Selective cleavage of the linear ether bond in benzyl glycidyl ether and triphenylmethyl glycidyl ether by potassium alkalide as two-electron-transfer reagent

Article abstract of DOI:10.1016/S0022-328X(02)01628-5

The linear ether bond was exclusively cleaved in benzyl glycidyl ether and triphenylmethyl glycidyl ether under the influence of K^-, K⁺(15-crown-5)₂ (1), whereas the strongly strained three-membered oxacyclic ring remained undisturbed. Potassium glycidoxide and benzylpotassium were found as the primary reaction products of benzyl glycidyl ether with 1. Subsequently, benzylpotassium reacted with benzyl glycidyl ether giving the next potassium glycidoxide molecule and bibenzyl. Benzyl phenyl ether was used as a model compound to explain the mechanism of bibenzyl formation. The reaction of triphenylmethyl glycidyl ether with 1 resulted in potassium glycidoxide and stable triphenylmethylpotassium. After treating with a quenching agent a new glycidyl ether or glycidyl ester was obtained from potassium glycidoxide. These results were found when the reaction occurred at the excess of glycidyl ether. In another case, i.e. at the excess of 1 further reactions took place with the participation of potassium anions and various new compounds were observed in the reaction mixture after benzylation or methylation. Thus, the method of substrates delivery influences the course of studied processes in a decisive way.

Full text of DOI:10.1016/S0022-328X(02)01628-5

Journal of Organometallic Chemistry 660 (2002) 6ꢂ13
/
www.elsevier.com/locate/jorganchem
Selective cleavage of the linear ether bond in benzyl glycidyl ether and
triphenylmethyl glycidyl ether by potassium alkalide as two-electron-
transfer reagent
Zbigniew Grobelny a, Andrzej Stolarzewicz a,*, Barbara Morejko-Buz a,
˙
Adalbert Maercker b
a Centre of Polymer Chemistry, Polish Academy of Sciences, 34, Marie Curie-Sklodowska Str., 41-819 Zabrze, Poland
b Institut fur Organische Chemie, Universitat Siegen, D-57068 Siegen, Germany
¨
¨
Received 20 March 2002; received in revised form 25 May 2002; accepted 25 May 2002
Abstract
The linear ether bond was exclusively cleaved in benzyl glycidyl ether and triphenylmethyl glycidyl ether under the influence of
ꢀ, Kꢁ(15-crown-5)2 (1), whereas the strongly strained three-membered oxacyclic ring remained undisturbed. Potassium
K
glycidoxide and benzylpotassium were found as the primary reaction products of benzyl glycidyl ether with 1. Subsequently,
benzylpotassium reacted with benzyl glycidyl ether giving the next potassium glycidoxide molecule and bibenzyl. Benzyl phenyl
ether was used as a model compound to explain the mechanism of bibenzyl formation. The reaction of triphenylmethyl glycidyl
ether with 1 resulted in potassium glycidoxide and stable triphenylmethylpotassium. After treating with a quenching agent a new
glycidyl ether or glycidyl ester was obtained from potassium glycidoxide. These results were found when the reaction occurred at the
excess of glycidyl ether. In another case, i.e. at the excess of 1 further reactions took place with the participation of potassium anions
and various new compounds were observed in the reaction mixture after benzylation or methylation. Thus, the method of substrates
delivery influences the course of studied processes in a decisive way. # 2002 Elsevier Science B.V. All rights reserved.
Keywords: Benzyl glycidyl ether; Potassium alkalide; Linear ether bond cleavage
1. Introduction
atom of the substituent R. Mechanisms of the ether
bond cleavage in the reaction of organoÃalkaliÃmetal
compounds or alkali metals with linear and cyclic ethers
/
/
Glycidyl ethers possess two kinds of ether bonds, i.e.
the cyclic and the linear ones:
were thoroughly discussed in a review [1].
The oxirane ring opening in glycidyl ethers occurs in
the presence of potassium alkoxides [2ꢂ4] or potassium
/
hydroxide [5,6] exclusively in the b-position [5]. Recently
we have found that the same ring opening took place in
the reaction of Kꢀ, Kꢁ(15-crown-5)2 (1) with phenyl
glycidyl ether (R: Ph) being at the excess in the system
[7]. Cleavage of the cyclic ether bond by 1 was followed
in that case by g-elimination. Potassium phenoxide and
potassium cyclopropoxide were the main reaction pro-
ducts. The crown ether ring was opened in a side
reaction. Potassium tetraethylene glycoxide vinyl ether
was the product of the latter. The rate of the process was
very high at ambient temperature. The linear ether bond
in phenyl glycidyl ether was stable in the presence of 1
where R is an alkyl or an aryl group.
The cyclic ether bond can be opened either in the a-
position (the CHÃ
(the CH2ÃO bond scission). The cleavage of the linear
/
O bond scission) or in the b-position
/
ether bond is also possible in two ways. This occurs
either between the oxygen and the carbon atom of the
glycidyl group or between the oxygen and the carbon
* Corresponding author. Tel.:
2712969
ꢁ
/
48-32-2716077; fax:
ꢁ/48-32-
E-mail address: andrzej.stolarzewicz@cchp-pan.zabrze.pl (A.
Stolarzewicz).
similarly as in all described earlier systems [2ꢂ6]. The
/
0022-328X/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 3 2 8 X ( 0 2 ) 0 1 6 2 8 - 5

Z. Grobelny et al. / Journal of Organometallic Chemistry 660 (2002) 6ꢂ
/13
7
attack of the base occurred only at the oxirane ring of
glycidyl ether.
The behavior of Kꢀ in chemical processes differs
from those of typical anions such as HOꢀ or ROꢀ
[12,14,15]. The potassium anion possesses two valence
electrons in its 4s2 orbital. At the beginning of the
reaction Kꢀ transfers one electron to an acceptor
molecule giving Ko. Subsequently, the potassium atom
transfers the second electron, and finally Kꢁ is formed
from Kꢀ, as recently shown by Perrin in his excellent
work [14]. It means that the potassium anion behaves as
two-electron-transfer reagent. Until now only one
exception was found from this rule [16].
In order to confirm the role of Ko in the present work
an additional experiment was carried out using benzyl
glycidyl ether. Methanol was introduced into the reac-
tion mixture as the quenching agent. If Ko remained in
the system, it should react with methanol giving hydro-
gen. However, no gaseous products were observed. As
mentioned the same result was found for triphenyl-
methyl glycidyl ether. The lack of hydrogen in the
protonated samples indicated that Ko was converted
into Kꢁ. Consequently, the two-electron-transfer me-
chanism was proposed also in these studied systems.
The assumption was made that in the first step of the
reaction of 1 with benzyl glycidyl ether one electron is
transferred from Kꢀ to the aromatic ring. It results in
Ko and the radical anion 2 (Scheme 1). The latter
undergoes fragmentation to potassium glycidoxide (3)
and benzyl radical 4. The linear ether bond cleavage
occurs between the oxygen and the benzyl group. The
surprising regioselectivity of the reaction probably is due
to the high resonance stabilization of 4. Butyl glycidyl
ether 5 is formed from 3 after its treating with butyl
bromide. Benzyl radical recombines with Ko giving
In this work we describe results of our investigations
concerning the reaction of benzyl glycidyl ether (R:
CH2Ph) and triphenylmethyl glycidyl ether (R: CPh3)
with 1 in tetrahydrofuran solution. They were model
systems for the study of initiation step in the polymer-
ization of these ethers by potassium alkalide (the name
introduced by Dye [8] for salts containing alkali metal
anions and complexed metal cations). Owing to that the
experiments were conducted at the excess of glycidyl
ethers in relation to 1, i.e. under the polymerization
conditions [9ꢂ11]. However, the influence of increased
/
concentration of 1 on the course of these reactions was
also studied.
2. Results and discussion
The reaction of 1 with benzyl glycidyl ether proceeded
instantaneously. The blue color of 1 solution vanished
immediately after mixing of reagents. It means that all
potassium anions took part in the process [12]. After
that butyl bromide was added to the mixture as a
quenching agent. Two compounds, i.e. butyl glycidyl
ether and bibenzyl were then identified in the analysed
samples. Their yields were small, about 10% for both
products, because the reaction was conducted at the
excess of benzyl glycidyl ether. No gaseous products
were found before and after butylation. Thus, potassium
glycidoxide and bibenzyl were the real reaction pro-
ducts.
Then, the reaction of 1 was carried out with the excess
of triphenylmethyl glycidyl ether. Discoloration of the
solution of 1 was also instantaneous after its introducing
into the ether solution. Butyl glycidyl ether and butyl-
triphenylmethane were found in the samples treated
with butyl bromide. The second product was only
identified by the analysis of its mass spectrum. There-
fore, in a separate experiment the reaction mixture was
quenched by methanol. In this case triphenylmethane
was obtained in the system. Gaseous products were not
found before quenching as well as in the butylated or
protonated samples. It showed that potassium glycid-
oxide and stable triphenylmethylpotassium were pri-
marily formed in the reaction mixture.
benzylpotassium (6) which then takes part in the Wurtz
¨
coupling with benzyl glycidyl ether being in the excess.
Bibenzyl 7 and potassium glycidoxide (3?) without
crown ether are finally formed.
This mechanism, however is questionable. Dimeriza-
tion of benzyl radical might be expected to occur giving
the same final product of the studied process. Proofs
that benzyl radicals really do not dimerize to bibenzyl
but give benzylpotassium are presented in the further
part of this work.
It was then assumed that in the reaction of 1 with
triphenylmethyl glycidyl ether the initial products in-
volve potassium glycidoxide (3) and triphenylmethyl
radical 9 (Scheme 2). The latter recombines with Ko to
triphenylmethylpotassium (10). However, it does not
No products of the oxirane ring opening were
observed in both series of experiments. It indicated
that only the linear ether bond was cleaved by using
potassium anions, whereas the strongly strained three-
membered ring remained undisturbed. According to our
best knowledge such a phenomenon is described in the
literature for the first time. A degenerate rearrangement
of potassium glycidoxide by an intramolecular nucleo-
philic attack to the oxirane ring, however, cannot be
excluded [13].
take part in the Wurtz coupling with triphenylmethyl
¨
glycidyl ether. Strong mesomeric stabilization of the
triphenylmethyl anion, as it was observed earlier [17],
and/or steric effects probably play a significant role in
this case.
The organometallic compounds 6 and 10 did not
cleave the oxirane ring. In contrast to our earlier
findings [7,15,18] the reaction of these compounds

8
Z. Grobelny et al. / Journal of Organometallic Chemistry 660 (2002) 6ꢂ13
/
Scheme 1.
with crown ether was also not observed. Therefore, not
every organometallic compound opens the crown ring,
as it might be considered on the ground of the cited
works. The reason of this phenomenon seems to be clear
for 10 because of its high stability. It is probable that in
the case of 6 the large excess of benzyl glycidyl ether
causes the lack of its reaction with crown ether. In
conclusion it should be stated that the absence after
methylation of the volatile derivative of 15-crown-5
opening product, i.e. tetraethylene glycol methyl vinyl
ether, cannot be a proof that the formation of organo-
metallic compounds does not occur.
prevails in that system [19ꢂ
not observed in our work.
/
21]. Such a phenomenon was
Results presented in this work differ from those
obtained by us in the reaction of 1 with phenyl glycidyl
ether [7]. The aromatic ring in that glycidyl ether is
connected directly with the oxygen atom possessing an
electron donating character. It increases the electron
density of the aromatic ring. Therefore, the electron
transfer from Kꢀ to the oxirane ring is preferred
resulting in such a case in the cleavage of the cyclic
ether bond. The introduction of only one carbon atom
between the aromatic group and the oxygen changes
dramatically the direction of the electron transfer from
Kꢀ to the glycidyl ether molecule. A similar effect was
observed earlier in the reaction of metallic lithium with
It is worth noting that 9 obtained in the reaction of
metallic zinc with triphenylmethyl chloride was found to
exist in an equilibrium with a dimeric form which even

Z. Grobelny et al. / Journal of Organometallic Chemistry 660 (2002) 6ꢂ
/13
9
benzyl radical and phenoxide anion was expected before
the protonation. The amount of benzyl phenyl ether in
the reaction mixture was 10 times lower than that of
potassium metal and 18-crown-6 using for the prepara-
tion of the metal solution. The relationship between
metal and crown ether was initially equal to 1:1 which
indicated that half of 18-crown-6 molecules remained
free in the solution, i.e. did not complex potassium
cations. These free 18-crown-6 molecules can react with
potassium anions in the same way as it was proposed by
Schue´ and coworker [24] for cryptands, i.e. giving
electrons and potassium cations complexed by the
ligand. Thus, electrons prevailed in such system which,
therefore, was named by the authors as ‘a blue solution
of solvated electrons’.
Angelo [25] investigated the cleavage of several ethers
using sodium naphthalenide as the single electron
transfer reagent. Benzyl phenyl ether was selected as
the model compound. The most likely mechanism of its
reaction with naphthalenide was analogous to that
elucidated by Kornblum and coworkers [26]. The
reduction of the ether lead to the radical anion in which
an added electron was situated in the benzyl ring.
Ejection of phenoxide anion provided the benzyl radical.
That after reduction and protonation was converted
into toluene. Traces of bibenzyl were found in the
reaction mixture. The authors concluded that it might
be formed from the benzyl radical coupling.
Woolsey and coworkers [27] used a sodiumꢂhexam-
/
ethylphosphoramide system (prepared at about 2:1
molar ratio of the components) in tetrahydrofuran at
ambient temperature and sodium solution in liquid NH3
with tetrahydrofuran at ꢀ78 8C for the reaction with
/
benzyl phenyl ether. The reductive cleavage of the ether
was observed also in this case. Bibenzyl as one of the
products was formed in 4 and 50% yields, respectively,
probably by the coupling of benzyl radicals as the
authors mentioned.
Cleavage of nitrosubstituted benzyl phenyl ethers by
2,4,6-tri-tert-butylnitrobenzene radical anion was stu-
died by Guthrie and Shi [28]. p-Nitrosubstituted ether
gave 1,2-bis(4-nitrophenyl)ethane, i.e. di-p-nitroderiva-
tive of bibenzyl, as the main reaction product. Then,
naphthyl benzyl ethers were found to undergo scission
Scheme 2.
1-naphthol in NH3ꢂ
/
C2H5OH [22]. In that system a
selective reduction of the non-substituted aromatic ring
took place due to deactivating influence of the electron
donating hydroxyl group connected to the neighboring
aromatic ring.
In the present work the cleavage of the ether bond
occurred between the oxygen and the benzyl or triphe-
of the CH2Ã
anthracene or fluoranthene [29]. In the case of benzyl 1-
naphthyl ether the major product was bibenzyl in 40ꢂ
/
O bond when treated with radical anions of
nylmethyl substituent, i.e. the OÃ
regioselective carbonÃoxygen bond cleavage in aromatic
ethers and esters by potassiumꢂ18-crown-6ꢂtetrahydro-
/
R bond is broken. The
/
/
/
/
75% yields. For benzyl 2-naphthyl ether the main
product was toluene, and bibenzyl was produced in
furan solution at 0 8C were studied earlier by Fish and
Dupon [23]. In the case of benzyl phenyl ether two
products were found after protonation, i.e. toluene and
phenol in almost equimolar amounts. It was claimed
that Kꢁ formed a 1:1 complex with 18-crown-6; Kꢀ
gave Ko and a solvated electron. Then, the electron
only 5ꢂ20% yields. That work was continued by using
/
naphthylmethyl phenyl ethers for the reaction involving
the electron transfer from fluoranthene radical anion or
biphenyl radical anion [30].
The electron transfer from the donor to the ether
molecule with the formation of benzyl radical as the
intermediate product was postulated in all mechanisms
provided the cleavage of the benzylÃ
Decomposition of the intermediate radical anion to
/oxygen bond.

10
Z. Grobelny et al. / Journal of Organometallic Chemistry 660 (2002) 6ꢂ13
/
presented in the Refs. [23ꢂ
/
31]. However, the kind of the
formed could be converted into toluene by tert-butanol
end product depended on the reaction conditions. That
could be bibenzyl or a benzylmetallic compound (tolu-
ene after protonation).
or it reacted with benzyl phenyl ether being in the excess
and bibenzyl 7 was formed after ejection of 12?. A lack
of toluene in the 1:10 experiment shows that in this case
benzylpotassium did not react with crown ether. Its ring
opening products could be formed in the reaction of 1
with methyl iodide, as described in Ref. [15] or in the
reaction conducted in the presence of tert-butanol.
It is worth noting that by the use of the excess of 1 and
at ambient temperature more products were obtained in
the sample methylated after discoloration of the reaction
mixture, i.e. after the decay of potassium anions. Many
of them were identified as toluene, anisole, and 1,2-
diphenylpropane. Ethylbenzene, bibenzyl, 1,2,3-triphe-
nylpropane, tetra- and triethylene glycol dimethyl ethers
were found only in small amounts. It means that
potassium anions reacted also with bibenzyl, which
had been formed during the process. The intermediate
organometallic compound obtained in this case attacked
then the next benzyl phenyl ether molecule. The
presence of 1,2,3-triphenylpropane in the methylated
sample confirms formation of the metalated bibenzyl as
well as the proposed course of its reaction with benzyl
phenyl ether. These findings are in a good agreement
with the experimental results described by Collins et al.
Therefore, we decided to inspect the course of the
reaction of 1 with benzyl phenyl ether, treating it as the
model reaction, and to solve the problem of bibenzyl
formation in our system. Firstly, benzyl phenyl ether
was used in the excess. Bibenzyl and anisole were
observed in the reaction mixture after quenching with
methyl iodide. Then, 1 was used in the excess and the
temperature was diminished to 0 8C. Initially a solution
of benzyl phenyl ether in tetrahydrofuran containing
tert-butanol was slowly dropped into the solution of 1
until its discoloration. This phenomenon was observed
at the molar ratio of reagents equal to 4:10, respectively.
Toluene, anisole, tetra- and triglyme were obtained in
the reaction mixture after quenching with methyl iodide.
In other experiments benzyl phenyl ether in tetrahydro-
furan solution was slowly introduced into the 10-fold
excess of 1. The mixture was still dark blue. Methyl
iodide was then dropped into this mixture until its
discoloration in order to destroy unreacted potassium
anions. Ethylbenzene, anisole, tetraglyme, triglyme and
tetraethylene glycol methyl vinyl ether were found in this
case as the reaction products.
[32]. In that work reactions of sodiumꢂ
in the solvent system glymeꢂtriglyme (without crown
ether) on a series of model compounds were studied to
determine whether the BirchꢂHuckel reduction can be
used in this solvent system to cleave carbonÃ
/potassium alloy
It can be concluded that the scission of the ether bond
occurred between the oxygen and the CH2 group in
benzyl phenyl ether resulting in potassium phenoxide 12
and benzyl radical 4 (Scheme 3). The latter recombined
with Ko and did not dimerize. Benzylpotassium 6
/
/
¨
/carbon
bonds. Sodium anions, potassium cations and small
amounts of potassium anions could be expected to exist
in such sodiumꢂpotassium alloy solution [33]. Ethyl-
/
benzene and 1,2-diphenylpropane were obtained in that
case as methylated products of the bibenzyl decomposi-
tion. Toluene was also formed, however, a mechanism
of this reaction was not presented.
It was then found that at the excess of 1 further
reactions with the participation of potassium anions
occurred also in the case of benzyl glycidyl ether. 1,2,3-
Triphenylpropane and toluene were identified among
other compounds in the benzylated sample obtained at
the 5:1 molar ratio of 1 to benzyl glycidyl ether. The
latter was not observed in the reaction mixture. It means
that potassium glycidoxide underwent destruction quan-
titatively. A weak signal of bibenzyl was found in the
chromatogram of benzylated samples. On the other
hand, toluene, ethylbenzene, 1,2-diphenylpropane,
1,2,3-triphenylpropane, tetraethylene glycol methyl vi-
nyl ether as well as tetra- and triethylene glycol dimethyl
ethers were formed in the system when methyl iodide
was used as the quenching agent.
The 10:1 molar ratio of 1 to benzyl glycidyl ether and
0 8C was applied in a separate experiment. Ethylben-
zene, i.e. the methylated derivative of benzylpotassium,
and not bibenzyl was detected in this case in the reaction
Scheme 3.

Z. Grobelny et al. / Journal of Organometallic Chemistry 660 (2002) 6ꢂ
/13
11
mixture after methylation. That was in a good agree-
ment with the results obtained with benzyl phenyl ether.
Thus, a method of substrates delivery influences the
course of studied reactions in a decisive way. In this
work we wanted to explain a possible mechanism of the
initiation reaction in the polymerization of selected
aromatic glycidyl ethers by 1. The polymerization occurs
always at the excess of monomer in relation to the
initiator. Thus, only the results obtained in the first part
of this work, i.e. at the excess of glycidyl ethers, can be
taken into account in further investigations concerning
their polymerization by potassium alkalide. However, a
big excess of 1 was needed to use for explanation of the
reaction mechanism. These results show additionally
that a simple titration of 1 with an ether or other
compounds until discoloration of the former, which
usually undergoes at the molar ratio of reagents close to
1:1, may provide wrong conclusions.
of 14 MHz. Chemical shifts were referenced to Me4Si
serving as an internal standard for 1H and 13C measure-
ments. Deuterated C3H6O was used as the solvent. IR
spectra were taken on a FTS-40A BIO-RAD instru-
ment.
Benzyl glycidyl ether (Aldrich) was heated over CaH2
for 6 h and then distilled under dry Ar atmosphere; the
fraction boiling at 70 8C/11 Torr was collected. Triphe-
nylmethyl glycidyl ether and benzyl phenyl ether (both
Aldrich) were dried in vacuum at elevated temperature
prior to use. Tetrahydrofuran (POCH) was boiled over
CuCl to decompose peroxides and then over CaH2 for
10 h, and finally it was distilled at 66 8C. This fraction
was dried over metallic potassium for 20 h and redis-
tilled prior to use. 15-Crown-5 (Aldrich) was dried
under vacuum at 50 8C for 8 h. Potassium (Fluka)
was purified in boiling tetrahydrofuran and then dis-
tilled under high vacuum. Butyl glycidyl ether, glycidyl
butyrate, triphenylmethane and bibenzyl (all Aldrich)
were used for GC-MS analyses.
3. Conclusions
4.1. Kꢀ, Kꢁ(15-crown-5)2 alkalide (0.1 M solution in
tetrahydrofuran)
The kind of substituent in the glycidyl ether molecule
exerts a great influence on the mechanism of its reaction
with potassium alkalide. For the first time to our
knowledge it has been demonstrated the possibility of
the selective cleavage of the linear ether bond without
the opening of the oxirane ring present in the same
reagent molecule. Potassium glycidoxide is formed in
this reaction. Until now the synthesis of this alkoxide by
the use of alkalide solutions has not been described in
the literature.
Blue potassium alkalide solution was obtained by
dissolution of the potassium mirror in 0.2 M 15-crown-5
tetrahydrofuran solution, as in Ref. [34]. After 25 min
contact time the solution was filtered through a glass frit
and then used as the reagent.
4.2. Bibenzyl
Crown ether does not take part in the studied
processes as a reagent. Its role is limited to the
complexation of potassium cation in spite of the
formation of organometallic intermediates.
Its NMR spectra were presented in Ref. 35a Mass
spectrum (m/e): 182 [Mꢁ, 94]; 178 (7); 165 (7); 91 (100);
65 (12). IR (neat): 3028, 1600, 1492, 752, 700 cmꢀ1
.
4. Experimental
4.3. Butyl glycidyl ether
GC-MS analyses were run on a 30 m long DB 1701
fused silica capillary column, using a Varian 3300 gas
chromatograph equipped with a Finnigan MAT 800 AT
ion trap detector. The use of quenching agents enabled
the analysis of liquid compounds formed from the non-
volatile reaction products. Diethylene glycol dimethyl
ether was used as the internal standard for the yield
measurement. Gaseous products were searched for by
GC with a 2.4 m long stainless steel column packed with
Its NMR spectra were presented in Ref. 35b. Mass
spectrum (m/e): 131 [Mꢁ, 1]; 87 (24); 75 (41); 73 (64); 57
(100); 55 (90); 43 (67); 41 (87). IR (neat): 2960, 2871,
1465, 1253, 913 cmꢀ1
.
4.4. 1,2-Diphenylpropane
Its NMR and mass spectra were presented in Ref.
[36].
Al2O3, 0.2ꢂ0.3 mm, deactivated with 5% K2CO3, in an
/
INCO 505 gas chromatograph equipped with a flame
ionization detector. 1H-, 13C-, and 39K-NMR spectra
were recorded at 20 8C on a Varian VXR-300 multi-
nuclear pulsed spectrometer operating at the H reso-
nance frequency of 300 MHz, the 13C resonance
frequency of 75 MHz, and the 39K resonance frequency
4.5. 1,2,3-Triphenylpropane
1
Its NMR and mass spectra were presented in Refs.
[37,38].

12
Z. Grobelny et al. / Journal of Organometallic Chemistry 660 (2002) 6ꢂ13
/
4.6. Glycidyl butyrate
4.12. Model processes
Its NMR spectra were presented in Ref. 35c. Mass
spectrum (m/e) 145 [Mꢁ, 1]; 116 (21); 89 (2); 71 (100);
Benzyl phenyl ether in tetrahydrofuran solution was
selected for this purpose. Its reactions with 1 occurred at
various molar ratio of reagents, from 5:1 to 1:10,
respectively. Initial concentration of benzyl phenyl ether
in the solution was equal to 0.5 M in the experiments
conducted at its excess. In other experiments it was
equal to 0.1 M. The reaction temperature was equal to
25 or to 0 8C. In one experiment tert-butanol was
present in the benzyl phenyl ether solution; the molar
ratio of tert-butanol to ether was equal to 2:1. Methyl
iodide was used as the quenching agent.
56 (14); 43 (92). IR (neat): 1740, 1255, 1180, 911 cmꢀ1
.
4.7. Tetraethylene glycol dimethyl ether (tetraglyme)
Its NMR and mass spectra were presented in Ref.
[18].
4.8. Tetraethylene glycol methyl vinyl ether
The method of its synthesis as well as its NMR and
mass spectra were presented in Ref. [18].
Acknowledgements
This work was supported by the Committee of
Scientific Research in Poland (project No. 3 TO9A 097
18).
4.9. Triethylene glycol dimethyl ether (triglyme)
Its NMR and mass spectra were presented in Ref.
[18].
References
4.10. Triphenylmethane
[1] A. Maercker, Angew. Chem. Int. Ed. Engl. 26 (1987) 972 (and
references therein).
Its NMR spectra were presented in Ref. 35d. Mass
spectrum (m/e): 244 [Mꢁ, 100]; 165 (78); 152 (16); 139
(4); 77 (29); 51 (6). IR (neat): 3021, 1596, 1492, 757, 699
[2] P. Banks, R.H. Peters, J. Polym. Sci. Part A-1 8 (1970) 2595.
[3] C.C. Price, M.K. Akkapeddi, J. Am. Chem. Soc. 94 (1972) 3972.
[4] (a) A. Stolarzewicz, D. Neugebauer, Z. Grobelny, Macromol.
Chem. Phys. 196 (1995) 1295;
cmꢀ1
.
(b) A. Stolarzewicz, D. Neugebauer, Z. Grobelny, Macromol.
Chem. Phys. 196 (1995) 1301.
4.11. General procedure
[5] A. Dworak, Z. Jedlin´ski, Polymer 21 (1980) 93.
[6] A. Stolarzewicz, Makromol. Chem. 187 (1986) 745.
[7] Z. Grobelny, A. Stolarzewicz, A. Maercker, W. Demuth, J.
Organomet. Chem. 590 (1999) 153.
The reaction was conducted under dry Ar atmosphere
at 25 8C in a reactor described in Ref. [12]. Kꢀ, Kꢁ(15-
crown-5)2 (10 cm3 of 0.1 M) blue solution was dropped
into 10 cm3 of 0.5 M benzyl glycidyl ether or triphe-
nylmethyl glycidyl ether tetrahydrofuran solution while
mixing, i.e. the molar ratio of reagents was equal to 1:5,
respectively, and glycidyl ethers were still in the excess.
In another series of experiments 0.1 M benzyl glycidyl
ether tetrahydrofuran solution was dropped into 10 cm3
of 0.1 M Kꢀ, Kꢁ(15-crown-5)2 solution until its
discoloration, i.e. until all potassium anions were in
the system exhausted. Potassium anions were still in the
excess in this mixture, except the final moment of the
reaction. In both series of experiments the reaction
mixture was then treated with one of the appropriate
quenching agent, i.e. with butyl bromide, MeOH, MeI,
benzyl bromide or butyryl chloride. The products
obtained were identified by matching their mass spectra,
retention times, as well as NMR and IR spectra with
those of authentic compounds. In a separate experiment
1 cm3 of 0.1 M benzyl glycidyl ether in tetrahydrofuran
solution was introduced into 10 cm3 of 0.1 M Kꢀ,
Kꢁ(15-crown-5)2 solution. That reaction was quenched
with MeI.
[8] J.L. Dye, Angew. Chem. 18 (1979) 587.
[9] A. Stolarzewicz, Z. Grobelny, G.N. Arkhipovich, K.S. Kazanskii,
Makromol. Chem. Rapid Commun. 10 (1989) 131.
[10] A. Stolarzewicz, Z. Grobelny, Makromol. Chem. 193 (1992) 531.
[11] G. Buika, J.V. Grazulevicius, A. Stolazewicz, Z. Grobelny,
Macromol. Chem. Phys. 196 (1995) 1287.
[12] A. Stolarzewicz, Z. Grobelny, J. Grobelny, Spectrochim. Acta
Part A 56 (2000) 1257.
[13] cf. B.A. Ershov, O.A. Ermakov, A.I. Kol’tsov, T.I. Temnikova,
Zh. Org. Khim. 2 (1966) 933 [Chem. Abstr. 65 (1966) 12116f].
[14] Ch.L. Perrin, J. Wang, M. Szwarc, J. Am. Chem. Soc. 122 (2000)
4569.
[15] Z. Grobelny, A. Stolarzewicz, A. Maercker, J. Organomet. Chem.
604 (2000) 283.
[16] Z. Grobelny, A. Stolarzewicz, G. Adamus, G. Buika, J.V.
Grazˇulevicˇius, J. Organomet. Chem 595 (2000) 66.
[17] K.K. Brandes, R. Suhrman, R.J. Gerdes, J. Organomet. Chem. 32
(1967) 741.
[18] Z. Grobelny, A. Stolarzewicz, M. Czaja, W. Demuth, A.
Maercker, J. Organomet. Chem. 64 (1999) 8990.
[19] V.D. Sholle, G.G. Rozantsev, Russ. Chem. Rev. 42 (1973) 1011.
[20] H. Lankamp, W.T. Nauta, S. MacLean, Tetrahedron Lett. (1968)
249.
[21] J.M. McBride, Tetrahedron 30 (1974) 2009.
[22] C.D. Gutsche, H.H. Peter, Org. Syn. Coll. 4 (1963) 887.
[23] R.H. Fish, J.W. Dupon, J. Organomet. Chem. 53 (1988) 5230.

Z. Grobelny et al. / Journal of Organometallic Chemistry 660 (2002) 6ꢂ
/13
13
[24] J. Lacoste, F. Schue´, J. Organomet. Chem. 231 (1982) 279.
[25] B. Angelo, Bull. Chim. Soc. Fr. (1966) 1081.
[33] J. Grobelny, M. Soko´ l, Z. Jedlin´ski, Magn. Res. Chem. 29 (1991)
679.
[26] R.C. Kerber, G.W. Urry, N. Kornblum, J. Am. Chem. Soc. 87
(1965) 4520.
[34] Z. Grobelny, A. Stolarzewicz, M. Soko´ l, J. Grobelny, H.
Janeczek, J. Phys. Chem. 96 (1992) 5193.
[35] Aldrich Library of 13C and 1H FT NMR Spectra 1(2), 8 A (a);
1(1), 373 A (b); 1(1), 1088 B (c); 1(2), 8 C (d).
[36] J.M. Robinson, G.T. Daniel, S.J. Hale, J. Organomet. Chem. 51
(1986) 109.
[27] K.M. Patel, R.J. Baltisberger, V.I. Stenberg, N.F. Woolsey, J.
Organomet. Chem. 47 (1982) 4250.
[28] P. Maslak, R.D. Guthrie, J. Am. Chem. Soc. 108 (1986) 2628.
[29] P. Maslak, R.D. Guthrie, J. Am. Chem. Soc. 108 (1986) 2637.
[30] R.D. Guthrie, B. Shi, J. Am. Chem. Soc. 112 (1990) 3136.
[31] N.L. Holy, Chem. Rev. 74 (1974) 243 (and references therein).
[32] C.J. Collins, H.-P. Hombach, B. Maxwell, M.C. Woody, B.M.
Benjamin, J. Am. Chem. Soc 102 (1980) 851.
[37] A.R. Katritzky, M.F. Gordeev, J.V. Greenhill, P.J. Steel, J.
Chem. Soc. Perkin Trans. 1 (1992) 1111.
[38] K. Przewocki, M.G. Voronkov, L. Przewocka, Z. Varnke, Zh.
Org. Khim. 11 (1989) 2382.