Tin-Free Radical Cyclization Reactions Using the Persistent Radical Effect

Full text of DOI:10.1002/(SICI)1521-3773(20000317)39:6<1108::AID-ANIE1108>3.0.CO;2-A

COMMUNICATIONS
planarity, and number of p electrons, see K. Müllen, Chem. Rev. 1984,
84, 603 ± 646.
[10] The reductions of 1 and 2 were carried out as previously described;
K. Müllen, T. Meul, P. Schade, H. Schmickler, E. Vogel, J. Am. Chem.
Soc. 1987, 109, 4992 ± 5003.
[11] A triad of (4n‡2)p dications, [4n]annulenes, and (4n‡2)p dianions
analogous to 1/3/5 and 2/4/6 has been reported by K. Müllen, T. Meul,
E. Vogel, U. Kürschner, H. Schmickler, O. Wennerström, Tetrahedron
Lett. 1985, 3091 ± 3094.
[12] Depending on whether the p-electron system is planar and fourfold or
just fourfold, the Soret and Q transitions belong to symmetry Eu or E.
[13] P. R. Callis, T. W. Scott, A. C. Albrecht, J. Chem. Phys. 1983, 78, 16 ±
22.
Tin-Free Radical Cyclization Reactions Using
the Persistent Radical Effect**
Scheme 1. Radical 5-exo-cyclization with use of the persistent 2,2,6,6-
tetramethylpiperidin-1-oxyl (TEMPO) radical.
Armido Studer*
Organotin compounds have found widespread application
in conducting various types of radical reactions.[1] Despite this
important utility, there are drawbacks associated with tin-
based radical chemistry, namely, toxicity of organostannanes
that necessitate special handling of disposal, and in many
instances problems with product purification. Therefore,
various research groups have been looking for alternatives.[2]
Herein we report new tin-free radical cyclization reactions
based on the persistent radical effect (PRE).
The PRE[3, 4] is a general principle that explains the highly
specific formation of the cross-coupling product (R1 R2)
between two radicals R1 and R2 when one species is persistent
(long-lived) and the other transient; the two radicals must be
formed at equal rates. The initial buildup in concentration of
the persistent species, caused by self-termination of the
transient radical, steers the reaction to follow a single pathway
for the cross-reaction. The PRE has already been used in
various chemical systems[3, 5] and is important in stable free
radical polymerization (SFRP).[6]
heating.[6, 7] C O bond homolysis in 1 will therefore lead to
the persistent TEMPO radical and the transient radical 2.[6±10]
Radical 2 can either be trapped by TEMPO to reform 1, or it
can undergo a 5-exo cyclization (the 6-endo cyclization is
omitted in Scheme 1) to form a new radical 3, which after
trapping with TEMPO affords 4. The trapping of 3 with
TEMPO is irreversible, because the C O bond in an alkoxy-
amine derived from TEMPO and a primary alkyl radical is too
strong to be homolytically cleaved.[7] Due to the low concen-
tration of TEMPO during the isomerization,[6b] 3 is long-lived
and the cyclization from 2 to 3 is probably reversible.[11]
According to the PRE, the coupling products 5 ± 7 from the
transient radicals 2 and 3 should be formed in very low yields,
and the isomerization of 1 should occur almost quantitative-
ly.[4] Furthermore, degenerate radical reactions (reversible
initial C O bond homolysis in our system) have been shown
to suppress potential side reactions by reforming starting
material when the desired cyclization does not proceed
effectively.[5b]
To test the concept of the PRE in cyclization reactions,
alkoxyamine 1 was prepared from 1-bromo-1-phenyl-5-hex-
ene and the Ca alkoxide (2 equiv) of 1-hydroxy-2,2,6,6-
tetramethylpiperidine in refluxing THF (14 h, 80%).[12, 13]
The isomerization reaction depicted in Scheme 1 was studied
under different conditions (Table 1). In tert-butylbenzene and
N,N'-dimethyl-N,N'-propylene urea (DMPU), no isomeriza-
tion was observed (entries 1 and 2). In DMF after 16 h, 4 was
isolated in 56% yield (trans:cis ˆ 2.5:1)[14] along with the
6-endo product 8 (10%, trans:cis ˆ 1:1; entry 3). In tBuOH
(0.1m) clean but slower isomerization occurred (entry 4) and
lowering the concentration (0.1m ! 0.01m) has no significant
effect (entry 5). Interestingly, the reaction was accelerated by
addition of camphorsulfonic acid (CSA, entry 6),[15] with the
best results obtained in tBuOH (0.02m) with 10% CSA
(entry 7). The isomerization products were isolated in 83%
yield along with 9 (2%).[16]
In Scheme 1, an application of the PRE for a 5-exo
cyclization with the 2,2,6,6-tetramethylpiperidin-1-oxyl
(TEMPO) radical as persistent species is suggested. It is
known that a-phenyl-substituted alkoxyamines such as 1 have
weak C O bonds which are homolytically cleaved upon
[*] Dr. A. Studer
Laboratorium für Organische Chemie
Eidgenössische Technische Hochschule
ETH-Zentrum, Universitätstrasse 16, 8092 Zürich (Switzerland)
Fax : (‡ 41)1-632-1144
E-mail: studer@org.chem.ethz.ch
[**] I am grateful to Prof. Dieter Seebach for generous financial support
and to Prof. Hanns Fischer for helpful discussions. Dr. Sylvain Marque
is acknowledged for conducting the ESR experiments, and Dr. Volker
Gramlich for carrying out the X-ray analysis. I also thank Christian
Wetter and Elisabeth Baier for conducting some experiments.
Supporting information for this article is available on the WWW under
http://www.wiley-vch.de/home/angewandte/ or from the author.
1108
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3906-1108 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 6

COMMUNICATIONS
Table 1. Isomerization of 1 under different conditions (130 ± 1328C, sealed
tube).
Entry
Solvent
Conc. [m]
t [h]
4 [%]
8 [%]
9 [%]
1
2
3
4
5
6
7
tBuPh
DMPU
DMF
tBuOH
tBuOH
tBuOH
tBuOH
0.1
0.1
0.1
0.1
0.01
0.01
0.02
22
14
16
16
36
24[d]
24[d]
< 2[a]
< 2[a]
56
< 2[a]
< 2[a]
10
20
< 2[a]
< 2[a]
< 2[a]
< 2[a]
< 2[a]
2
n.d.[b]
n.d.[c]
53
n.d.[b]
n.d.[c]
10
70
13
[a] In the 300-MHz 1H NMR spectrum of the crude product, no signals of
the corresponding compound were observed. [b] After 16 h, 62% of the
starting material remained. Products 4 and 8 were not isolated. [c] After
36 h, about 10% of starting material remained. Products 4 and 8 were not
isolated. [d] 10% CSA was added.
Scheme 2. Cross-over experiment with an equimolar mixture of the
alkoxyamines 1 and 10. a) tBuOH (0.02 m), 1308C, 10% CSA, 24 h;
b) TBAF, THF. TBAF ˆ tetrabutylammonium fluoride; TBDMS ˆ tert-
butyldimethylsilyl.
It is known that the reaction of nitroxides with various
C-centered radicals is dependent on solvent, with decreasing
rates in polar protic solvents.[17] Furthermore, the homolytic
cleavage is faster in polar solvents.[18] These synergistic
effects[6b] lead to the observed rate enhancement in the
isomerization upon using polar solvents.
An ionic mechanism (C O heterolysis) can be excluded
since clean cationic isomerizations in tBuOH are unlikely. To
further corroborate the radical nature of the process, we also
conducted ESR experiments. A deoxygenated sample of 1 in
tBuOH/tBuPh (1/1) containing an excess of the C-radical
scavenger
2,2,10,10-tetraperdeuteromethyl-isoindolin-15N-
oxyl (10 equiv) was heated to the reaction temperature, and
the evolution of the ESR signal of the TEMPO radical was
followed by continuous wave (CW) ESR spectroscopy.[19]
Almost 100% of the TEMPO was generated, thus establish-
ing initial C O bond homolysis under these conditions.
To exclude a mechanism where isomerization occurs in a
solvent cage, we also conducted a cross-over experiment. The
alkoxyamines 1 and 10 (1/1 mixture) were isomerized under
the optimized conditions. After desilylation, roughly equal
amounts of the scrambled alkoxyamines 4 (19%), 11 (19%),
12 (16%), and 13 (16%) were isolated (Scheme 2).[20]
To further examine the scope and the limitations of the
reaction, we tested whether isomerization also occurs with
alkoxyamines 14a ± j (variation of substitutent R, Ta-
ble 2).[21±23] For bromide 14a, a clean reaction occurred, and
15a was isolated in 71% (trans:cis ˆ 2.7:1)[24] with 16a (8%,
entry 1). Isomerization of 14b led to 17b (10%), and the
products 15b and 16b were isolated in 54% yield (15b:16b ˆ
5.8:1, entry 2). Clean reactions were observed for heteroar-
enes 14c and 14d (entries 3 and 4). With 14e, no 6-endo
product was formed, and 15e was isolated in 67% yield
(d.r. ˆ 1:1, entry 5) along with 10% of 17e. The reaction with
nitrile 14 f afforded 61% of the 5-exo product and 7% of 16 f
(entry 6). No isomerization occurred with 14g (R ˆ H), 14h
Table 2. Isomerization of 14a ± f under optimized conditions (tBuOH,
0.02m, 24 h, 130 ± 1328C, 10% CSA, sealed tube). Yields refer to
chromatographically (SiO2) purified compounds.
Entry
Starting
material
15 [%]
d.r. (15)
(trans:cis)
16 [%][a]
17 [%]
1
2
3
4
5
6
14a
14b
14c
14d
14e
14 f
71
46
67
57
67
61
2.7:1
2.8:1
2.1:1
1.6:1
1.1:1
1.1:1[c]
8
8
11
12
< 2[b]
7
< 2[b]
10
5
5
10
< 2[b]
[a] The 6-endo product was formed as a 1:1 mixture of the diastereoisom-
ers. [b] In the 300-MHz 1H NMR spectrum of the crude product, no signals
of the corresponding compound were observed. [c] The relative config-
uration of the two isomers was not assigned.
(R ˆ Me), and 14i (R ˆ SPh). It is likely that the C O bond in
14g ± i is too strong to be effectively cleaved.[7] As expected,
sulfone 14j was not stable at higher temperatures (elimination
of PhSO2H).
Angew. Chem. Int. Ed. 2000, 39, No. 6
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3906-1109 $ 17.50+.50/0
1109

COMMUNICATIONS
We then tested alkoxyamines 18 ± 22 (R ˆ 1-phenyl-5-
hexenyl, variation of the nitroxide) in the isomerization.[25±28]
The reactions were conducted in tBuOH (0.02m) with or
To further illustrate the potential of the method, we
conducted a radical cascade reaction (Scheme 3). Enone
30[31] was isomerized to give the angular triquinane 31 in 78%
yield as the expected mixture of stereoisomers.[32] N O bond
cleavage with Zn in AcOH/THF/H2O (3/1/1) afforded the
hemiacetals 32a, b (40%) and the alcohols 33a, b (13%). The
relative configuration of the major isomer 32a was assigned
by X-ray crystallographic analysis.[33]
without added CSA (10%) at 1308C, and the time necessary
for complete consumption of the starting material was
measured. The success of the isomerization is dependent on
the bond dissociation energy (BDE) of the C O bond (see
above), which should correlate with the reaction time. The
fastest isomerization (3 h) was observed for 18. This accel-
eration, as compared to the cyclization of 1 (24 h with CSA), is
not surprising, since it is known that alkoxyamines derived
from di-tert-butyl nitroxide have very weak C O bonds.[18]
Only 4.5 h were necessary to completely isomerize phospho-
nate 19. Addition of CSA led to decomposition of the starting
alkoxyamine (for 18 and 19).
Based on results from Ingold and Beckwith on the
stabilization of nitroxides by intermolecular hydrogen bond-
ing,[17] we decided to prepare new nitroxides capable of
intramolecular hydrogen bonding (23).[29] In addition to
decreasing the trapping rate of the nitroxide with C-centered
radicals, the hydrogen bonding should also influence the rate
of C O bond homolysis (late transition state). Indeed, 21 was
observed to isomerize faster (7 h; 12 h without CSA) than its
silylated congener 20 (16 h with CSA). With 22, reaction was
already completed after 4 h (9 h without CSA).
Scheme 3. Radical cascade reaction.
In conclusion, the new tin-free radical cyclization reactions
afford products which can be considered as protected alcohol
derivatives.
Received: November 4, 1999 [Z14225]
The advantage of our new alkoxyamines over derivatives
derived from commercially available nitroxides is document-
ed by the failure of 24 (no reaction) and 25 (decomposition) to
isomerize and by the successful isomerization of triol 26
(52%).[30] We also conducted a slow 6-exo cyclization of a
phenyl-stabilized radical. Reaction of 27 under Bu3SnH
conditions (1.2 equiv of hydride, syringe pump, 7 h, benzene,
0.1m) afforded only 11% of the cyclized product besides the
dehalogenation product 28. Alkoxyamine 29, however, yield-
ed the corresponding 6-exo product in 48% yield. An
additional advantage, besides higher yield and lower toxicity
in the reaction of 29 as compared to the tin hydride mediated
cyclization of 27, is the fact that the cyclization product is
functionalized (protected alcohol) in the former case.
[1] A. G. Davies, Organotin Chemistry, Wiley-VCH, Weinheim, 1997;
D. P. Curran, Synthesis 1988, 417; D. P. Curran, Synthesis 1988, 489.
[2] P. A. Baguley, J. C. Walton, Angew. Chem. 1998, 110, 3272; Angew.
Chem. Int. Ed. 1998, 37, 3072.
[3] B. E. Daikh, R. G. Finke, J. Am. Chem. Soc. 1992, 114, 2938.
[4] H. Fischer, J. Am. Chem. Soc. 1986, 108, 3925; D. Rüegge, H. Fischer,
Int. J. Chem. Kinet. 1989, 21, 703; T. Kothe, S. Marque, R. Martschke,
M. Popov, H. Fischer, J. Chem. Soc. Perkin Trans. 2 1998, 1553.
[5] a) For early work see references cited in ref. [3]. See also D. C.
Harrowven, G. Pattenden, Tetrahedron Lett. 1991, 32, 243; B. Giese, P.
Erdmann, T. Göbel, R. Springer, Tetrahedron Lett. 1992, 33, 4545; B. P.
Branchaud, G.-X. Yu, Organometallics 1993, 12, 4262; J. Hartung, B.
Hertel, F. Trach, Chem. Ber. 1993, 126, 1178; P. A. Macfaul, I. W. C. E.
1110
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3906-1110 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 6

COMMUNICATIONS
Arends, K. U. Ingold, D. D. M. Wayner, J. Chem. Soc. Perkin Trans. 2
1997, 135; A. Bravo, H.-R. Bjorsvik, F. Fontana, L. Liguori, F. Minisci,
J. Org. Chem. 1997, 62, 3849; T. Kothe, R. Martschke, H. Fischer, J.
[25] Olefins 18 ± 22 were prepared from 1-bromo-1-phenyl-5-hexene and
the corresponding protected nitroxides (tert-butyldimethylsilyl
(TBDMS) protecting groups for 20[26] and orthoester protection for
triol 22 (d.r. ˆ 1:1, see the Supporting Information)) using a CuI-
mediated process.[27] Phosphonate 19 (d.r. ˆ 1:1) was prepared from
the corresponding nitroxide[28] in an analogous manner. In the
isomerizations of 18 ± 22 the corresponding cyclized compounds (5-
exo and 6-endo) were obtained in high yields (>60%).
[26] While we were carrying out our experiments, similar alkoxyamines
were suggested for SFRP : D. Benoit, V. Chaplinski, R. Braslau, C. J.
Hawker, J. Am. Chem. Soc. 1999, 121, 3904. The nitroxides were
prepared in analogy.
Chem. Soc. Perkin Trans.
2 1998, 503. b) Degenerate radical
reactions: B. Quiclet-Sire, S. Zard, Pure Appl. Chem. 1997, 69, 645.
[6] a) D. H. Solomon, E. Rizzardo, P. Cacioli, US-A 4,581,429, 1985;
M. K. Georges, R. P. N. Veregin, P. M. Kazmaier, G. K. Hamer,
Macromolecules 1993, 26, 2987; C. J. Hawker, J. Am. Chem. Soc.
1994, 116, 11314; C. J. Hawker, Trends Polym. Sci. 1996, 4, 183; b) H.
Fischer, Macromolecules 1997, 30, 5666.
[7] M. V. Ciriano, H.-G. Korth, W. B. van Scheppingen, P. Mulder, J. Am.
Chem. Soc. 1999, 121, 6375.
[8] Heterolytic O N bond cleavage is possible in a-acyl-substituted
alkoxyamines under acidic conditions: W. G. Skene, T. J. Connolly,
J. C. Scaiano, Tetrahedron Lett. 1999, 40, 7297.
[9] Homolytic O N bond cleavage in phenyl-substituted, TEMPO-
derived alkoxyamines has not been observed.[7] The O N bond is
stronger than the C O bond in PhCH2-TEMPO (BDE(C O):
[27] K. Matyjaszewski, B. M. Woodworth, X. Zhang, S. G. Gaynor, Z.
Metzner, Macromolecules 1998, 31, 5955.
[28] S. Grimaldi, J.-P. Finet, A. Zeghdaoui, P. Tordo, D. Benoit, Y. Gnanou,
M. Fontanille, P. Nicol, J.-F. Pierson, Polym. Prepr. Am. Chem. Soc.
Div. Polym. Chem. 1997, 38, 651.
[29] Intramolecular hydrogen bonding has already been suggested: E. G.
Janzen, J. I-Ping Liu, J. Magn. Reson. 1973, 9, 510.
[30] Due to the complex NMR spectra, the 5-exo:6-endo ratio was not
determined.
1
[7]
32 kcalmol
;
BDE(O N) ꢀ 53 kcalmol 1 [10]). In strained cyclic
alkoxyamines, N O bond homolysis is possible : F. M. Cordero, I.
Barile, F. De Sarlo, A. Brandi, Tetrahedron Lett. 1999, 40, 6657, and
references therein.
[31] The synthesis of 30 (d.r. ˆ 1:1, racemic) is described in the Supporting
[10] H. G. Aurich, Nitrones, Nitronates and Nitroxides (Eds.: S. Patai, Z.
Rappoport), Wiley, Chichester, 1989.
[11] H. Pines, N. C. Shi, D. B. Rosenfield, J. Org. Chem. 1966, 31, 2255; C.
Walling, A. Cioffari, J. Am. Chem. Soc. 1972, 94, 6064.
[12] The Ca alkoxide was prepared by treating a suspension of TEMPO in
H2O with 1 equiv of the calcium salt of l-ascorbate dihydrate.[13]
[13] C. J. Hawker, G. G. Barclay, A. Orellana, J. Dao, W. Devonport,
Macromolecules 1996, 29, 5245.
[14] The relative configuration was assigned after transformation of 4 to
the corresponding known alcohol (Zn, AcOH, H2O, THF).
[15] CSA was also used as an additive in SFRP: M. K. Georges, R. P. N.
Veregin, P. M. Kazmaier, G. K. Hamer, M. Saban, Macromolecules
1994, 27, 7228; P. G. Odell, R. P. N. Veregin, L. M. Michalak, D.
Brousmiche, M. K. Georges, Macromolecules 1995, 28, 8453; R. P. N.
Veregin, P. G. Odell, L. M. Michalak, M. K. Georges, Macromolecules
1996, 29, 4161; M. V. Baldovi, N. Mohtat, J. C. Scaiano, Macro-
molecules 1996, 29, 5497; T. J. Conolly, J. C. Scaiano, Tetrahedron Lett.
1997, 38, 1133.
Information.
[32] D. P. Curran, S.-C. Kuo, J. Am. Chem. Soc. 1986, 108, 1106; A. Matzeit,
H. J. Schäfer, C. Amatore, Synthesis 1995, 1432.
[33] Crystallographic data (excluding structure factors) for the structure
reported in this paper have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication no.
CCDC-137891. Copies of the data can be obtained free of charge on
application to CCDC, 12 Union Road, Cambridge CB21EZ, UK (fax:
(‡44)1223-336-033; e-mail: deposit@ccdc.cam.ac.uk).
Pseudoprolines: Targeting a cis Conformation
in a Mimetic of the gp120 V3 Loop of HIV-1**
[16] 1,2,3,3a,8,8a-Hexahydrocyclopent[a]indene[11] was formed in very low
yield (<2%) under our conditions.
[17] A. L. J. Beckwith, V. W. Bowry, K. U. Ingold, J. Am. Chem. Soc. 1992,
114, 4983.
Angela Wittelsberger, Michael Keller, Leo Scarpellino,
Luc Patiny, Hans Acha-Orbea, and Manfred Mutter*
[18] G. Moad, E. Rizzardo, Macromolecules 1995, 28, 8722.
[19] The ESR experiments were carried out in the laboratory of Prof.
Hanns Fischer at the University of Zürich. For experimental reasons
the reactions were conducted in tBuOH/tBuPh (1/1). We also
performed the isomerization under the conditions applied in the
ESR experiment. Clean but slower isomerization was observed in
tBuOH/tBuPh (1/1) without the use of CSA. Experimental details are
given in: S. Marque, C. Le Mercier, P. Tordo, H. Fischer, Macro-
molecules, submitted.
[20] The 6-endo products also formed in the cyclization are omitted in
Scheme 2 for reasons of clarity. For similar cross-over experiments in
polymerizations, see: C. J. Hawker, Acc. Chem. Res. 1997, 30, 373.
[21] Arenes 14a ± d were prepared in analogy to 1 (unoptimized, 23 ±
33%). Ester 14e was obtained by treatment of the corresponding Li
enolate with CuCl2 and TEMPO (59%).[22] Nitrile 14 f was prepared
after deprotonation (lithium diisopropylamide, LDA) of 6-cyano-1-
hexene and subsequent addition of a suspension of CuCl2 and
TEMPO in THF at 788C (67%). Olefin 14g was prepared from
6-iodo-1-hexene, Mg, and TEMPO (61%).[13] Treatment of 6-iodo-1-
heptene with TEMPO (6 equiv) and (Me3Si)3SiH (4 equiv) in reflux-
ing benzene (0.1m) afforded 14h (97%). Thioether 14i was prepared
from the corresponding chloride[23] in analogy to 1. Sulfone 14j was
available from phenyl-(5-hexenyl)sulfone after deprotonation (LDA)
and subsequent oxidation (CuCl2) in the presence of TEMPO (36%).
[22] R. Braslau, L. C. Burrill II, M. Siano, N. Naik, R. K. Howden, L. K.
Mahal, Macromolecules 1997, 30, 6445.
Pseudoprolines (YPro) have been introduced recently as
synthetic proline analogues readily obtained by cycloconden-
sation of the amino acids cysteine, threonine, or serine with
aldehydes or ketones.[1] Their application as structure-dis-
rupting, solubilizing protecting groups in solid-phase peptide
synthesis[2, 3] was followed by conformational investigations
concerning the YPro preceding peptide bond.[4, 5] In fact, the
propensity of the amino acid proline for forming a Xaai 1-Proi
[*] Prof. Dr. M. Mutter, Dipl.-Chem. A. Wittelsberger, Dr. M. Keller,
Dr. L. Patiny
Institute of Organic Chemistry
University of Lausanne
BCH-Dorigny, 1015 Lausanne (Switzerland)
Fax: (‡41)21-692-39-55
E-mail: Manfred.Mutter@ico.unil.ch
L. Scarpellino, Prof. Dr. H. Acha-Orbea
Ludwig Institute for Cancer Research
Lausanne Branch and Institute of Biochemistry
University of Lausanne ISREC
Ch. des Boveresses 155, 1066 Epalinges (Switzerland)
[**] We are grateful to Dipl.-Biol. Raymond Jacquet for helpful advice.
This work was supported by the Swiss National Science Foundation.
[23] D. L. Tuleen, T. B. Stephens, J. Org. Chem. 1969, 34, 31.
[24] The relative configuration of 15a±d was assigned in analogy to that of 4.
Supporting information for this article is available on the WWW under
http://www.wiley-vch.de/home/angewandte/ or from the author.
Angew. Chem. Int. Ed. 2000, 39, No. 6
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3906-1111 $ 17.50+.50/0
1111