вход по аккаунту


Anions Stabilized by -Nitro Groups The Acidity and ortho-Metalation of NitrocubanesЧPenta- and Hexanitrocubanes.

код для вставкиСкачать
Anions Stabilized by P-Nitro Groups: The Acidity
and ortlzo-Metalation of NitrocubanesPenta- and Hexanitrocubanes**
Kirill Lukin, Jianchang Li, Richard Gilardi, and
Philip E. E a t o n *
Nitro groups are powerfully electron withdrawing. On both
resonance and inductive scales they are more potent than any
other uncharged functionality.". 21 Anions with a-nitro substituents are highly stabilized by resonance delocalization of
charge; an 1-nitronate (the aci-nitro anion) is 10''' times less
basic than a ketone e n ~ l a t e . [The
~ I acidity of carboxylic acids is
increased about one thousandfold by the inductive effect of an
cw-nitro group.[41 We show here that nitro substituents have a
dramatic effect on the acidity and chemistry of cubanes.
Cubane uses an s-rich orbital to bond substituents. Thus,
cubyl hydrogens are more acidic than those'on more usual hydrocarbons; for example, cubane is approximately 60000 times
more acidic than c y c i o h e ~ a n eExperiments
with mono- and
1,4-dinitrocubanes, designed to probe for substituent-enhanced
acidity (refer to the effects of amide substituents[6])were entirely
frustrated by the ease with which base destroys these compounds. Not even the simplest exchange reaction could be
achieved. If one can use transient colorations as an indicator,
single electron transfer mechanisms probably intervened and
somehow unraveled the cubane framework; nothing of substance could be isolated. It is possible that an anion was formed,
but that /%elimination of nitrite then occurred. Although this
happens readily with enolizable /I-nitro ketones and the like,"'
elimination of nitrite is much less likely here, for the product
would be a cubene. Cubenes are anti-Bredt olefins, highly
pyramidalized and difficult to access, but they d o give isolable
1,3,5,7-Tetranitrocubane( I ) , first made here in 1992 and now
much more easily available,['. "1 has four identical hydrogen
atoms each surrounded by three nitro groups (the molecule has
pseudo-tetrahedral symmetry). The possibility of deprotonating
to an anion (2) much stabilized by the inductive electron-withdrawing effects ( - I effect) of three vicinal (B) nitro groups and,
perhaps more importantly, by coordination of the electron-rich
oxygen atoms of these groups with the counterion seemed particularly attractive. Indeed, we have found that the acidity of the
cubane hydrogen atoms is so enhanced in 1 that their exchange
can be effected rapidly even with bases as mild as sodium
methoxide in methanol. Thus, reaction of [H,]-1 with 0.125 M
CD,ONa in CD,OD for 45 minutes at room temperature,
followed by quenching with hydrochloric acid and extraction
with EtOAc, gave [D,]-l in 87 % yield. The reverse experiment,
that is, reaction of [DJ-1 with CH,ONa in CH,OH, returned
[*] Prof. P. E. Eaton. Dr. K. Lukin
Department of Chemistry. The University of Chicago
5735 S . Ellis Avenue. Chicago. IL 60637 (USA)
Fax: Int. code +(312) 702-2053
Dr. J. Li
Geo-Centers Inc.
762 Route 15 South, Lake Hopatcong, NJ 07849 (USA)
D r R. Gilardi
Laboratory for the Structure of Matter
The Naval Research Laboratory. Washington. D.C. 20375 (USA)
We are grateful to Dr. Ravi Shankar and Dr. Gene Wicks far their initial
studies of lower nitrocubanes and to Dr. Eric Punzalan and Dr. A. BashirHashemi for samples of I . This work was supported by The U.S. Army Research. Development and Engineering Center (Geo-Centers) and the Office of
Naval Research.
The anion that must intervene in these exchanges exists at
only very low concentration in methanolic CH,ONa. It could
not be detected in this milieu by NMR spectroscopy, nor did we
have any success trapping it with electrophiles other than acid.
However, 'H NMR examination of the reaction of 1 with sodium bis(trimethylsily1)amide in [DJTHF at -75 "C showed
clearly the formation of the salt Na-2: 'H NMR 6 = 5.54 (s).
This salt is stable for hours at - 75 "C, but it decomposes rapidly
above -50°C.
We could estimate the pK, of 1 from low temperature proton
NMR measurements on equilibrium mixtures with various
bases. Almost complete metalation (salt formation) occurs in
T H F with one equivalent of sodium bis(trimethylsilyl)amide,
but none with cw-lithiobenzyl cyanide. As the pK, in T H F of
bis(trimethylsilyl)amine is 25.8["] and that of benzyl cyanide is
18.7.['21 the pK, of 1 must be between about 20 and 24.5. We
have narrowed this to between 20.5 and 22.5 by noting that I is
partially converted into its anion by 1.5 equivalents of potassium tert-butoxide (the pK, of tert-butyl alcohol in T H F is given
as 21 .6[131).
The salt of tetranitrocubane can be used to achieve further
substitution on the cubane framework. For example, reaction of
Na-2 in T H F at - 75 "C with carbon dioxide followed by neutralization gave the tetranitrocubane carboxylic acid 3 in 85 YO
yield (characterized as its methyl ester: 'H NMR: 6 = 3.84 (s,
3H), 6.22 (s, 3H); 13C NMR: 6 = 54.3 (OCH,), 67.0 (3C, CH),
72.0, 73.4, 75.2 (3C, CNO,), 159.9 (C=O)). Quenching 2 with
chlorotrimethylsilane gave the tetranitro(trimethylsily1)cubane
4 in 80% yield ('H NMR: 6 = 0.16 (s, 9H), 6.07 (s, 3H);
I 3 C N M R : 6 = - 3.3 (3C, SiCH,), 67.8 (3C, CH), 71.7, 74.4
(3C, CNO,), 75.5). These conversions are the first to our
knowledge of substitutions by way of "ortho-metalations" in
which the nitro group activates a ,8 C-H bond and stabilizes the
When a larger excess of sodium bis(trimethylsily1)amide was
used in these reactions, yet more highly substituted cubanes
were obtained. Treatment of 1 with five equivalents of base
followed by quenching with carbon dioxide and neutralization
gave a 70: 30 mixture of 3 with 5, the tetranitro diacid (charac-
terized as its dimethyl ester: ' H N M R : 6 = 3.90 (s, 6 H ) , 6.34 (s.
2 H ) ; 13C N M R : 6 = 54.5 (2C, OCH,), 66.3 (2C, CH), 72.5
(2C). 74.4 (2C). 76.5 (2C), 159.3 (2C, C=O)). Similarly, the
use of three equivalents of base followed by reaction with
chlorotrimethylsilane gave a 20:80 mixture of 4 with 6, the
bissilylated product ('H N M R : 6 = 0.19 (s, 18H). 6.10 (s, 2H);
" C N M R : 6 = - 3.2 (6C, CH,), 67.1 (2C. CH), 72.7 (2C),
75.4 (2C). 75.8 (2C)). We are reasonably certain that 5 and 6
result from sequential anion formations from 3 and 4.There is
no evidence from the low temperature N M R study for dianion
formation from 1.
As anions derived from 1 are trapped effectively with simple
electrophiles. we hoped that reaction with electrophilic nitrating
reagents would allow us access to more highly nitrated cubanes.
Such cubanes are predicted to be powerful, shock-insensitive,
high-density explosives." Until now the most highly nitrated
cubane known was 1, obtained from the corresponding tetraacid by oxidation of the derived amine, an approach not applicable to cubanes with vicinal nitro groups.",
As best we can determine, there are no firm examples of
successful direct nitration of localized (non-resonance-stabilized) organometallic compounds of alkali metals reported."'. "] In line with this, we were quite unsuccessful in nitrating salts of 1 with NO,BF,, acetyl nitrate, tetranitromethane,
amyl nitrate, or similar reagents.['*] However. condensation of
excess N 2 0 , into a frozen solution of anion 2 in T H F at
--I96 C followed by warming produced a 60:40 mixture of (7) and 1. This strongly suggests that
N,O, effects oxidation of 2 to the radical, which then either
abstracts a hydrogen atom from the solvent (THF) giving 1 or
reacts with NZO,to give 7 (Scheme 1).
The detailed structure of 7 was obtained b j single-crystal
X-ray analysis.["1 Comparison to that of tetranitrocubane I
showed that the fifth nitro group in 7, even though surrounded
by three others, causes no significant disturbance in molecular
geometry---bond angles and lengths are essentially unperturbed.
Molecules of 7 close-pack in the crystal; each oxygen has nine
non-bonded 0 .. .O contacts shorter than 3.0 8, of which four
are between 2.75 and 2.92 A. This close packing leads to a density of 1.96 gcm-3 at 21 "C, a value exceeded by few other
C H N O compounds.
Our methodology for nitration of I by ortho-metalation can
be extended to nitration of pentanitrocubane 7 via its salt Na-8.
Thus, formation of anion 8 from 7 with sodium bis(trimethy1sily1)amide followed by nitration with N,O, gave a 30:70 mix(9) and pentanitrocubane 7,
ture of 1,2,3,4,5,7-hexanitrocubane
presumably via a radical intermediate (cf. Scheme 1).
The separation of these two polynitrocubanes is very difficult.
Nonetheless, we obtained enough of 9 to identify it unambiguously: 'H N M R ([DJacetone): 6 = 6.73 (s); 13C N M R :
6 = 65.0 (C6,8-H), 78.9 (C1. ,-NO,), 83.1 (C3, ,-NO,). 91 .0
(C2.,-NO2). The assignment was confirmed fully by X-ray analysis of the crystalline bis-acetonitrile solvate of 9.fz01
Received: October 12. 1995
Revised version: January 25. 1996 [Z 8469;8470 I€]
German version: Aiigric.. Chivn. 1996, 108, 938-940
Keywords: carbanions . cubanes . ortho-metalation . nitro compounds
C. Gardner. E. C. Lupton. J Ani. C h i v . Soc. 1968. 90. 4328.
R. W. Taft, I. S. Lewis. J A m . Chem. Sor. 1958. 80. 2436.
R. G . Pearson. R. L. Dillon. J Am. Chem. Soc 1953. 75. 2439.
H. L. Finkleiner. M. Stiles. J. Am Chon. Soc. 1963, 85, 616.
[S] R. Dixon. A. Streitwieser, P. G. Williams. P E. Eaton, J. .4n7. Chem Snc 1991.
113. 357.
[6] Review: P. E. Eaton. Angeu-. Chen?. 1992. 104. 1447: A i q e i v . Chrni. l i i r . Ed.
Engl. 1992, 31, 1421
[7] N. Ono in Nitro Conipoccnds. Rewnr Adrunrr.s in Svnr/ri.~urind Chrmi.srrv
(Eds.: H. Feuef. A. T. Nielsen). VCH Publishers. New York, 1980, p. 86.
[8] P. E. Eaton. M. Maggin!. J Am. Chem. Soc. 1988. 110, 7230.
[9] P. E. Eaton, Y. Xiong. R. Gilardi, J Am. Chem So<. 1993. 115. 10195.
[lo] For recent improvements in the synthesis of 1. see a ) A. Bdshir-Hashemi.
Angrir. Chem. 1993. 105. 5x5; Angeir. Chiwi. Int. Ed. E q / . 1993. 32. 612:
b) R. L. Hertzler, P. E. Eaton, Abstr. Pup. Am. Chfin. Soi. 1993, paper 12X.
( 1 11 R. R. Frdser. T. S. Mansour. S. Savard. J Org. Chrm. 1985. 50. 3232
(121 R. R. Frdser. T. S. Mansour. S. Savard. Can. J Chein. 1985. (13, 3505.
(131 C. A. Brown. J. C h ~ mSoc.
Chiwf. Coinmiin. 1974. 680.
1141 Even the anion of 1,3,5-trinitrobenzene has not been obtained by direct proton
abstraction. It and other o-nitrophenyl anions have been made by lithium-forbromine exchanges below - 100 'C. Such anions have been trapped by carboxylation. Fordetails, see: G. Kobrich, P. Buck, C/wm Brr. 1970. 103. 1412;
P. Buck, G. Kobrich. hid. 1970, 103, 1420.
Scheme I
Re:iction of N,O,with Nd-2
Pentanitrocubane 7, the first cubane containing vicinal nitro
groups. is colorless and highly crystalline ('H N M R
([D,]acetone): 6 = 6.43 (s); 1 3 C N M R : 6 = 66.5 (C4.6,8-H),
72.8 (C,-NO,), 78.8 (Cl, 3. ,-NO), 93.0 (C,-NO,)). Differential
scanning calorimetry (20 "C min- ') indicated a stability similar
to earlier p o l y n i t r o c ~ b a n e s , ~
setting in above
A i i g i ~ !C'lfiw.
Ed EngI. 1996. 35, No. 8
VCH Verlug.sgese/l.schafrmhH. 0-69451 Wi~inheIin.1996
O570-0833!96135ON-OH65 S 15 00 + 2 5 0
[15] a) E. E. Gilbert. U.S. Army Research Development and Engineering Center,
Pic;itiiiny, NJ. privatecommunication. 1979, b) J. Alster. 0.Sandus, R. Genter,
N.Slagg. J. P. Ritchie, M. J. S. Dewar. paper at the Working Group Meeting
o n High-Energy Molecules, 1981: c ) A. P. Marchand. Terrtiherlron 1988. 44,
[16] J. Thiele. Ber. Dtsdi. Chenr. Ges. 1907. 33, 666.
[I 71 a) H . Feuer in The Cheniistr~of Anrino, Nrtroso und Nitro Compoiod.t ond Theb.
Deriwrnes, S~ipphiientF(Ed.: S. Patai), Wiley. New York, 1982. p. 805: b) for
an interesting special example. see M . E. Sitzmann. L. A. Kaplan, 1. Angers, J.
Org. Chern. 1977. 42, 563.
[18] Similarly, poor results were obtained with nitrosylating reagents.
1191 Crystal data of 7: monoclinic (P2,;c). u = 6.637(3). b = 23.275(14), c =
7.860(5) A,p =113.21(5) , V =115.8(11) A 3 , Z = 4,p,,,,, =1.959gcrW3.The
thin-plate crystal was small (0.01 x 0.35 x 0.40 mm), leading to a somewhat
weak set of diffraction intensities. and thus low precision in the numerical
results. Of 1542 collected reflections [Z(Cu,,) = 1.54178 .&. T = 294(2) K, H/20
scan mode, 2ernAx
= 96.1, 1051 were unique, and 684 with I > 20(1) were used
for refinement. Lorentz. polarization, and absorption corrections (integration,
T,,,= 0.98. T,,, = 0.72) were applied to data. The structure was solved and
refined using SHELX programs (G. M. Sheldrrck. SHELXSR6 and
SHELXL93. University of Gottingen. Germany). Full-matrix least squares on
F Z refinement varied 218 parameters. atom coordinates and anisotropic thermal parameters for all non-H atoms and isotropic thermal parameters for
hydrogen atoms. Final R = 0.0668, wR2 = 0 1647 with final difference Fourier
excursions of0.294 and - 0 . 3 0 2 e k ' [20b].
[20] a) Crystal data of 9 . 2MeCN. monoclinic, C2:c. u = 13.0102(7). h =
7.7242(6), c = 18.0758(12) A. 6 = 95.404(7)-. V =1808.4(2) A3. Z = 4.
= 1.676 gcm". The asymmetric unit comprises half o f 9 and one acetonitrile molecule; a crystallographic twofold axis passes through the centers of the
di- and tetra-substituted faces ofthecube. Crystal size 0.36 x 0.6x 0.75 mm. Of
1567collected reflections(L(Cu,,) = 1.54178 A. T = 2?3(2) K.0.211scan mode.
20,,, =115,'), 1241 were unique ((> 20(1)). and were used for refinement
Lorentz. polarization. and absorption corrections (integration.
= 0.65,
T,,, = 0.47) were applied to data. The structure was solved and refined as
described above. Final R = 0.0407. wR2 = 0.1062 with final difference Fourier
excursions of 0.278 and -0.186 e k ' . The indicated precision in the numerical results is about four times higher than that for 7; the appearance of bonding-electron density peaks in the final electron density map indicated aboveaverage diffraction quality. b) Crystallographic data (excluding structure factors) for the structures reported in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplementary publication
no. CCDC-179.6. Copies of the data can be obtained free of charge on application to The Director, CCDC. I 2 Union Road, Cambridge CB2 lEZ, U K (fax:
Int. code +(1223) 336-033, e-mail:
less, as the pKa of cyclopentadiene is not unlike that of methanol, and we already knew that methoxide in [DJmethanol was
basic enough to effect D-for-H exchange on 1, it seemed possible
that an aminostannane could metalate 1. Indeed, reaction of 1
with excess (diethy1amino)trirnethyl~tannanef~~
proved to be
very smooth and easily followed by ' H NMR ([DJTHF, 20 " C ) .
After 15 min, more than 50% of 1 was converted into the monotin derivative (6 = 5.94 (s)). Another singlet (S = 5.99) corresponding to the ditin compound grew to its maximum (60%
yield) in about 1.5 hours. A further singlet (6 = 6.05) for the
tritin derivative rose to a maximum (70% yield) after 7.5 h.
After 48 hours all signals from cubane hydrogen atoms were
gone, and the formation of the tetratin compound 2 was com-
plete. On a preparative scale (0.1 mmol), reaction of 1 with
seven equivalents of the base in THF at room temperature for
24 hours resulted in 80 YO yield of 3,3,5,7-tetranitro-2,4,6,8tetrakis(trimethy1stannyl)cubane (2), a stable, white solid, easily
purified by column c h r o m a t ~ g r a p h y . [ ~ ]
Single-crystal X-ray structural analysis of 2 (Fig. 1) confirmed the assigned structure and revealed disorder amongst the
Stable Tin and Lead Derivatives of
Nitrocubanes: Their Formation and Use
in Multiple Functionalization**
Kirill Lukin, Jianchang Li, Richard Gilardi, and
Philip E. Eaton*
The discovery that 1,3,5,7-tetranitrocubane (1) is acidic
enough to permit preparation of its monosodium salt led us to
attempt making polymetallic derivatives with metals forming
more covalent bonds.[']
Trialkyl(amin0)stannanes can metalate weak acids such as
acetonitrile and cyclopentadiene.[*]The mechanism of such metalations is not known, but we suspect i t is not simple. Nonethe-
Prof. P. E. Eaton, Dr. K. Lukin
Department of Chemistry. The University of Chicago
5735 S. Ellis Avenue. Chicago, 1L 60637 (USA)
Dr. J. Li
Geo-Centers Inc.
762 Route I S South. Lake Hopatcong, NJ 07849 (USA)
Dr. R. Gilardi
Laboratory for the Structure of Matter
The Naval Research Laboratory Washington, D C. 20375 (USA)
We are grateful to our colleague Prof. Larry Sita for a stimulating conversation
about tin amides. We thank Dr. Eric Punzalan for samples of 1. This work was
supported by The U.S. Army Research. Development and Engineering Center
(Geo-Centers) and by the Office of Naval Research.
Fig. 1. The molecular structure of crystalline 2. Anisotropic thermal ellipsoid envelopes are shown at the 25% population density level. For clarity, only one nitro
group and the three closest stannyl groups are shown. The observed disorder (55:45)
of the nitro groups is illustrated by the shaded (primed) and tinshaded constituent
nitro groups.''' This is noteworthy, as commonly the intermolecular electrostatic interactions amongst highly polar nitro
groups of nitro compounds, including other nitrocubanes,@'
result in a high degree of ordering. Apparently the bulky trimethylstannyl groups of 2 serve as 'umbrellas,' distancing the nitro
groups of adjacent molecules and thus reducing their interactions. The closest intermolecular approaches of oxygen atoms in
2 are greater than 5.5 A, whereas in tetranitrocubaiie 1 each of
the four nitro groups is less than 3.0 A from a nitro group atom
on a neighboring molecule in the crystal.
The general order for the rate of cleavage of carbon-tin
bonds with electrophiles is ethynyl > phenyl > vinyl > methyl
> butyl and thus correlates apparently with the acidity of the
-0833/96/3508-0866$ 15.1X1+.25/f)
An.g-en. Chem. Int. Ed. Engl. 1996.35. No. 8
Без категории
Размер файла
378 Кб
acidity, stabilizer, group, nitrocubanesчpenta, anion, hexanitrocubanes, nitra, ortho, metalation
Пожаловаться на содержимое документа