Acetylenic cyclophanes as fullerene precursors: Formation of C60H6 and C60 by laser desorption mass spectrometry of C60H6(CO)12

Article abstract of DOI:10.1002/(SICI)1521-3773(19980518)37:9<1226::AID-ANIE1226>3.0.CO;2-H

A logical precursor of macrocycle C₆₀H₆, cyclophane C₆₀H₆(CO)₁₂ (1) represents a building block in a possible total synthesis of C₆₀. In Fourier transform ion cyclotron resonance laser deso

Full text of DOI:10.1002/(SICI)1521-3773(19980518)37:9<1226::AID-ANIE1226>3.0.CO;2-H

COMMUNICATIONS
Barbara, CA) along with an external pulse/function generator (Model HP
8111 A). Tips were electrochemically etched from Pt/Ir wire (80%/20%,
diameter 0.2 mm) in an aqueous solution of 2n KOH and 6n NaCN.
Typically, a tunneling current of 0.5 ± 1 nA and a bias voltage of 0.2 ± 1 V
(sample negative) were employed. The STM images were corrected for
instrumental drift.
Acetylenic Cyclophanes as Fullerene
Precursors: Formation of C60H6 and C60 by
Laser Desorption Mass Spectrometry of
C60H6(CO)12**
Yves Rubin,* Timothy C. Parker, Salvador J. Pastor,
Satish Jalisatgi, Christophe Boulle, and
Charles L. Wilkins*
Received: August 15, 1997 [Z10823IE]
German version: Angew. Chem. 1998, 110, 1281 ± 1284
Keywords: chirality
microscopy
´ monolayers ´ scanning tunneling
In previous reports[1] we proposed that highly unsaturated
macrocyclic cyclophanes such as 1 and 2 may function as
precursors of the fullerene C60 and its endohedral metal
complexes in a process analogous to the coalescence anneal-
ing of mono- and polycycles with sp-hybridized carbon atoms
during the gas-phase formation of fullerenes from evaporated
graphite (Scheme 1).[2, 3] Formation of endohedral transition
metal complexes by this route is particularly appealing in view
of the exceptional physical and chemical properties of C60.[4]
Here we report for the first time that the acetylenic macro-
cycle 3 effectively leads to C60H6 and C60 ions in the gas phase
in laser desorption mass spectroscopic experiments.[5]
[1] L. Pasteur, C. R. Seances Acad. Sci. 1848, 26, 535 ± 539.
[2] G. Binnig, H. Rohrer, C. Gerber, E. Weibel, Phys. Rev. Lett. 1982, 49,
57 ± 61.
[3] G. Binnig, C. F. Quate, C. Gerber, Phys. Rev. Lett. 1986, 56, 930 ± 933.
[4] R. Viswanathan, J. A. Zasadzinski, D. K. Schwarz, Nature 1994, 368,
440 ± 443.
[5] D. P. E. Smith, J. Vac. Sci. Technol. B 1991, 9, 1119 ± 1125.
[6] S. J. Sowerby, W. M. Heckl, G. B. Petersen, J. Mol. Evol. 1996, 43,
419 ± 424.
[7] J. P. Rabe, S. Buchholz, Phys. Rev. Lett. 1991, 66, 2096 ± 2099.
[8] C. J. Eckhardt, N. M. Peachy, D. R. Swanson, J. M. Takacs, M. A.
Khan, X. Gong, J.-H. Kim, J. Wang, R. A. Uphaus, Nature 1993, 362,
614 ± 616.
[9] a) F. Stevens, D. J. Dyer, D. M. Walba, Angew. Chem. 1996, 108, 955 ±
957; Angew. Chem. Int. Ed. Engl. 1996, 35, 900 ± 901; b) D. M. Walba,
F. Stevens, N. A. Clark, D. C. Parks, Acc. Chem. Res. 1996, 29, 591 ±
597.
[10] a) P. Nassoy, M. Goldman, O. Bouloussa, F. Rondelez, Phys. Rev. Lett.
1995, 75, 457 ± 460. b) I. Weissbuch, M. Berfeld, W. Bouman, K. Kjaer,
J. Als-Nielsen, M. Lahav, L. Leisorowitz, J. Am. Chem. Soc. 1997, 119,
933 ± 942.
[11] P. Vanoppen, P. C. M. Grim, M. Rücker, S. De Feyter, G. Moessner, S.
Valiyaveettil, K. Müllen, F. C. De Schryver, J. Phys. Chem. 1996, 100,
19636 ± 19641.
Á
[12] P. C. M. Grim, S. De Feyter, A. Gesquiere, P. Vanoppen, M. Rücker, S.
Valiyaveettil, G. Moessner, K. Müllen, F. C. De Schryver, Angew.
Chem. 1997, 109, 2713 ± 2715; Angew. Chem. Int. Ed. Engl. 1997, 36,
2601 ± 2603.
[13] Positive is defined as going clockwise from the normal towards the
orientation of the organized structure. The normal can be related to
the graphite axis (q) or the lamellar axis (f; Figure 2).
[14] Our conclusions are based upon approximately 50 STM images for
each enantiomer. None of the STM images obtained are in contra-
diction with the results outlined in the text.
[15] This compound forms two types of monolayer structures. Some
domains are completely disorded and contain ringlike structures,
individual small lamellae, etc. Other domains are composed of
lamellae, but the lamellar width (20 Æ 1 Š) and the intralamellar
distance between two head groups (8.4 Æ 0.2 Š) are much smaller than
the values obtained for ISA molecules. The angle between the alkoxy
chains and the lamellar axis measures approximately 43 Æ 28. No
codeposition of solvent molecules was observed. These large devia-
tions are most likely due to the presence of an ether functionality in
the alkyl chain and the absence of a stereogenic center. In ISA these
stereogenic centers force, probably because of sterical hindrance, the
side chains to adapt an extended conformation that is almost
perpendicular to the lamellar axis.
Scheme 1. Structures of the acetylenic macrocycles 1 ± 3 and of Ih-C60
.
When the thermochemistry of macrocycle 2 was studied by
matrix assisted laser desorption ionization Fourier transform
(MALDI-FT) and atmospheric pressure chemical ionization
[*] Prof. Y. Rubin, Dr. T. C. Parker, Dr. S. Jalisatgi, Dr. C. Boulle
Department of Chemistry and Biochemistry
University of California, Los Angeles
Los Angeles, CA 90095-1569 (USA)
Fax: (‡1)310-206-7649
E-mail: rubin@chem.ucla.edu
[16] K. Freudenberg, W. Lwowski, Justus Liebigs Ann. Chem. 1955, 594,
76 ± 88.
[17] M. Hjalmarsson, H.-E. Högberg, Acta Chem. Scand. Ser. B 1985, 39,
793 ± 796.
Prof. C. L. Wilkins, S. J. Pastor
Department of Chemistry
University of California, Riverside
Riverside, CA 92521 (USA)
Fax: (‡1)909-787-4713
E-mail: cwilkins@citrus.ucr.edu
[**] We are grateful to the Office of Naval Research (N00014 ± 94 ± 1 ±
0534, Y.R.) and the NIH (grant GM44606, C.L.W.) for financial
support.
1226
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3709-1226 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1998, 37, No. 9

COMMUNICATIONS
(APCI) negative-ion mass spectrometry, it was found that the
parent molecular ion (C60H18) showed little propensity to lose
hydrogen.[1a] This seemed somewhat surprising in view of the
facile Bergman cyclization of enediynes to generate 1,4-
benzene diradicals.[1b] Since effective leaving groups should
facilitate the formation and consecutive rearrangement of the
more highly unsaturated precursor C60H6 (1), we decided to
embark on the synthesis of the cyclobutenedione 3. The
cyclobutenedione moiety was previously found to release
acetylenic bonds very efficiently in the gas phase with
concomitant loss of carbon monoxide, for example in the
generation of the carbocycles C18, C24, and C30 in Fourier
transform ion cyclotron resonance laser desorption (FT-ICR
LD) mass spectroscopic experiments or in the preparative-
scale synthesis of triynes and higher polyynes by flash vacuum
pyrolysis.[2a±c, 6]
9a under a number of oxidative coupling conditions; the
attempts always resulted in rapid polymerizations. It was
suggested that the homoconjugated nature of the spiroacetal
moieties in 8a (or 9a) confers unusual reactivity upon this
compound.[8] This was indeed the case: Cyclization of the
dimethoxyacetal 8b by a modified Hay procedure gave
macrocycle 9b as a stable, waxy, yellow solid in good yields
(Table 1). For the preparation of acetals 8a and 8b, we
Table 1. Selected spectroscopic data for 4b, 5b, 7b, 9b, and 3.
4b: 1H NMR (400 MHz, CDCl3): d ˆ 0.23 (s, 9H), 1.06 (m, 21H), 3.92 (s,
3H), 4.15 (s, 3H); 13C NMR (100 MHz, CDCl3): d ˆ 1.18, 11.12, 18.48,
58.40, 59.57, 79.56, 91.23, 102.37, 135.07, 165.46, 180.09; IR (neat): nÄ ˆ 2945,
2156, 1784, 1650 cm 1; EI-HR-MS: calcd for C20H36O4Si2: 396.2152, found:
396.2147
5b: 1H NMR (200 MHz, CDCl3): d ˆ 1.08 (m, 21H), 3.52 (s, 6H), 4.38 (s,
3H); 13C NMR (50 MHz, CDCl3): d ˆ 11.07, 18.42, 53.15, 61.01, 91.86,
94.69, 99.39, 112.24, 184.07, 186.15; IR (neat): nÄ ˆ 2945, 2148, 1790,
1614 cm 1; EI-HR-MS: calcd for C18H30O4Si: 338.1913, found: 338.1915
The preparation of cyclophane 3 proved to be much more
challenging than expected. The protected ethylene spiroacetal
7a was first prepared in analogy to the synthesis of the
ethylene acetal precursors of the carbo[n]cycles (Scheme 2).[7]
However, the deprotected alkyne 8a could not be cyclized to
7b: 1H NMR (400 MHz, CDCl3): d ˆ 1.09 (m, 63H), 3.52 (s, 18H) 3.53 (s,
18H), 7.59 (s, 3H); 13C NMR (100 MHz, CDCl3): d ˆ 11.10, 18.50, 52.00,
52.05, 73.45, 75.41, 83.68, 85.52, 97.86, 108.15, 108.44, 109.96, 122.77, 133.42,
136.50, 138.88; IR (KBr): nÄ ˆ 2943, 2207, 2130, 1600, 1576, 1463 cm 1; UV/
Vis (CH2Cl2): lmax (e) ˆ 235 (79200), 269 (74400), 291 (71200), 308 (70300),
318 sh (74800), 326 sh (85200), 341 (107000), 368 nm 1 (96400); FAB-HR-
‡
MS: calcd for C75H102O12Si3 (M ): 1278.6679, found: 1278.6646; elemental
analysis calcd for C75H102O12Si3: C 70.38, H 8.03; found: C 70.41, H 8.03
9b: 1H NMR (400 MHz, CDCl3): d ˆ 3.48 (s, 36H), 3.50 (s, 36H), 7.55 (s,
6H); 13C NMR (100 MHz, CDCl3): d ˆ 52.08, 52.09, 73.12, 75.08, 78.19,
85.22, 87.07, 87.97, 108.52, 108.54, 122.43, 136.26, 136.61, 138.00; IR (KBr):
nÄ ˆ 2942, 2206, 1576, 1463, 1085 cm 1; UV/Vis (CH2Cl2): lmax (e) ˆ 280
1
(197000), 300 sh (112000), 322 (80900), 341 (84700), 382 (70500), 414 nm
‡
(48400); FAB-HR-MS: calcd for C96H78O24Na (M ‡ Na ): 1637.4781,
found: 1637.4724
3: 1H NMR (400 MHz, CD2Cl2): d ˆ 7.70 (s); 13C NMR (100 MHz,
[D8]THF): d ˆ 79.1, 79.4, 80.7, 97.1, 102.8, 108.1, 122.8, 137.8, 175.3, 178.4,
192.0, 192.5; IR (KBr): nÄ ˆ 1784, 2136, 2198 cm 1; UV/Vis (CH2Cl2): lmax
(rel. intensity) ˆ 245 (100), 282 (98), 316 sh (63), 347 (59), 407 sh (56),
1
442 nm sh (37)
followed a disconnection similar to the synthesis of macro-
cycle 2.[1a] This required the preparation of the half-protected
enediynes 6a and 6b in gram quantities.[6] Liebeskindꢁs
methodology was used most effectively to form siloxy ketones
4a and 4b, which were converted into monoacetals 5a and 5b
in a one-pot reaction.[9] Addition of a second alkyne moiety to
5a or 5b and conversion into the stable cyclobutenedione
bis(acetal)s 6a and 6b was achieved by in situ treatment of the
initially formed lithium alkoxides with concentrated H2SO4 or
3m HCl followed by a second acetalization step and selective
removal of the trimethylsilyl protecting groups. Both 6a and
6b were converted into their yellow copper(i) acetylides and
coupled with 1,3,5-tris(bromoethynyl)benzene[1a] in pyridine
or THF to give the stable cyclization precursors 7a and 7b.
These precursors were deprotected just prior to cyclization to
the unstable terminal alkynes 8a or 8b by treatment with
tetrabutylammonium fluoride and subsequent filtration
through silica gel (Et2O).
ꢀ
Scheme 2. Synthesis of the protected macrocycle 9b. a) TIPSC CLi, THF,
788C; b) TMSCl, 78 to 258C; c) TMSOCH2CH2OTMS or TMSOMe,
ꢀ
cat. TMSOTf, THF, 55 or 258C; d) TMSC CLi, Et2O, 78 to 458C;
e) concd H2SO4, 258C, or 3m HCl, 45 to 258C; f) K2CO3, MeOH;
g) LiHMDS, THF, then CuBr; h) 1,3,5-tris(bromoethynyl)benzene, pyri-
dine or THF, 258C; i) TBAF, THF; j) CuCl ´ TMEDA (3 equiv), ODCB,
258C; k) CF3COOH, 258C, 4 ± 6 h. All OCH2CH2O-bridged acetals are
spiroannulated. TMS ˆ SiMe3; TIPS ˆ SiiPr3; Tf ˆ CF3SO2; LiHMDS ˆ
lithium hexamethyldisilazide; TBAF ˆ tetra-n-butylammonium fluoride;
TMEDA ˆ N,N,N',N'-tetramethylethylenedimamine; ODCB ˆ o-dichloro-
benzene.
Deprotection of the methoxyacetal moieties of macrocycle
9b necessitated some fine-tuning. Surprisingly, we found that
the cyclobutenedione macrocycle 3 was considerably less
stable than the carbon oxide precursors of C18, C24, and
Angew. Chem. Int. Ed. 1998, 37, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3709-1227 $ 17.50+.50/0
1227

COMMUNICATIONS
C30;[2a, 7] it could be characterized only with difficulty. Macro-
cycle 3 was prepared by dissolution of 9b in trifluoroacetic
acid. After three to six hours an orange-yellow precipitate had
formed, which was redissolved in CH2Cl2 after evaporation of
the acid. In CH2Cl2 3 decomposed to a brown precipitate
within one to two hours. In [D8]THF a 13C NMR spectrum
could be obtained; it shows 12 lines in the expected range
(Table 1).
implied by its fragmentation to C60H6(CO)4, C60H6(CO)3,
C60H6(CO)2, and C60H6(CO) in the negative-ion spectrum
(calcd: 838.0272, 810.0322, 782.0373, 754.0424; found: m/z
(deviation) ˆ 838.0096 (21 ppm), 810.0231 (11 ppm), 782.0327
(6 ppm), 754.0460 (5 ppm)). The fullerenic nature of the C60
ions can be unambiguously inferred from the characteristic
loss of C2 units to give peaks at m/z ˆ 696 (C58) and 672
(C56).[11] At higher laser powers in both negative and positive
modes, the degree of C2 loss increases greatly. On the other
We were very pleased to find that macrocycle 3, despite its
instability, undergoes clean loss of its carbonyl groups in FT-
ICR LD-MS experiments (Figure 1). Fragment ions are
.
hand, fragmentation with C2 loss is not observed from C60H6 ,
indicating that its structure is most likely non-fullerenic in
nature and possibly that of 1.
The formation of fullerene ions (e.g. C60, C70) from the
precursors of the carbocycles C18, C24, and C30 was an
important observation.[2a,b] Interestingly, this conversion was
only found in the positive-ion mode. Cations of carbon
clusters are generally highly excited thermally in LD-MS
experiments, which tends to induce intermolecular coales-
cence and fragmentation reactions.[10, 11] In this study, we find
concurrently that positive ions give rise to increased frag-
mentation and coalescence. In contrast, this is the first time
that rearrangement of a precursor such as 1 to C60 is found to
occur in the negative-ion mode under seemingly milder
conditions. It is likely that the three-dimensional framework
of cyclophane 1 lends itself ideally to fullerene isomerization.
On the other hand, the negative ions of the carbocycles C18,
C24, and C30 do not react with neutral species because of their
excess charge.[2a,b, 12] In addition, the three-dimensional shape
of precursors such as 1 may display better suitability for metal
complexation in the potential formation of endohedral
metallofullerenes.
Preliminary attemps at preparative isolations showed that
the precipitates obtained from the solutions of 3 in trifluoro-
acetic acid or from pulsed femtosecond laser irradiation at
400 nm of a dilute solution of 3 in THF did not contain
buckminsterfullerene or other apolar material. Further ex-
periments aimed at the incorporation of metals into C60 and
preparative-scale isolations (e.g. by flash vacuum pyrolysis)
are in progress.
Figure 1. Negative-ion ICR-LD mass spectrum of macrocycle 3 after
desorption at 10.6 mm.
observed with both CO2 (10.6 mm) and N2 lasers (337 nm), but
much more intense ion signals are obtained with the former.
In the negative-ion mode, prominent ions corresponding to
.
.
C60H6 and C60 appear at m/z ˆ 726 and 720 together with
their isotopic clusters, whose intensities match superbly with
the theoretical ratios (Figure 1, inset). The resolution of the
ICR-LD mass spectrum is excellent and reaches values of up
to 7500. With authentic C60/C70 as external standards (for
calibration after LD-MS measurements of 3), the observed
Experimental Section
ꢀ
4b: To iPr3SiC CH (4.53 g, 24.8 mmol) in THF (25 mL) was added 2.5m
.
.
13CH6
12
12
12
nBuLi (9.90 mL, 24.8 mmol) dropwise at 08C. The solution was stirred for
30 min and then added to dimethyl squarate (3.35 g, 23.6 mmol) in THF
(100 mL) at 78 oC dropwise with a syringe. After 1 h at 78 oC TMSCl
(3.26 g, 30.0 mmol) was added dropwise, and the reaction was allowed to
warm to 258C. Aqueous workup (NaHCO3) and removal of the solvent in
vacuo gave 4b (9.36 g, 99%) as a pale yellow oil.
masses for
C
(719.9951), C60H6 (726.0380),
C
60
59
.
12
(727.0416), and
C
13C2H6 (728.0599) are within 8, 13, 13,
58
and 8 ppm of their theoretical values (720.0006, 726.0475,
727.0509, and 728.0542). In the positive-ion mode (spectrum
.
‡
not shown), no fragment ion corresponding to C60H6 is
.
‡
5b: To 4b (9.36 g, 23.6 mmol) in THF (10 mL) was added MeOTMS
(2.60 g, 24.8 mmol) followed by TMSOTf (0.30 mL, 1.5 mmol). The
reaction was initially exothermic, and the mixture was stirred for 30 min
at 258C. Aqueous workup (NaHCO3) and removal of solvent in vacuo gave
a brown oil, which was purified by filtration on silica gel (Et2O/hexanes 1/2)
to give 5b (7.99 g, 99%) as a yellow oil.
observed; only the peak for C60 and less intense peaks for the
higher fullerenes C62 ± C70 appear. In one instance C2n clusters
centered around multiples of C60 (C120, C180) were also
observed; they are attributed to oligomerization of the
unstable precursor during laser desorption or to coalescence
.
‡
of C60 or radical cationic intermediates with neutral species in
the gas phase.[10] In the negative-ion spectrum, fragment ions
for C62 ± C70 are also found, but with lower intensities
(Figure 1).
ꢀ
6b: 1) To Me3SiC CH (0.66 mL, 0.46 g, 4.7 mmol) in Et2O (10 mL) at 08C
was added dropwise 2.5m nBuLi (1.9 mL, 4.7 mmol). After the reaction
mixture had been stirred for 30 min the temperature was lowered to 78
oC. Compound 5b (1.50 g, 4.43 mmol) in Et2O (20 mL) was added
dropwise. The reaction mixture was stirred for 1 h at 458C. To the
resulting red solution was added 3m HCl (20 mL), and the reaction mixture
was stirred at 258C for 1 h. Aqueous workup (NaHCO3) and removal of the
The parent ion of 3 is not observed in either the positive-
or negative-ion mass spectra, but its involvement is clearly
1228
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3709-1228 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1998, 37, No. 9

COMMUNICATIONS
solvent in vacuo gave the desired enediyne (1.67 g, 93%) as a brown oil,
which was used directly in the next step. 2) To the enediyne (1.67 g,
4.13 mmol) in MeOTMS (2.0 mL, 1.5 g, 15 mmol) was added TMSOTf
(0.10 mL, 0.12 g, 0.52 mmol). After the reaction mixture had been stirred
for 48 h at 258C, aqueous workup (NaHCO3), removal of the solvent in
vacuo, and filtration through silica gel (Et2O/hexanes 1/1) gave the desired
bis(acetal) (1.38 g, 74%) as a yellow oil. 3) To the bis(acetal) (7.34 g,
16.3 mmol) in MeOH (50 mL) was added K2CO3 ( ꢁ 100 mg). After the
reaction mixture had been stirred for 15 min at 258C, aqueous workup,
removal of the solvent in vacuo, and filtration through silica gel (Et2O/
hexanes 1/1) gave 6b (5.80 g, 94%) as a yellow oil. It was diluted with THF
(100 mL) and used directly in the next step.
[6] Y. Rubin, S. S. Lin, C. B. Knobler, J. Anthony, A. M. Boldi, F.
Diederich, J. Am. Chem. Soc. 1991, 113, 6943 ± 6949.
[7] Y. Rubin, C. B. Knobler, F. Diederich, J. Am. Chem. Soc. 1990, 112,
1607 ± 1617.
[8] The conformationally locked lone pairs on both O atoms of the
cyclopentadienone ethylene acetal create a similar situation. This is
manifested by the dimerization rate, which is 500000 times faster over
that of cyclopentadiene and 1070 times faster than that of the
dimethoxyacetal: P. E. Eaton, R. A. Hudson, J. Am. Chem. Soc. 1965,
57, 2769 ± 2771. We are indebted to Professor Philip Eaton at the
University of Chicago for his insightful suggestion.
[9] a) L. S. Liebeskind, K. R. Wirtz, J. Org. Chem. 1990, 55, 5350 ± 5358;
b) L. S. Liebeskind, R. W. Fengl, K. R. Wirtz, T. T. Shawe, ibid. 1988,
53, 2482 ± 2488.
[10] a) C. Yeretzian, K. Hansen, F. Diederich, R. L. Whetten, Nature 1992,
359, 44 ± 47; b) R. D. Beck, P. Weis, G. Bräuchle, M. M. Kappes, J.
Chem. Phys. 1994, 100, 262 ± 270; c) Z. X. Xie, Z. Y. Liu, C. R. Wang,
R. B. Huang, F. C. Lin, L. S. Zheng, J. Chem. Soc. Faraday Trans. 1995,
91, 987 ± 990.
[11] a) Q. L. Zhang, S. C. OꢁBrien, J. R. Heath, Y. Liu, R. F. Curl, H. Kroto,
R. E. Smalley, J. Phys. Chem. 1986, 90, 525 ± 528; b) P. Wurz, K. R.
Lykke, ibid. 1992, 96, 10129 ± 10139; c) G. Ulmer, E. E. B. Campbell,
R. Kühnle, H. G. Busmann, I. V. Hertel, Chem. Phys. Lett. 1991, 182,
114 ± 119.
7b: To 6b (6.28 g, 16.6 mmol) in THF (100 mL) at 788C was added
LiHMDS, which was prepared from (Me3Si)2NH (3.50 mL, 2.68 g,
16.6 mmol) and 2.5m nBuLi (6.60 mL, 16.6 mmol) in THF (15 mL). After
30 min CuBr (2.38 g, 16.6 mmol) was added, and the mixture warmed to
08C. To this solution was added 1,3,5-tris(bromoethynyl)benzene[1a] (2.03 g,
5.25 mmol). After the reaction mixture had been stirred for 12 h at 258C,
aqueous workup (NaHCO3) and removal of the solvent in vacuo gave a
brown oil, which was purified by chromatography on silica gel (hexanes to
hexanes/Et2O 3/1) to afford 7b (3.25 g, 48%) as a colorless foam.
8b: To 7b (500 mg, 0.391 mmol) in THF (10 mL) was added five drops of
H2O followed by 1.0m TBAF (2.0 mL, 2.0 mmol). The reaction mixture was
stirred for 2 h, diluted with hexanes (50 mL), and filtered through silica gel
(Et2O). The resulting yellow fractions were concentrated to 5 mL and
diluted with ODCB (400 mL) for the immediately following cyclization.
[12] A possible mechanism for the rearrangement of C60H6 (1) to C60 is
available from the authors upon request.
9b: To the solution of 8b obtained above was added CuCl (119 mg,
1.20 mmol) followed by TMEDA (1.0 mL, 0.77 g, 6.6 mmol). After the
reaction mixture had been stirred for 1 h under air, the solution was poured
on top of a pad of silica gel and eluted with CHCl3 to remove most of the
ODCB. Elution with Et2O/CHCl3 (1/2) provided a yellow fraction with the
desired product. The solvent was removed, and the brown oil purified by
chromatography on silica gel (Et2O/CHCl3 5/95). Removal of the solvent in
vacuo gave cyclophane 9b (150 mg, 48%) as a yellow foam.
A Dendritic Macrocyclic Organic Polyradical
with a Very High Spin of S ˆ 10**
Andrzej Rajca,* Jirawat Wongsriratanakul,
Suchada Rajca, and Ronald Cerny
3: Aliquots of 9b (5 - 30 mg) were dissolved in trifluoroacetic acid (1 ±
3 mL). Over 3 ± 6 h a precipitate began to form (the reaction could be
monitored by 1H NMR spectroscopy in CF3CO2D). After 6 ± 12 h the
trifluoroacetic acid was removed in vacuo at 258C in the absence of light.
[D8]THF ( ꢁ 0.5 mL) was immediately added to the remaining orange-red
solid under argon. This solution was used for 13C NMR spectroscopy. For
the LD-MS experiments, anhydrous CH2Cl2 or THF were added to dissolve
the precipitate, and the solution was used immediately.
Organic molecules with a very high spin possess a large
number of ferromagnetically coupled unpaired electrons.[1]
The design and synthesis of such molecules must overcome
the challenging problem of maintaining strong through-bond
interactions between multiple sites within the molecule.[1d, 2]
Highly efficient generation of unpaired electrons and/or
multiple pathways for ferromagnetic coupling provided
organic molecules with S greater than 5.[3±6] The highest spin
for an organic molecule to date, a nonacarbene with S ˆ 9, was
reported by Iwamura and co-workers in 1993.[3]
Received: March 4, 1998 [Z11555IE]
German version: Angew. Chem. 1998, 110, 1353 ± 1356
Keywords: alkynes ´ cage compounds ´ carbon allotropes ´
fullerenes ´ mass spectrometry
Polyradical 1 was designed as an ªorganic spin clusterº.[6, 7]
Because ferromagnetic coupling through 1,3-phenylene is
significantly stronger than that through a 3,4'-biphenylene
unit, unpaired electrons in the four dendritic branches and the
macrocyclic core can effectively be lumped into component
spins (S').[6] Such a ferromagnetically coupled spin pentamer
with S' ˆ 5/2, 5/2, 5/2, 5/2, 2 should have a ground state with
S ˆ 12 (Scheme 1). We describe here the synthesis and
magnetic studies of polyradical 1.
[1] a) Y. Rubin, T. C. Parker, S. I. Khan, C. L. Holliman, S. W. McElvany,
J. Am. Chem. Soc. 1996, 118, 5308 ± 5309; b) Y. Rubin, Chem. Eur. J.
1997, 3, 1009 ± 1016.
[2] a) Y. Rubin, M. Kahr, C. B. Knobler, F. Diederich, C. L. Wilkins, J.
Am. Chem. Soc. 1991, 113, 495 ± 500; b) S. W. McElvany, M. M. Ross,
N. S. Goroff, F. Diederich, Science 1993, 259, 1594 ± 1596; c) F.
Diederich, Y. Rubin, O. L. Chapman, N. S. Goroff, Helv. Chim. Acta
1994, 77, 1441 ± 1457; d) N. S. Goroff, Acc. Chem. Res. 1996, 29, 77 ± 83.
[3] a) G. von Helden, M. T. Hsu, N. Gotts, M. T. Bowers, J. Phys. Chem.
1993, 97, 8182 ± 8192; b) J. M. Hunter, J. L. Fye, M. F. Jarrold, Science
1993, 260, 784 ± 786; c) J. M. Hunter, J. L. Fye, E. J. Roskamp, M. F.
Jarrold, J. Phys. Chem. 1994, 98, 1810 ± 1818; d) K. B. Shelimov, D. E.
Clemmer, M. F. Jarrold, J. Chem. Soc. Dalton Trans. 1996, 567 ± 574;
e) D. L. Strout, G. E. Scuseria, J. Phys. Chem. 1996, 100, 6492 ± 6498.
[4] a) F. T. Edelmann, Angew. Chem. 1995, 107, 1071 ± 1075; Angew.
Chem. Int. Ed. Engl. 1995, 34, 981 ± 985; b) S. Nagase, K. Kobayashi, T.
Akasaka, Bull. Chem. Soc. Jpn. 1996, 69, 2131 ± 2142.
[*] Prof. A. Rajca, Mr. J. Wongsriratanakul, Dr. S. Rajca, Prof. R. Cerny
Department of Chemistry
University of Nebraska
Lincoln, NE 68588 (USA)
Fax: (‡1)402-472-9402
E-mail: arajca@unlinfo.unl.edu
[**] This research was supported by the National Science Foundation
(CHE-9510096). We thank Dr. Richard Schoemaker for the NMR
spectral determinations and Ms. Nissakorn Thongkon for her
assistance with synthesis.
[5] For a similar approach to C60H6, see Y. Tobe, N. Nakagawa, K.
Naemura, T. Wakabayashi, T. Shida, Y. Achiba, J. Am. Chem. Soc., im
Druck.
Angew. Chem. Int. Ed. 1998, 37, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3709-1229 $ 17.50+.50/0
1229