) in mol. [d] These values are for methanol containing 5% water. Somewhat higher values are expected for pure methanol (about k0.5 log K J

tated diffusion down the concentration gradient of the salt. No transport is detectable in the absence of carrier.

2) Carrier-cation pairs forming very stable cryptate com- plexes display efficient extraction of the salt into the organic phase.

3) The rate of transport strongly depends on both the cation and the carrier (anion effects are not considered here). It is of the same order of magnitude as observed for antibiotic- mediated transport across a bulk liquid membraneI61.

4) The transport selectivities differ widely for each of the carriers (1)-(4). However, they are all lower than the com- plexation selectivities in methanol.

5) In contrast to observations for antibiotic-mediated trans- port the relative transport rates are not proportional to complex stability and extraction efficiency. In the case of (I) they are actually opposite to each other. Thus 12 pmol of (I), whose K + cryptate is very stable, transports only 3 pmol of potassium picrate in 100 h! The rate of transport is deter- mined either by a diffusional step or by the rate of hetero- geneous reaction at the interfaces[*].

6) Under the present conditions there is an optimum stability of the cryptate complex for efficient transport : log K , z 5. Thus (2) and (3) display the same complexation selectivity for Na+ and K + but opposite transport selectivity. Interestingly, this value amounts to 4.9 for the efficient natural K+ carrier valino- mycin (in methanol['').

7) Finally, comparison of the carrier properties of ( 1 ) and ( 4 ) shows the effect of a simple structural modification on the sequence of stability constants and transport rates. ( 4 ) carries Na+ and K + much faster than (I); also the very high Cs+/K+ stability of ( 4 ) surpasses that found for most antibioticd'l. This may simply be rationalized by the fact that ( 4 ) retains the same intramolecular cavity size as (1) but removal of two oxygen binding sites lowers all stability constants and places the cryptate complex of K + and ( 4 ) in the optimal stability range for efficient transport. This structural modification transforms a specific K + receptor ( 1 ) into a specific K + carrier ( 4 ) . The same principle may be used for the design of other specific carriers.

Transport in Organic Chemistry, Part 3.-Part 2: J . M . Lehn, A. Morad- pour, and J . P. Behr, J. Amer. Chem. SOC. 97, 2532 (1975). J . M . Lehn, Struct. Bonding 16, 1 (1973); B. Dietrich, J . M . Lehn, J . P. Sauuage, and J . Elanzat, Tetrahedron 29, 1629 (1973); 8. Dietrich, J . M . Lehn, and J . P. Sauuage, ibid. 29, 1647 (1973); J . M . Lehn and J . P . Sauuage, Chem. Commun. 1971, 440; J . M . Lehn, J . P . Sauuage, and B. Dietrich, J. Amer. Chem. SOC. 92, 2916 (1970). S . G. A. McLaughlin, G. Szabo, S . Ciani, and G. Eisenmann, J. Membrane Biol. 9, 1 (1972). The experimental set-up and the transport cell cf. ref. [ S ] . Two water phases (2x401111) (IN and OUT) are bridged by a chloroform mem- brane (8.5 ml) in which the carrier is dissolved (1.5 mmol I - ' , i .e. 12.7pmo1/8.5ml). The area of the interface is about 2cm2. Initial alkali metal picrate concentrations were lOmmoll-' and 0.01 mmoll-' in the IN and OUT water phases, respectively. All layers are stirred at 300rpm and maintained at 20.Ok0.5"C. The transport process was followed by monitoring the increase in picrate (UV absorption at 354 nm) and cation (atomic absorption) concentrations in the OUT phase. The membrane phase was also analyzed at the end of each experiment. Since the carriers are bases the set of experiments described here were performed using as water phases a iminodiethanol/HCl buKer at pH 10. The transport rates were determined from the slope of the tangent at each point after phase equilibration (about 1 h after addition of carrier). K. Sollner in J . N . Sherwood, A. I! Chadwick, W M . Muir, and F. L. Swinton: Diffusion Processes. Gordon and Breach, London 1971, Vol. 2, p. 655. R. Ashton and L. K . Steinrauf, J. Mol. Biol. 49,547 (1970); R. W Roeske, S . Isaac, 'I: E. Xing, and L. K . Steinrauf, Biochem. Biophys. Res. Commun. 57, 554 (1974). E. Grell, 'I: Funck, and F. Eggers, Symposium on Molecular Mechanisms of Antibiotic Action on Protein Biosynthesis and Membranes. University of Granada, June 1971. J . A. Jacquez, Biochim. Biophys. Acta 79, 318 (1964); P. Liiuger and G. Stark, ibid. 211,458 (1970); M . Kirch and J . M . Lehn, to be published. B. C . Pressman and D. H . Haynes in D. C. Tostrson: The Molecular Basis of Membrane Function. Prentice-Hall, Englewood Cliffs 1969, p. 21 1 ; B. C . Pressman, Ann. N. Y. Acad. Sci. 147, 829 (1 969).

Polyenes as FourStage Redox Systems[**]

By SiegFied Hllnig, Michael Horner, and Peter Schillingp]

Granted suitable structural prerequisites['', stepwise transfer of two electrons often occurs (Wurster radical cations, semi-

Received: April 18, 1975 [ Z 235 IE] German version: Angew. Chem. 87, 542 (1975)

CAS Registry numbers:

14). 31250-03-0; Na. 7440-23-5; K, 7440-09-7; Cs, 7440-46-2 ( I ) , 23978-09-8; (2 ) . 31255-22-8; (3). 31255-19-3;

[*] Prof. S. Hunig [ +]. Dipl.-Chem. M. Horner, and Dr. P. Schilling Institut fur Organische Chemie der Universitat 87 Wiirzburg, Am Hubland (Germany)

[ +] To whom correspondence should be addressed. [**I This work was supported by the Fonds der Chemischen Industrie, by Dechema Frankfurt and by BASF, Ludwigshafen.

556 Angew. Chem. internat. Edit. 1 Vol. 14 (1975) No. 8

quinone radical anions). However, very few four-stage redox systems are known in which the oxidized form oxk of the two-stage cathodic redox system is identical with the reduced form RED, of the two-stage anodic redox system (SEM denotes the radical ion):

- e - e - e

+ C + e + e + e RED,? e S E M ~ 2 oxk = RED, s SEM: + ox:@

Characteristic representatives are polycyclic arenes such as anthraceneL2] and perylenet3!

We now describe polyenes as four-stage redox systems. Of long-chain polyenes (e. g. 0-carotene) it is known only that they can be reversibly reduced to radical anions (SEMf)[4al or oxidized in one step to dications (OX:@)‘4b1. However, oxidation of e. g. butadiener5] and hexatriene[‘l to radical cations (SEM:) occurs irreversibly.

Because of the great stability of the pentadienyl cation (2)17] which is formed on protonation of the crossed triene ( I ) ( ” , we synthesized the vinylogous polyenes (3)-(5)[’] which contain this system and studied their redox behavior[g!

(a) , R = H ( b ) , R = CH,

In dichloromethane from which the last traces of nucleo- philes have been removed by suspended (3)-(5) behave reversibly at -60°C both in the direct current polaro- gram and in the cyclic voltammogram[”], so that E t and E l can be determined. Thus we have here the first two-stage reversible oxidation of polyenes not involving aromatization during the transition into OX$@[”].

+ e

- e +r (3)-(5) SFMF - (3) - ( 5 ) SEAT?

The reduction potentials of (3)-(5) appear with very negative values so that measurements must be performed in DMF/C6H614]. All three polyenes show reversible electron transfer for the step OX,/SEMF (=EL,) even at room tempera- ture with half-lives for SEMf of 10-20s. However, even at - 60 “C and with 20 VJs scan rate only for (5) could the stageSEMp/REDie (= E:) beobtained reversibly, the half-life for REDie being 4s (cf. Table 1). The polyene ( 5 ) can thus occur in five oxidation stages, which are connected by four

reversible single-electron transfers. (3) and ( 4 ) differ there- from only insofar as the lifetime of REDfe is too short for observation of a reversible redox step.

Table 1. The potentials Ef, E:, Eq, and E; [V] of the polyenes (3)-(5) determined from direct current polarograms or cyclic voltammograms (200 mV/s) against Ag/AgCl/acetonitrile with Bu,” B e as supporting elec- trolyte at - 60°C; also the semiquinone formation constants &EM.

E: E: K ~ E M E; Ei K ~ E M in DMF/benzene (2: 1) in dichloromethane

(3) -2.44 irr. -1.95 - +0.60 +0.68 8.1 x 10’ (4 ) -2.8 irr. -1.83 - +0.59 +0.68 1 4 x 1 0 2 (5) -1.99 -1.67 4 x 1 0 ’ +0.56[a] +0.60[a] 4

[a] Index potentials according to B. Elema, Rec. Trav. Chim. Pays-Bas 54, 76 (1935).

That the special structure of (3)-(5) is responsible for the unusual redox behavior is indicated by comparison with the “isolated end groups” (1 a) and (1 b): the maxima in the cyclic voltammogram show irreversible oxidation [(1 a): +1.5OV; (1 6): +1.23V] and irreversible reduction [ ( I a): - 2.80 V; ( I 6) : - 2.93 V]. Further, these potentials are appre- ciably more positive than E t or negative than E: of (3)-(5).

Whereas for the polycyclic arenes the potentials El and E: lie symmetrically around the zero point this is not so for the polyenes (3)-(5). With a zero point potential of the working electrode of -0.22V (against Ag/AgC1 in CH3CN) and E!=+O.60V determined for (I) , E: should appear at - 1 . I 5 V, whereas the experimental value is - 1.96 V.

This behavior may be due to the donor action of the methyl groups. According to MO calculations[13’ they lower the HOMO in arenes and raise the LUMO, which is equivalent to a cathodic shift of the oxidation and reduction potential. The unexpectedly large difference between K&M and K ~ M of (5) could have the same reason. In arenes these values are “I.

As has been shown for several vinylogous two-stage redox systems, there is a good linear relation between the Coulomb repulsion integral JmmsEM and log KSEM, with K ~ E M becoming rapidly smaller with increasing chain 16]. According to Table 1 this is not so for KiEM of (3)-(5). The cause of this deviation can be seen in the twisting of the n-system by the methyl group^"^. 17! This could affect the whole rr-sys- tem and thus K& would be smallest for the shortest polyene ( 3 ) .

Received: April 25, 1975 [Z 237 IE] German version: Angew. Chem. 87, 548 (1975)

CAS Registry numbers: (3 ) , 55712-11-3; ( 4 ) , 55712-12-4; (5). 55712-13-5

[l] E. Weitz, Angew. Chem. 66, 658 (1954); S. Hiinig, Liebigs Ann. Chem. 676, 32 (1964); S. Hunig, H. Balli, H . Conrad, H . Friedrich, J . Gross, G. Kiesslich, G. Ruider, G. Sauer, and D. Scheutzow, Pure Appl. Chem. IS, 109 (1967).

[2] 0. Hammerich and V. D. Parker, 3. Amer. Cbem. SOC. 96, 4289 (1974); B. S. Jensen and

[3] V. D. Parker, .I. Amer. Chem. SOC. 96, 4289 (1974). [4] a) l! G. Mairanovskii, A. A. Engovatou, and G. J . Samokhualov, Zh.

Org. Khim. 1970, 6 (3). 632; b) l! G. Mairanouskii, A. A. Engovatou, N. T lofle, and G. J . Samokhalou in L. G. Feoktistou: Novosti Elektro- khim. Org. Soedin., Tezisy Dokl. Vses. Soveshch. Elektrokhim. Org. Soedin. 8th, 1973. Zinatne, Riga 1973; Chem. Abstr. 82, 36616n (1975); Elektrokhimiya 11, 184 (1975).

D. Parker, J. C. S. Chem. Comm. 1974, 367.

[5] K . K . Barnes and C . K . Mann, J. Org. Chem. 32, 1322 (1967). [6] A. Zweig, A. K . Hoflmann, D. L. Mariele, and A. H . Maurer, Cbem.

Commun. 1967, 106. [7] (2a): W u. E. Doering, M . Saunders, H . G. Boyton, H . W Earhart,

E. F. Wadley, W R. Edwards, and G. Laber, Tetrahedron 4, 178 (1958); C. Jut& W Miiller, and E. Mitller, Chem. Ber. 99, 2479 (1966); ( 2 b ) : S. Hiinig and P. Schilling, ibid., in press.

Angew. Chem. internat. Edit. / Vol. 14 (1975) / No. 8 557

S. Huiiig and P. Schilling. Liebigs Ann. Chem., in press. M. Hornrr and S. Hunig. Liebigs Ann. Chem., in press. 0. Hurnnirridi and I/: D. Purker. Electrochim. Acta 1973, 537. The criteria are fully satisfied [cf. R S. Nicholsori and I . Shuin, Anal. Chem. 36, 706 (1964)l. S. f f u n i g . D. Scheurzorc. and H . J. fr iedrich, Angew. Chem. 7 6 , 818 (1964); Angew. Chem. intcrnat. Edit. 3. 701 (1964). G. J. Horrink and J. ran Scliootm, Rec. Trav. Chim. Pays-Bas 71. 10x9 (1952); E . S. Pysh and N . C . Yung, J . Amer. Chem. SOC. 85, 2124 (1963). P. Curski. S. Hiinig. D. Sclrrurmt'. and R. Zuhrudnik, Tetrahedron

S. Hunig, D. Scheufzow, H. Schluf , and H . Piirtrr. Liebigs Ann. Chem. 1974, 1436. S. Hunig and H c'. Sreinmrrzrr, Tetrahedron Lett. 1971, 643; Liehigs Ann. Chem., in press; S. Hunig. F . L i ~ ~ l i a r t , and D. Scheurxw, ihid., in press. S. Hunig. J . Croas. and W Schrrtk. Liebigs Ann. Chem. 1973, 324.

25.4781 I i 969).

Building Units for Oligosaccharides. Synthesis of a-Gly- cosidically Linked 2-Amino Sugar Oligosaccharides

By Hans Paulsen and Wolfgang Stenzel"] The problem of direct selective synthesis of cc-glycosidically

linked 2-amino sugar oligosaccharides has not previously been solved. An indirect preparation that affords an cc-glycoside preferentially is possible by addition to nitrosoglycals accord- ing to Lemieux's method['- 21, but the difficulty of selective hydrogenation of the 2-ketoxime produced then remains to be solved.

The azido group is an excellent "blocking group" for amino groups; it exerts no neighboring group effect. After having found a simple method of preparing 2-azido-2-deoxy-~-glu- C O S ~ [ ~ ~ we were able to show that halogen derivatives of this compound can be used to accomplish glycoside syntheses with good yields and high selectivity.

The 2-azido sugar ( l a ) affords the 3.4di-0-benzyl com- pound f l h ) whose 1,6-anhydro ring can be opened by acetol- ysis, yielding the diacetate ( 2 ) (64%; m.p. 97.C; [a]do= +62"). The a-bromide (3) (82%; m.p. 84°C; [cr]do= + 101") can be obtained by treatment of (2) with HBr in dichloromethane. Compound (3) can be used for selective synthesis of P-glycosides; for example, it reacts with benzyl alcohol in the presence of Ag2C03 to give a 72% yield of a product that consists to the extent of 802, of the 0-glucoside ( 6 a ) (m.p. 83°C; [a],$O= -4").

The desired a-glycoside synthesis can be effected in the following way: The cr-bromide (3) can be rearranged to the P-chloride by tetraethylammonium chloride[41 in acetonitrile. The rearrangement should be followed polarometrically and must be stopped at exactly the maximum negative rotation. For glycoside synthesis the extremely reactive P-chloride ( 4 ) (80 % ; a syrup) should be used directly since it anomerizes to the e-chloride on standing.

The C-6-unsubstituted compound ( 6 h ) , which is readily obtained from ( 6 a ) , reacts with ( 4 ) in dichloromethane in the presence of AgzCO3, Drierite and catalytic amounts of AgCIOd". '1 in 5 min at - 5 "C to give a 75 yield of a disaccharide containing 90 y> of the a-glycoside (7) which crystallizes well (m. p. 91 'C; [a];'= +49"). The linkage step can be continued without difficulty. If the disaccharide (10) obtained by partial deblocking of ( 7 ) is treated with ( 4 ) under the same conditions as before, a 71 % yield of a trisac- charide is isolated, 90 'x of which was shown to be the cr-glyco- side ( 1 1 ) (m.p. 77-82°C; [a],$o= +72").

Reaction of ( 4 ) with less reactive secondary hydroxyl groups can also be accomplished. The C-3-unsubstituted compound ( I a), with ( 4 ) , gives a 36 yield of a product containing 80% of the a-glycoside ( 1 2 ) (m.p. 110-111"C;

N3

191

~.

[*] Prof. Dr. H. Paulsen and Dipl.-Chem. W. Stenzel Institut fur Organische Chemie und Biochemie der Universltat 2 Hamburg 13. Papendamm 6 (Germany)

[cr]6'= +990"). A suitable C-4-unsubstituted reactant could be obtained from the 1,6-anhydro sugar (5). The epoxide ring of ( 5 ) could be opened by means of NaN3 to afford

558 Angcw. Chrm. blrr~rnat. Edir. ,' Vid. 14 (IY75) 1 No 8