An Unusual Reduction of Ethylene Occurring during the Thermal Decomposition of Alkalides and Electrides

Article abstract of DOI:10.1016/S0040-4039(00)93422-1

Alkalides and electrides thermally decompose by a reductive elimination process leading to the formation of ethylene and a ring opened complexant which has lost an ethylenic unit.Surprisingly, gas phase nmr also shows the formation of ethane and butane.We postulated that these compounds arise from a reduction of the first-formed ethylene by the alkalide or electride.

Full text of DOI:10.1016/S0040-4039(00)93422-1

Tetrahedron Letters, Vol.32, No.38, pp 5039-5042, 1991
Printed in Great Britain
0040~1039/91 $3.00 + .00
Pergamon Press pie
An Unusual Reduction of Ethylene Occurring during the
Thermal Decomposition of Aikalides and Electrides.
P. M. Cauliez, J. E. Jackson* and J. L. Dye*
Department ofChemistry and Center forFundamental Materials Research,
Michigan State University, East lansing, Michigan 48824.
Abstract: AIkalides and electrides thermally decompose by a reductive eliminationprocess leading to theformation of
ethylene and a ring opened complexant which has lost an ethylenic unit. Surprisingly, gas phase nmr also shows the
formation of ethane and butane. We postulate that these compounds arise from a reduction of the first-formed ethylene
by the alkN'__~__ or electride.
Alkalides are ionic salts consisting of an alkali metal cation encapsulated in an organic complexant (crown
ether, cryptand, aza crown ether...) and of an alkali metal anion. Since the 1974 preparation and characterization
of Na+C222Na-,1 the first crystalline alkalide, many other alkalide salts have been prepared. Analogous species
in which the anion is an electron are called electrides. To date, crystal structures have been obtained for more
than 25 alkalides and four electrides.
Alkalides and eleetrides are potent reducing agents. They should be 2 and 1 electron reductants,
respectively. Their application for organic reductions has been limited to only a few examples, in which the
alkalide is prepared in situ by reaction of a solution of a crown ether (or cryptand) with a metal mirror in an inert
solvent.2'3 Recently, they have been used to prepare nanoscale metal particles by homogeneous reduction of
metal salts.4
In addition to their high reactivity as reductants, alkalides and electrides are thermally sensitive. They can
be prepared and studied in pure crystalline form only at low temperatures. They are stable for long periods of
time below about - 60 °C, but when warmed to ambient temperatures, they decompose irreversibly in periods
that range from a few seconds to several days. Decomposition processes have been studied only in solution,
where cleavage of the solvent by the strongly reducing alkalide or electride occurs.5 Because of our interest in
the stability of crystalline alkalides and electrides, we have investigated the products and possible mechanisms of
their solvent-free degradation.
We first examined the decomposition products of the compounds (MCt0+M'- and (MCn)+e- (where M and M'
are alkali metals, C is a crown ether and n is 1 or 2). Crystalline samples of active compounds were thermally
decomposed at room temperature under an inert atmosphere. Depending on the alkalide or electride studied, this
process took from a few minutes for the more sensitive compounds to several days for the more stable species.
Complete decomposition was indicated by fading of the color, after which the sample was quenched with D20.
5039

5040
IH and 13C NMR showed that a substantial amount of crown ether remained unchanged. The main dissolved
organic product (after the quench) was the polyethylene glycol resulting from loss of an ethylenic unit from the
crown ether. These results suggest that the main pathway is a "classical" reduction of the crown ether,
analogous to the reduction of vicinal dihalides,6 leading to olef'ms (Scheme 1). The alkali metal cation in the
crown ether may assist the cleavage by polarizing the C-O bonds which are broken.
Scheme 1
r"o"-i
o( o
o( o-
+
II
(o
o]
(o
o-
~,.~°-~-)'Jm
~,,10~m
analogous to
x
R
2 e
X
R
÷
X~'R
X -
R
In order to probe this pathway more completely, the alkalide or electride was thermally decomposed in a
sealed NMR tube, and the gaseous products in the headspace were examined by 1H NMR. In addition to the
expected ethylene, these experiments surprisingly revealed significant quantities of ethane and butane among the
decomposition products.7 Despite substantial effort, we have been unable to establish the source of the extra
protons in these saturated products.
Quantitative analysis was accomplished using NMR and a gas evolution apparatus which has been
previously described in detail.8 Cold alkalide samples were placed in a weighed vessel which was then
connected to the apparatus and evacuated before decomposition. A calibrated pipet was used to measure the gas
volume and a mercury leveling bulb was used to measure its pressure. The gases were then condensed in an
NMR tube and the composition of this mixture was determined by gas phase NMR. Relaxation times for all the
gases studied were determined, and sufficient delay times were used to yield reliable integrations. The non-
volatile residue was weighed, dissolved in D20 and its composition determined by IH and 13C NMR.
Examples of the results obtained for an alkalide [Cs+(18C6)2Cs-] and an electride [K+(15C5)2e-] are
presented in Table 1. In both cases the quantities of ethylene, ethane and butane found account quite well for the
amount of crown ether converted. Thus, we obtain rather good carbon balances for the decomposition. We may
also examine the "electron balance" by counting the reducing equivalents consumed. Reduction of the crown
ether to polyethylene glycol is expected to need 2 electrons per crown ether. Similarly, formation of ethane and
butane from ethylene requires 2 additional electrons in each case. Since the alkalide is expected to give 1 electron
per crown ether, while for the electride it is only 1/2 electron per crown ether, we can estimate "electron yields"
by dividing the number of electrons used to produce the reduced species by the number of electrons available.
These "yields" are respectively ca. 76 % and ca. 93 %. While the second is quantitative within experimental
error, it is not clear why the first one is too small. Some possible explanations are as follows: i) decomplexation
occurred during the decomposition yielding alkali metal that might not be a strong enough reductant to react with
the crown ether or with ethylene; ii) sample was partially decomposed before the decomposition study;

5041
iii) sample was contaminated by an excess of crown ether;, iv) reaction was not completed (our only criterion to
determine the end of the decomposition is the color of the alkalide).
: Products Obtained by Decomposition of an Alkalide and an Electride.
Crown Ether Derivatives (%)
Gases (mol % relative to Starting Material)
Compound
Un~acted
Crown Ether
74.4
Polyethylene
Glycol
25.6
Ethylene I Ethane
Butane
Total Yield of
Gases*
Cs+(18C6)2Cs -
K+(15C5)2e-
10.9
12.4
7.9
0.6
4.4
2.5
27.6
18.0
79.9
20.1
P
*This molpercent is given in terms ofC2 units (the butane percentage is multiplied by 2, the number of C2 units incorporated)
In search of the origins of the reduction products, we used fully deuterated ethylene to demonstrate that
gaseous ethylene can be reduced by a crystalline alkalide at low temperature. In a typical procedure, a cold active
alkalide was quickly loaded, under an inert atmosphere of nitrogen, in the reaction vessel. After evacuation to
ca. 10-5 Tort, the vessel was connected to another vessel that contained the desired amount of ethylene-d4, and
the latter was condensed into the reaction vessel. The temperature was maintained at - 40 °C (a temperature at
which the alkalide is relatively stable) until the alkalide was completely decolorized (roughly a week in the case
of Cs+(18C6)2Cs-). The gases were then condensed into an NMR tube which was then sealed and examined by
1H and 2H gas phase NMR. Three deuterated gases were observed in these experiments: unreacted ethylene-d4,
ethane-d4 and butane-d8 (Scheme 2). Typical yields obtained (based on ethylene-d4) were ca. 5 % of ethane-d4
and ca. 70 % of butane-d8.
Scheme 2
D2C;CD2
25 %
Alkalide
D2C=CD2
l=,,
HD2C-CD2H
5 %
-
4 0 °C
+
HD2C-D2C-CD2-CD2H
70 %
It appears from the labeling studies above that ethane and butane are produced during normal
decomposition by further reaction of the initially formed ethylene with undecomposed crystalline alkalides or
electrides. This sequential pathway was confmned by an experiment in which the gaseous products were
continuously removed from the decomposition mixture by trapping them in a side arm at liquid nitrogen
temperature. Ethylene was then the only gas observable by 1H gas phase NMR, implying its intermediacy in the
paths to alkanes.
The observation of alkene reduction by alkalides in solution has been reported.3 However, the strongly
negative electron affinity of ethylene in the gas phase9 makes the present observation particularly intriguing. In
the tight of recent matrix isolation studies,10 where an electron was trapped by an alkene in alkane matrices, we
think it likely that ethylene in the polar environment of the solid alkalide may similarly trap an electron. The
resulting radical anion would then produce ethane and butane by various sequences of protonation or hydrogen
atom abstraction, fiaxher electron transfer and dimerization, in unknown order.11
An unresolved problem about our suggested mechanism is the nature of the proton source. Despite careful
examination of the reaction mixtures, we have not been able to identify any other products than the ones reported

5042
above. Specifically 1H and 13C NMR of the quenched solutions do not show any olefmic or aldehydic products.
Such compounds would be expected if a proton were abstracted from the crown ether. When methyl iodide was
used as a quencher, polyglymes were obtained instead of the polyethylene glycols, but no other methylated
products were observed. This rules out the presence of carbanions in the reaction mixture. Further experiments
to identify the proton source are in progress.
Acknowledgments
: This work was supported by the Camille and Henry Dreyfus Foundation via a
distinguished young faculty award to J.E.J. and by NSF Grant No. DMR-87-14751
to LL.D. The authors thanks Dr. Evelyn P. Jackson and Mr. Ronald T. Kaufmann for
performing some of the preliminary solution NMR studies.
References and Notes
1.
2.
Dye, J. L.; Ceraso, J. M.; Lok, M. T.; Barnett, B. L.; Tehan, F. J. J. Amer. Chem. Soc. 1974, 96, 608.
see for example: (a) Jedlinski, Z.; Misiolek, A.; Glowkowski, W.; Janeczek, H.; Wolinska, A.
Tetrahedron 1990, 46, 3547-3558. (b) Barrett, A. G. M.; Godfrey, C. R. A.; Hollinshead, D. M.;
Prokopiou, P. A.; Barton, D. H. R.; Boar, R. B.; Joukhadar, L. J.; McGhie, J. F.; Misra, S. C. J.
Chem. Soc. Perkin Trans. I, 1981, 1501-1509. (c) Ohsawa, T.; Mitsuda, N.; Nezu, J.; Oishi, T.
Tetrahedron Lett., 1989, 30, 845-846 and references cited therein.
3.
4.
5.
6.
Mathre, D. J.; Guida, W. C. Tetrahedron Lett. 1980, 21, 4773-4776.
Tsai, K.-L.; Dye, J. L. J. Amer. Chem. Soc. 1991, 113, 1650.
Jedlinski, Z.; Stolarzewicz, A.; Grobelny, Z. Makromol. Chem. 1986, 187, 795.
for a review, seeBaciocchi, E. t,2-dehalogenations and related reaction~ In The Chemistry ofFunctional
groups, Supplement D; Patai, S.; Rappoport, Z. Eds; John Wiley and Sons Ltd.: New York, 1983; pp.
161-201.
7.
The formation of all gases was confirmed by gas chromatography. The headspace gas phase NMR
technique has recently been applied to the study of polyphosphoric acid catalyzed hydrocarbon formation
from methanol: Jackson, J. E.; Bertseh, F. M. J. Amer. Chem. Soc. 1990, 112, 9085.
Dewald, R. R.; Dye, J. L. J. Phys. Chem. 1964, 68, 128.
8.
9.
Jordan, K. D.; Burrow, P. D. Chem. Rev. 1987, 87, 557-588 and references cited therein.
10. Muto. H.; Nunome, K.; Matsura, K. J. Amer. Chem. Soc. 1991, 113, 1840-1841.
11. A similar mechanism has recently been proposed for the reaction of lithium with ethylene at low
temperature: Van Eikema Hommes, N. J. R.; Bickelhaupt, F.; Klumpp, G. W. Angew. Chem., Int. Ed.
Engl. 1988, 27, 1083.
(Received in USA 21 May 1991)