Frequency of C-allylations on oligoglycinates via N-ylides

Article abstract of DOI:10.1021/jo300484x

The highly selective mono-C-allylation of oligoglycinates such as a diethylenetriaminepentaacetate, an iminodiacetate, and an ethylenediaminetetraacetate via insertion of a vacuum operation between the N-allylation and C-migration steps is reported. It is contrastive that one-pot N-allylation-C-allylation procedure gave a mixture including multiallylated products. In the reaction with N-ylides, gem-C-diallylation and α,α′-C-diallylation of oligoglycinates are strongly inhibited even with the use of an excess of allyl bromide and base. A mechanism to explain this control of the frequency of C-allylation on oligoglycinates via N-ylides is also proposed.

Full text of DOI:10.1021/jo300484x

Article
pubs.acs.org/joc
Frequency of C-Allylations on Oligoglycinates via N-Ylides
Hisao Nemoto,* Shin-ya Tamagawa, Kenji Yatsuzuka, and Tomoyuki Kawamura
Department of Pharmaceutical Chemistry, Institute of Health Biosciences, Graduate School of The University of Tokushima, 1-78-1
Sho-machi, Tokushima 770-8505, Japan
S
* Supporting Information
ABSTRACT: The highly selective mono-C-allylation of
oligoglycinates such as a diethylenetriaminepentaacetate, an
iminodiacetate, and an ethylenediaminetetraacetate via in-
sertion of a vacuum operation between the N-allylation and C-
migration steps is reported. It is contrastive that one-pot N-
allylation−C-allylation procedure gave a mixture including
multiallylated products. In the reaction with N-ylides, gem-C-
diallylation and α,α′-C-diallylation of oligoglycinates are
strongly inhibited even with the use of an excess of allyl
bromide and base. A mechanism to explain this control of the frequency of C-allylation on oligoglycinates via N-ylides is also
proposed.
It is considered that the transient concentration of N-ylide 3
in each moment is low, since 3 can be instantaneously
converted to produce 4. Accordingly, mild bases such as
anhydrous potassium carbonate (K2CO3) can be strong enough
to induce the rearrangement. In contrast, ordinary carbanion
chemistry often requires the generation of stoichiometric
equivalents of an unstable (highly reactive) carbanion using
strong bases such as organolithiums or lithium amides.
Furthermore, the molar ratio of each reagent must often be
delicately controlled, and the exhaustive elimination of moisture
or air must be achieved. In spite of such laborious efforts, gem-
dialkylation can sometimes occur because of proton-transfer
processes between a starting enolate and a produced
monoalkylated carbonyl compound. In contrast, the gem-
diallylation of a glycinate via an N-allylated ylide has not been
reported even without rigorous control of the molar ratio
between 1 and an electrophlie such as allyl bromide.1−9
Accordingly, we concluded that rearrangement of 3 via [2,3]-
sigmatropy might be a more efficient and facile carbon−carbon
bond formation reaction of glycinates than ordinary carbanion
chemistry and the other two rearrangements.
INTRODUCTION
■
Base-promoted C-migration of one of four N-substituents of an
ammonium cation, as well as P-, S-, or Se-analogues, has been
classified as the Stevens rearrangement1−4 in a broad sense. In a
narrow sense, the Stevens rearrangement is often defined as
reaction only via [1,2]-sigmatropy. In fact, the reaction from N-
benzylic derivatives to ortho-substituted products via [2,3]-
sigmatropy has been discriminated as the Sommett−Hauzer
rearrangement.5−7 As shown in Scheme 1, the rearrangement of
Scheme 1. C-Allylation of Glycinates 1 via [2,3]-Sigmatropic
Rearrangement of N-Allyl Ylide 3
The carbon−carbon bond formation reaction of diethylene-
triaminepentaacetic acid (DTPA) pentaester 5 with organic
halide (R3X) to afford 6 as reported by us10 and two
independent groups11,12 is a typical example depicting the
laboriousness and difficulty of the unstable cabanion chemistry
(Scheme 2). By using an in situ generated ordinary unsatable
carbanion species, 6 was obtained in low to moderate yields.
Since the DTPA framework has five deprotonative methylenes,
use of excess base and electrophile, prolonging the reaction
period, or raising the reaction temperature increased the
formation of multiallylated byproducts.10 On the other hand,
N-allylated ammonium salts such as 2 via [2,3]-sigmatropy to
afford 4 via N-ylide 3 by Honda and Inoue et al.8 can also be
excluded from the narrowly defined Stevens rearrangement.
The rearrangement via a thermal [1,2]-sigmatropic concerted
pathway is theoretically prohibited, and the [1,2]-sigmatropic
products can often be produced via a higher energetic biradical
pathway.5−7,9 The Sommett−Hauzer rearrangement accom-
panies the collapse of stable aromatic conjugation. Therefore,
the [2,3]-sigmatropic rearrangement of the N-allylic ylide
probably has the lowest transition state energy level among the
three rearrangements.
Received: March 8, 2012
Published: May 1, 2012
© 2012 American Chemical Society
4696
dx.doi.org/10.1021/jo300484x | J. Org. Chem. 2012, 77, 4696−4705

The Journal of Organic Chemistry
Article
and 8L), probably a diastereomeric mixture of di- (8CL and 8LR),
Scheme 2. Previously Reported Carbon−Carbon Bond
Formation Reactions of DTPA Pentaesters 5
and a diastereomeric mixture of tri-C-allylated products 8CLR
,
along with the recovery of a non-negligible amount of starting
material 7. Furthermore, the molar ratio of 7 and 8 in the crude
mixture was irreproducible and extremely sensitive to the molar
ratio of each reagent, temperature, the reaction period, solvent,
and so on. At this point, we realized that, for the same reasons
affecting the carbanion chemistry in Scheme 2, it is difficult to
control conflicting reactivity issues (prevention of multi-
allylations and consumption of a satisfactory amount of starting
material) at the same time using a complicated oligoglycinate
such as 7. In fact, no papers or patents on the application of the
Stevens, Sommelet−Hauser, or Honda−Inoue rearrangements
to complicated oligoglycinates were found during a broad
search of the literature.1−9 Accordingly, the first problem with
carbanion chemistry (delicate control of various conditions)
was unsolved, although the second problem (exhaustive
elimination of moisture and air) was addressed.
Newly Developed Procedure for Selective Mono-C-
Allylation of DTPA Pentaethyl Ester (Newly Devised
Procedure). We have successfully demonstrated highly
selective mono-C-allyation of 7 to afford 8C in 63% isolated
yield via a newly devised procedure that involves mixing 7 and
allyl bromide without base (step A) followed by application of a
vacuum to remove excess allyl bromide (step B) and
irreversible C-migration (step C) (Scheme 4).13,14
suppression of the molar ratio of reagents versus 5, shortening
the reaction period, or lowering the reaction temperature
increased the recovery yield of 5. Furthermore, our reported
reaction10 was often irreproducible on a large scale in spite of
the laborious efforts taken, such as the exhaustive elimination of
moisture and air and careful control of the molar ratio of
substrate and reagents.
RESULTS AND DISCUSSION
■
Representative Result Using the Reaction of DTPA
Pentaethyl Ester and Allyl Bromide with Potassium
Carbonate Mixed in One Portion (Typical One-Pot
Procedure). After reviewing the published unsatisfactory
DTPA chemistry shown in Scheme 2,10 we were looking for
an alternative method for carbon−carbon bond formation at
the framework of 5. The potential for rearrangement chemistry
as a viable approach was an attractive option. We therefore
examined the reaction of 7 and various equivalents of allyl
bromide with various bases and solvents, at different temper-
atures, and reaction periods (Scheme 3).13,14 For convenience,
the nitrogen atoms at right side, center, and left side are
symbolized as NR, NC, and NL, respectively. Because of the
symmetric structure of 7, NL and NR are often merged in the
following discussion. The superscripted letter(s) on 8 indicates
the N-allylated position(s) of the intermediates 9 in Schemes 4
and 5.
This newly devised procedure for the highly selective
formation of 8C from 7 was successful for the following three
reasons: (i) the volatility of allyl bromide plus the nonvolatility
of 9C, (ii) the acceptable stability of 9C at low temperature
during the vacuum operation, and (iii) the thermodynamic
stability of 9C compared with the other ammonium halides
including 9L, 9CL, 9LR, and 9LCR (Scheme 5).
To examine the stability of 9C, the crude mixture was left for
2 days at room temperature after step B and then heated with
K2CO3. The yield of 8C and consumption of 7 were
dramatically decreased, and a non-negligible amount of 8L,
8LR, 8CL, and 8CLR was observed. That is to say, the result was
very similar to the ordinary one-pot procedure with 1 equiv of
allyl bromide. It is believed that 9C was converted back to 7 and
allyl bromide via equilibrium during the 2 days, and then step D
occurred. In contrast, when step C was started immediately
As we anticipated, carrying out the rearrangement was much
simpler and more facile than ordinary anhydrous carbanion
chemistry.10−12 The results, however, were unfortunately not
simple.15 Purification of the crude products afforded mono- (8C
Scheme 3. Allylation of 7 Using the Typical One-Pot Procedure
4697
dx.doi.org/10.1021/jo300484x | J. Org. Chem. 2012, 77, 4696−4705

The Journal of Organic Chemistry
Article
Scheme 4. Newly-Devised Procedure Involving a Vacuum Operation to Remove Excess Allyl Bromide
Scheme 5. Proposed Explanation for the Greater Thermodynamic Stability of 9C versus Other Salts
after step B, chemical yield and selectivity for 8C were excellent,
as we mentioned in the first sentence of this subsection.
It was expected that the ammonium cation 9C would be more
stable than the cation 9L since the quaternary ammonium salt
on NC is more stable than that on NL because of the stronger
basicity and nucleophilicity of NC, which bears one electro-
negative neighboring group (−CO2Et) compared to two
−CO2Et neighboring groups on NL. In fact, when step A was
continued for a long enough period (36 h), 8C and 8L were
obtained in 63 and <5% yield, respectively. In contrast, when
step A was continued only for a short period (4 h), 8C and 8L
were produced in 15 and 25% yield, respectively. Thus,
formation of 9L (= 9R) could be kinetically faster than that of
9C. It is likely that the concentration of 9L was decreased hour
by hour during step A as it underwent thermodynamic
convertion to 9C via either intramolecular allylic transformation
among the three nitrogens or deallylation−reallylation via 7.
Next, we considered why 9C is thermodynamically more
stable than the other salts such as 9L, 9CL, 9LR, and 9CLR. We
assumed that the di- and trication species 9CL, 9LR, and 9CLR are
less stable than the monocations 9L and 9C, mainly because of
the electronic repulsion of cationic nitrogens and the steric
repulsion of the allyl groups. Accordingly, we initially compared
the stability of 9C and 9L.
Discovery of the Frequency of C-Allylations. With
these results in hand, attention was again paid to the results
shown in Scheme 3. Using an unsophisticated mathematical
calculation, it was determined that the possible number of zero
to fully allylalted derivatives of 7 is 1024 (= 210) because of the
presence of 10 deprotonatable hydrogens in 7.16 After
diastereomeric and enantiomeric duplicates were merged, the
possible number of compounds was reduced to 108 (= 3 × 6 ×
6).16 Up to this point, however, the symmetry of 7 was ignored.
Next, gem-diallylated compounds were excluded on the basis of
the result of our 13C NMR analysis17 and the results reported
by various researchers,1−9,15 giving 18 (= 2 × 3 × 3)
compounds.16 Finally, six compounds were merged because
of the symmetry of 7 to leave 12 compounds,16 which are listed
in Table 1 (in fact, 11 compounds are listed because 7 itself is
excluded).
According to the mechanism shown in Scheme 1, all of the
generated intermediate(s) for the N-allylated ammonium
cations of the 11 compounds listed in Table 1 can be easily
illustrated. For example, 8LLR can be obtained via zero NC-, two
4698
dx.doi.org/10.1021/jo300484x | J. Org. Chem. 2012, 77, 4696−4705

The Journal of Organic Chemistry
Article
Table 1. Detected Products and Undetectable Compounds of Multi-Allylated Pentaethyl Diethylenetriaminepentaacetate 8 from
7 via the One-Pot Procedure
NL-, and one NR-allylated cationic intermediates. In this case,
the value is “0−2−1” in the column at the right-hand side of
Table 1.
the conditions of the ordinary one-pot procedure even in the
presence of an excess amount of allyl bromide and K2CO3.
Next, we considered the reasons why only one allylation can
occur on one nitrogen. Singly allylated product 17 can be, in
fact, smoothly produced from 15 via N-ylide 16 (Scheme 7).
The exclusion of gem-diallylation is consistent with the
results. The value for NC is either 0 or 1, and the value for NL
(also NR) is less than 3. It is obvious that all of the values for
the detected five products 8C, 8L, 8CL, 8LR, and 8CLR are either 0
or 1. In contrast, in the case of the undetected compounds 8LL,
8CLL, 8LLR, 8CLLR, 8LLRR, and 8CLLRR, the values always contain
the number 2 for at least one of the nitrogens. As a short
summary, it can be concluded that diallylation at Nα and Nα′ as
shown in Scheme 6 did not occur. For convenience, such a
Scheme 7. Hypothesis for Undetected gem-Diallylation and
gemN-Diallylation
Scheme 6. Hardly Observable Dialkylation at Nα and Nα′
(gemN-Alkylation)
diallylation at Nα and Nα′ is called a “gemN-diallylation” in this
manuscript. As far as we can determine, the gemN-diallylation
can occur via dual N-allylation (from 10 to 12, and from 13 to
14) and corresponding subsequent C-migrations (from 12 to
13, and from 14 to 11).
The above discussion led to us to understand the following
simple principle for this reaction: one amino functionality
manages only one C-allylation. As a result, neither the gem-
diallylated nor gemN-diallylated products were detected under
However, N-allylation of 17 (step E = second allylation) cannot
occur since one of the three N-substituents has been changed
to a secondary alkyl group after the first C-allylation.18
We believe that such a delicate increase in the steric
hindrance can inhibit the further N-allylation. In contrast, the
steps from 18 to either 21 or 22 via 19 or 20 cannot be
interfered with since either simple deprotonation or intra-
4699
dx.doi.org/10.1021/jo300484x | J. Org. Chem. 2012, 77, 4696−4705

The Journal of Organic Chemistry
Article
Scheme 8. C-Allylation of EDTA Derivative 23
Scheme 9. Allylation of Glycinates 26 and 27
molecular sigmatropy is not affected by such steric hindrance. It
is concluded, therefore, that extreme minimization of the
presence of 18 is the determinant for undetectable gem-
diallylated and gemN-diallylated compounds, although it is
difficult to decide whether N-allylation of 17 (step E) is
inhibited, or unstable salt 18 is immediately converted back to
17 and allyl bromide. The step from 17 to 18 corresponds to
the step from 13 to 14 in Scheme 6.
C-Allylation of Other Glycinate Derivatives. To explore
the scope of the reaction, we next focused our attention on the
C-allylation of other glycinates. The reaction of ethyl-
enediaminetetraacetic acid (EDTA) tetraethyl ester (23)19
afforded 24 and 25 in 43 and 26% yield, respectively using the
ordinary one-pot procedure illustrated in Scheme 3 (Scheme
8). In the case of the newly devised procedure illustrated in
Scheme 4, 25 was not detected at all, but 24 was produced in
51% yield. Neither gem-diallylation nor gemN-diallylation were
observed at all. Accordingly, highly selective mono-C-allylation
on the EDTA framework was also successful using the newly
devised procedure.
(27)20 (Scheme 9). In the case of 26, allylated product 30 was
obtained in more than 90% yield using both procedures, while
27 was also afforded the monoallylated product 31 in
satisfactory yield using both procedures. Accordingly, even
the ordinary one-pot procedure was synthetically acceptable in
the case of these simple glycinates. After careful observation
and comparisons, however, we discovered several significant
aspects in the case of 27. Although 26 was smoothly consumed
to afford 30 in excellent yield within a few hours in the one-pot
procedure, 27 hardly disappeared even after 48 h. Instead, a
very small amount of byproduct 34 was detected. It is believed
that an equilibrium between 29 and “32 + ethyl bromoacetate”
could lead to formation of 3421 via 33.22 In contrast, the
reaction of 27 via the newly devised procedure simply afforded
a mixture of 31 and recovered 27. Therefore, the suppression of
byproducts such as 34 is an advantage provided by the newly
devised procedure. Additionally, gemN-diallylating compound
35 was not detected at all in either the ordinary one-pot or the
newly devised procedure.
Finally, we isolated compound 28 as a model of the central
ammonium salt 9C and prepared 29 as a model of the edged
salt 9L because neither 9C nor 9L were separated or isolated
because of their instability. The salt 28 was prepared from 26
In contrast, at first glance, it was not apparent that the novel
procedure provided any advantages for mono-C-allylation of
ethyl glycinate 26 and iminodiacetic acid (IMDA) diethyl ester
4700
dx.doi.org/10.1021/jo300484x | J. Org. Chem. 2012, 77, 4696−4705

The Journal of Organic Chemistry
Article
1
Figure 1. H NMR spectra of the stable salt 28 at 60 °C.
1
Figure 2. H NMR spectra of the unstable salt 29 that generates allyl bromide.
and 10 equiv of allyl bromide in tert-butyl methyl ether for 48 h
at 40 °C. In a similar procedure, however, 29 was not isolated
as a precipitate from 27 and allyl bromide. Therefore, the salt
29 was briefly purified by preparative thin layer chromatog-
raphy.
indicates the presence of an equilibrium between 29 and “27 +
allyl bromide”. This result supports the idea that 9C is more
stable than 9L and confirms the highly selective formation of 8C
from 7.
Finally, it is worth noting that in our NMR analysis, we
observed the conversion of a quaternary ammonium halide
(QAX) to a “tertiary amine (TA) + allyl bromide” such as in
the case of 29 to 27 and 29 to 32. The study of the reaction
course from a QAX to a “TA + organic halide” has not been
1
In a solution of DMSO-d6 at 60 °C, the H NMR spectrum
of 28 did not change at all, even after 5 days (Figure 1). In
contrast, extricated allyl bromide was observed in the 1H NMR
spectrum of 29 in DMSO-d6 after 12 h (Figure 2), which
4701
dx.doi.org/10.1021/jo300484x | J. Org. Chem. 2012, 77, 4696−4705

The Journal of Organic Chemistry
Article
hexane/ethyl acetate (1/1) to afford very trace amount of a mixture of
di- and triallylated products (<0.5 mg), 8L (0.38 g, 0.66 mmol, 5%
yield), 8C (4.80 g, 8.30 mmol, 63% yield), and 7 (0.70 g, 1.30 mmol,
10% recovered yield). This result was confirmed by six repeated
examinations of the same scale and was highly reproducible even when
30.0 g (56.22 mmol) of 7 was used to afford 8C in 58% yield (18.7 g,
32.61 mmol). The major product 8C was easily discriminated from 8L
because 8C is more symmetric than 8L, which was easily judged by
NMR.
extensively investigated or reported, although the counter
course (from TA to QAX) has been too trite to mention.
Although the Hofmann elimination is very famous for a
reaction course from a QAX to a TA, the side product is an
olefin and not an organic halide.23,24
CONCLUSIONS
■
In the ordinary one-pot procedure for C-allylation of
oligoglycinates via N-allyl ylides, we found that the maximal
frequency of C-allylation is controlled by the inhibition of the
formation of both gem-diallylation and gemN-diallylation.
Accordingly, the maximal frequency of C-allylation is equal to
the number of amino groups and not the number of
deprotonative protons. By inserting a vacuum operation
between the N-allylation and C-migration steps, highly selective
mono-C-allylation was accomplished, probably because of the
prohibition of multication formation, and successfully applied
to compounds possessing DTPA, EDTA, and IMDA frame-
works. Synthetic applications using a mono-C-allylated
oligoglycinate are now under investgation.25
Characterization of Tetraethyl 2,2′,2″,2‴-(2,2′-(1-Ethoxy-1-
oxopent-4-en-2-ylazanediyl)bis(ethane-2,1-diyl))bis-
(azanetriyl)tetraacetate (8C). A colorless oil: FT-IR (neat) 3628,
3448, 3077, 2981, 2366, 2055, 1732, 1642, 1446, 1370, 1343, 1188,
1
1029, 917, 856, 808, 733 cm−1; H NMR (CDCl3, 400 MHz) δ 5.80
(ddt, J = 16.8, 10.0, 6.8, 1H), 5.08 (d, J = 16.8, 1H), 5.03 (dt, J = 10.0,
0.4, 1H), 4.20−4.13 (m, 10H), 3.57 (s, 8H), 3.50 (t, J = 7.6, 1H),
2.88−2.77 (m, 6H), 2.71−2.66 (m, 2H, 2.51 (ddd, J = 14.0, 7.6, 6.8,
1H), 2.35 (ddd, J = 14.0, 7.6, 6.8, 1H), 1.285 (t, J = 6.8, 12H), 1.276
(t, J = 6.8, 3H); 13C NMR (CDCl3, 100 MHz) δ 172.2 (C), 170.9 (C
× 4), 134.9 (CH), 116.5 (CH2), 63.6 (CH), 60.2 (CH2 × 4), 60.0
(CH2), 55.1 (CH2 × 4), 53.3 (CH2 × 2), 50.2 (CH2 × 2), 34.3
(CH2,), 14.3 (CH3), 14.1 (CH3 × 4); ESI-HRMS m/z [M + H]+ calcd
for C27H48O10N3 574.3340; found 574.3331.
Characterization of Diethyl 2,2′-(2-((2-((1-Ethoxy-1-oxo-
pent-4-en-2-yl)(2-ethoxy-2-oxoethyl)amino)ethyl)(2-ethoxy-2-
oxoethyl)amino)ethylazanediyl)diacetate (8L). A colorless oil:
FT-IR (neat) 3626, 3542, 3453, 3077, 2981, 2938, 2907, 2873, 2386,
2350, 2057, 1883, 1731, 1643, 1465, 1446, 1371, 1344, 1189, 1029,
EXPERIMENTAL SECTION
■
General Information. Reactions were carried out oven-dried or
flame-dried glassware under argon atmosphere unless otherwise noted.
N,N-Dimethylformamide (DMF) was distilled over calcium hydride
under argon atmosphere at reduced pressure, and stocked in a screw-
capped bottle. Sodium hydrogen carbonate was dissolved in
demineralized water to prepare a saturated aqueous solution
(NaHCO3 aq). Column chromatography was performed using silica
1
919, 861, 807, 725, 574 cm−1; H NMR (CDCl3, 400 MHz) δ 5.83
(ddt, J = 16.8, 10.0, 6.8, 1H), 5.08 (d, J = 16.8, 1H), 5.03 (d, J = 10.0,
1H), 4.19−4.12 (m, 10H), 3.60−3.45 (m, 9H), 2.90−2.77 (m, 8H),
2.49 (ddd, J = 10.0, 6.8, 6.8 1H), 2.40 (ddd, J = 10.0, 6.8, 6.8), 1.29−
1.26 (m, 15H); 13C NMR (CDCl3, 100 MHz) δ 172.3 (C), 171.7 (C),
171.4 (C), 171.1 (C × 2), 134.5 (CH), 116.9 (CH2), 64.3 (CH),
60.43 (CH2 × 2), 60.40 (CH2), 60.3 (CH2), 60.2 (CH2), 55.3 (CH2 ×
2), 55.1 (CH2), 53.2 (CH2), 52.8 (CH2), 52.7 (CH2), 52.3 (CH2),
50.7 (CH2), 34.9 (CH2), 14.5 (CH3), 14.34 (CH3), 14.31 (CH3 × 2),
14.28 (CH3); ESI-HRMS m/z [M + Na]+ calcd for C27H47O10N3Na
596.3159; found 596.3152.
gel 60N (spherical, neutral, 63−210 μm). H and 13C NMR spectra
1
were recorded at 300 and 75 MHz, or at 400 and 100 MHz,
respectively, in the presence of tetramethylsilane as an internal
standard unless otherwise noted. Multiplicities are indicated by s
(singlet), d (doublet), t (triplet), q (quartet), and m (multiplet).
Coupling constants, J, are reported in Hz (Hertz).
Representative Ordinary One-Pot Procedure of 7. A mixture
of 7 (100 mg, 0.187 mmol), allyl bromide (48.7 μL, 0.562 mmol), and
K2CO3 (233.0 mg, 1.087 mmol) in DMF (0.937 mL) was heated for 6
days at 40 °C with stirring equipped with calcium chloride tube. After
being cooled to room temperature, the reaction mixture was poured
into NaHCO3 aq (5 mL) and extracted with ethyl acetate (10 mL ×
3). The combined organic layers were dried over K2CO3 and
concentrated in vacuo. The residue was purified by column
chromatography eluted with hexane/ethyl acetate (3/1−1/1) to
afford 8C (21.2 mg, 0.0369 mmol, 20% yield), 8L (15.3 mg, 0.0266
mmol, 14% yield), and 7 (20.5 mg, 0.0383 mmol, 20% recovered
yield), along with an unidentified mixture of several compounds (35−
45 mg). The main components were purified by column
chromatography to afford a mixture of compounds probably assigned
as 8LR and 8CL (a mixture of diastereomeric and regioisomeric
diallylated products, 30.55 mg, 0.0498 mmol, 27% yield) and a mixture
of diastereomeric triallylated products of 8CLR (2.0 mg, 0.003 mmol,
2% yield). This result was merely one example, and each yield of all
the isolated compounds was delicately influenced by the molar ratio of
7, allyl bromide, and K2CO3, reaction period, or reaction temperature.
Newly Devised Procedure from 7 to Afford 8C. A mixture of 7
(7.00 g, 13.1 mmol) and allyl bromide (10.2 mL, 118.1 mmol) in
DMF (175 mL) was heated for 39 h at 40 °C with stirring equipped
with calcium chloride tube. After being cooled to 0 °C, the reaction
mixture was concentrated in vacuo to remove excess of allyl bromide
at 0 °C. The residue was dissolved in DMF (175 mL), and K2CO3
(16.3 g, 118.1 mmol) was added. After being heated at 80 °C for 70 h
with stirring equipped with calcium chloride tube, the resulting
mixture was poured into NaHCO3 aq (300 mL) and extracted with
ethyl acetate (300 mL × 3). The combined organic layers were washed
with brine (200 mL), dried over K2CO3, and concentrated in vacuo.
The residue was purified by column chromatography eluted with
Ordinary One-Pot Procedure of 23 and Allyl Bromide. A
solution of 23 (121 mg, 0.30 mmol) and allyl bromide (260 μL, 3.00
mmol) in DMF (1.0 mL) suspended with K2CO3 (373 mg, 2.70
mmol) was heated at 60 °C for 72 h. After being cooled to room
temperature, the resulting mixture was poured into NaHCO3 aq (4
mL) and extracted with ethyl acetate (25 mL × 3). The combined
organic layers were washed with brine (30 mL), dried over K2CO3,
and concentrated in vacuo. The residue was purified by silica gel
column chromatography eluted with hexane/ethyl acetate (3:1) to
afford 23 (19.0 mg, 0.047 mmol, 15% recovery yield), 24 (63.0 mg,
0.140 mmol, 43% yield), and 25 (34.0 mg, 0.070 mmol, 26% yield).
Newly Devised Procedure of 23 and Allyl Bromide. To a
solution of 23 (61.0 mg, 0.150 mmol) in DMF (0.5 mL) was added
allyl bromide (130 μL, 1.50 mmol) at room temperature, and the
mixture was stirred for 52 h at 40 °C. After being cooled to room
temperature, the resulting solution was concentrated in vacuo to
remove DMF and the excess allyl bromide. The residue was dissolved
in DMF (0.5 mL) and suspended with K2CO3 (187 mg, 1.35 mmol).
The resulting suspension was stirred for 41 h at 60 °C, cooled to room
temperature, poured into NaHCO3 aq (2 mL), and extracted with
ethyl acetate (15 mL × 3). The combined organic layers were washed
with brine (20 mL), dried over K2CO3, and concentrated in vacuo.
The residue was purified by silica gel column chromatography eluted
with hexane/ethyl acetate (3:1) to afford 23 (19.0 mg, 0.047 mmol,
31% recovery yield) and 24 (34.0 mg, 0.076 mmol, 51% yield).
Characterization of 24. A pale yellow oil: FT-IR (neat) 3445,
3078, 2981, 2937, 2907, 2873, 2360, 2342, 2064, 1889, 1732, 1643,
1541, 1521, 1447, 1371, 1343, 1191, 1030, 919, 858, 807, 752, 719
1
cm−1; H NMR (CDCl3, 400 MHz) δ 5.60 (ddt, J = 10.0, 17.2, 7.2,
1H), 5.01 (d, J =17.2, 1H), 4.95 (d, J = 10.0, 1H), 4.13−4.03 (m, 8H),
3.58−3.38 (m, 3H), 3.52 (s, 4H), 2.89−2.73 (m, 4H), 2.42 (dt, J =
4702
dx.doi.org/10.1021/jo300484x | J. Org. Chem. 2012, 77, 4696−4705

The Journal of Organic Chemistry
Article
14.4, 7.2, 1H), 2.33 (dt, J = 14.4, 7.2, 1H), 1.20 (bt, J = 7.2, 12H); 13
C
column chromatography eluted with hexane/ethyl acetate (9:1) to
afford 30 (49.0 mg, 0.14 mmol, 93% yield).
NMR (CDCl3, 100 MHz) δ 172.5 (C), 172.0 (C), 171.4 (C × 2),
134.7 (CH), 117.1 (CH2), 64.1 (CH), 60.4 (CH2 × 3), 60.3 (CH2),
55.2 (CH2 × 4), 52.8 (CH2), 52.6 (CH2), 50.8 (CH2), 34.7 (CH2),
14.3 (CH3), 14.13 (CH3 × 2), 14.09 (CH3); ESI-HRMS m/z [M +
Na]+ calcd for C21H36O8N2Na 467.2369; found 467.2364.
Newly Devised Procedure of 26 and Allyl Bromide. To a
solution of 26 (47.0 mg, 0.150 mmol) in DMF (0.5 mL) was added
allyl bromide (130 μL, 1.50 mmol) at room temperature, and the
mixture was stirred for 39 h at 40 °C. After being cooled to room
temperature, the resulting solution was concentrated in vacuo to
remove DMF and the excess allyl bromide. The residue was dissolved
in DMF (0.5 mL) and suspended with K2CO3 (187 mg, 1.35 mmol).
The resulting suspension was stirred for 70 h at 80 °C, cooled to room
temperature, poured into NaHCO3 aq (2 mL), and extracted with
ethyl acetate (15 mL × 3). The combined organic layers were washed
with brine (20 mL), dried over K2CO3, and concentrated in vacuo.
The residue was purified by silica gel column chromatography eluted
with hexane/ethyl acetate (9:1) to afford 30 (47.0 mg, 0.13 mmol,
90% yield).
Characterization of 25. A pale yellow oil (a 1:1 mixture of
diastereomers): FT-IR (neat) 3445, 3078, 2980, 2935, 2908, 2873,
2360, 2341, 2042, 1868, 1844, 1827, 1731, 1642, 1541, 1521, 1446,
1371, 1339, 1269, 1188, 1156, 1030, 980, 918, 857, 766, 710 cm−1; 1H
NMR (CDCl3, 400 MHz) δ 5.94−5.78 (m, 2H), 5.17−5.03 (m, 4H),
4.184 (q, J = 7.2, 4H), 4.176 (q, J = 7.2, 4H), 3.68−3.46 (m, 6H),
2.96−2.76 (m, 4H), 2.53 (dt, J = 14.1, 7.5, 2H), 2.33 (dt, J = 14.1, 7.5,
2H), 1.31 (t, J = 7.2, 6H), 1.30 (t, J = 7.2, 6H); 13C NMR (CDCl3, 75
MHz) δ 172.54 and 172.51 (C × 2), 171.9 (C × 2), 134.8 and 134.7
(CH × 2), 117.11 and 117.07 (CH2 × 2), 64.5 and 64.3 (CH × 2),
60.5 and 60.4 (CH2 × 2), 53.02 and 52.99 (CH2 × 2), 51.4 and 51.3
(CH2 × 2), 14.5 (CH3 × 2), 14.3 (CH3 × 2); ESI-HRMS m/z [M +
Na]+ calcd for C24H40O8N2Na 507.2682; found 507.2682.
Characterization of 30. A pale yellow oil: FT-IR (neat) 3439,
3062, 3026, 2979, 2935, 2858, 2738, 2366, 2335, 1944, 1869, 1802,
1729, 1641, 1604, 1541, 1496, 1454, 1369, 1339, 1271, 1225, 1178,
1
1136, 1098, 1030, 998, 915, 855, 746, 699 cm−1; H NMR (CDCl3,
Preparation of 26. A suspension of ethyl glycinate hydrochloride
(1.00 g, 7.16 mmol), 2-bromoethylbenzene (5.87 mL, 42.3 mmol),
K2CO3 (6.94 g, 50.2 mmol), and potassium iodide (1.43 g, 8.59
mmol) in acetonitrile (20 mL) was heated at reflux with stirring for 16
h. After being cooled to room temperature, the resulting mixture was
poured into NaHCO3 aq (10 mL) and extracted with ethyl acetate
(100 mL × 3). The combined organic layers were washed with brine
(100 mL), dried over K2CO3, and concentrated in vacuo. The residue
was purified by silica gel column chromatography eluted with hexane/
ethyl acetate (9:1) to afford 26 as a pale yellow oil (2.00 g, 6.44 mmol,
90% yield): FT-IR (neat) 3444, 3085, 3061, 3026, 2980, 2936, 2858,
2472, 1947, 1872, 1804, 1737, 1653, 1603, 1543, 1496, 1454, 1419,
1368, 1270, 1183, 1134, 1080, 1031, 977, 966, 909, 854, 747, 699, 571
cm−1; 1H NMR (CDCl3, 400 MHz) δ 7.30−7.23 (m, 4H), 7.22−7.14
(m, 6H), 4.16 (q, J = 7.2, 2H), 3.44 (s, 2H), 3.00−2.85 (m, 4H),
2.84−2.70 (m, 4H), 1.26 (t, J = 7.2, 3H); 13C NMR (CDCl3, 75 MHz)
δ 172.8 (C), 141.5 (C × 2), 130.1 (CH × 4), 129.7 (CH × 4), 127.4
(CH × 2), 61.8 (CH2), 57.7 (CH2 × 2), 56.6 (CH2), 35.7 (CH2 × 2),
15.7 (CH3); ESI-HRMS m/z [M + Na]+ calcd for C20H25O2N1Na
334.1783; found 334.1756.
Preparation of 28. A mixture of 26 (47.0 mg, 0.15 mmol) and
allyl bromide (130 μL, 1.50 mmol) in TBME (200 μL) was stirred for
48 h at 40 °C. The resulting suspension was filtered at room
temperature, and the residue was washed with TBME to afford 28 as a
white precipitate (44.0 mg, 0.10 mmol, 68% yield): FT-IR (KBr) 3052,
3023, 3000, 2953, 2360, 2342, 1962, 1889, 1824, 1752, 1636, 1601,
1558, 1541, 1497, 1472, 1438, 1405, 1302, 1272, 1208, 1099, 1038,
941, 916, 876, 763, 704, 594, 540, 506, 428 cm−1; 1H NMR (DMSO-
d6, 400 MHz) δ 7.41−7.26 (m, 10H), 6.17 (ddt, J = 10.4, 16.8, 7.2,
1H), 5.80 (d, J = 16.8), 5.71 (d, J = 10.4, 1H), 4.63 (s, 2H), 4.43 (d, J
= 7.2, 2H), 4.26 (q, J = 7.2, 2H), 3.81−3.77 (m, 4H), 3.16−3.12 (m,
4H), 1.26 (t, J = 7.2, 3H); 13C NMR (DMSO-d6, 75 MHz) δ 164.8
(C), 136.0 (C × 2), 129.1 (CH × 4), 128.8 (CH × 4), 128.1 (CH2),
127.1 (CH × 2), 125.5 (CH), 62.5 (CH2), 62.3 (CH2), 60.5 (CH2 ×
2), 56.7 (CH2), 27.9 (CH2 × 2), 13.8 (CH3); ESI-HRMS m/z [M]+
calcd for C23H30O2N1 352.2271; found 352.2268.
400 MHz) δ 7.29−7.22 (m, 4H), 7.20−7.11 (m, 6H), 5.70 (ddt, J =
10.0, 17.2, 7.2, 1H), 5.04 (d, J = 17.2, 1H), 5.00 (d, J = 10.0, 1H), 4.12
(q, J = 7.2, 2H), 3.47 (t, J = 7.2, 1H), 2.99−2.89 (m, 2H), 2.86−2.71
(m, 4H) 2.70−2.60 (m, 2H), 2.45 (dt, J = 7.2, 14.4, 1H), 2.26 (dt, J =
7.2, 14.4, 1H), 1.24 (t, J = 7.2, 3H); 13C NMR (CDCl3, 75 MHz) δ
174.0 (C), 141.8 (C × 2), 136.6 (CH), 130.3 (CH × 4), 129.7 (CH ×
4), 127.3 (CH × 2), 118.1 (CH2), 65.0 (CH), 61.5 (CH2), 54.9 (CH2
× 2), 37.1 (CH2), 35.9 (CH2 × 2), 15.9 (CH3); ESI-HRMS m/z [M +
H]+ calcd for C23H30O2N1 352.2277; found 352.2278.
Ordinary One-Pot Procedure of 27 and Allyl Bromide. A
solution of 27 (44.0 mg, 0.15 mmol) and allyl bromide (130 μL, 1.50
mmol) in DMF (0.5 mL) suspended with K2CO3 (187 mg, 1.35
mmol) was heated at 60 °C for 72 h. After being cooled to room
temperature, the resulting mixture was poured into NaHCO3 aq (2
mL) and extracted with ethyl acetate (15 mL × 3). The combined
organic layers were washed with brine (20 mL), dried over K2CO3,
and concentrated in vacuo. The residue was purified by silica gel
column chromatography eluted with hexane/ethyl acetate (6:1) to
afford 27 (8.0 mg, 0.027 mmol, 18% recovery yield) and 31 (33.0 mg,
0.10 mmol, 66% yield), (80% yield based on conversion of 27), along
with a trace amount of 34.
Newly Devised Procedure of 27 and Allyl Bromide. To a
solution of 27 (44.0 mg, 0.15 mmol) in DMF (0.5 mL) was added allyl
bromide (130 μL, 1.50 mmol) at room temperature, and the mixture
was stirred for 39 h at 40 °C. After being cooled to room temperature,
the resulting solution was concentrated in vacuo to remove DMF and
the excess allyl bromide. The residue was dissolved in DMF (0.5 mL)
and suspended with K2CO3 (187 mg, 1.35 mmol). The resulting
suspension was stirred for 70 h at 80 °C, cooled to room temperature,
poured into NaHCO3 aq (2 mL), and extracted with ethyl acetate (15
mL × 3). The combined organic layers were washed with brine (20
mL), dried over K2CO3, and concentrated in vacuo. The residue was
purified by silica gel column chromatography eluted with hexane/ethyl
acetate (6:1) to afford 27 (15.0 mg, 0.051 mmol, 34% recovery yield)
and 31 (23.0 mg, 0.069 mmol, 46% yield), (70% yield based on
conversion of 27).
Preparation of 29. A mixture of 27 (88.0 mg, 0.30 mmol) and
allyl bromide (260 μL, 3.00 mmol) in diethyl ether (100 μL) was
stirred for 5 days at room temperature. The resulting mixture was
directly purified by preparative thin layer chromatography eluted with
hexane/ethyl acetate (1:1) to afford 29 (41.0 mg, 0.144 mmol, 48%
yield) as a gummy solid.
Ordinary One-Pot Procedure of 26 and Allyl Bromide. A
solution of 26 (47.0 mg, 0.15 mmol) and allyl bromide (130 μL, 1.50
mmol) in DMF (0.5 mL) suspended with K2CO3 (187 mg, 1.35
mmol) was heated at 60 °C for 72 h. After being cooled to room
temperature, the resulting mixture was poured into NaHCO3 aq (2
mL) and extracted with ethyl acetate (15 mL × 3). The combined
organic layers were washed with brine (20 mL), dried over K2CO3,
and concentrated in vacuo. The residue was purified by silica gel
Characterization of 31. A pale yellow oil: FT-IR (neat) 3444,
3064, 3027, 2980, 2936, 2871, 2365, 2341, 1945, 1869, 1731, 1642,
1604, 1541, 1497, 1454, 1371, 1338, 1270, 1186, 1150, 1113, 1030,
1
978, 917, 856, 749, 700 cm−1; H NMR (CDCl3, 400 MHz) δ 7.30−
7.24 (m, 2H), 7.21−7.15 (m, 3H), 5.80 (ddt, J = 7.2, 10.1, 17.2, 1H),
5.08 (dd, J = 1.2, 17.2, 1H), 5.03 (bd, J = 10.0, 1H), 4.15 (q, J = 7.2,
2H), 4.14 (q, J = 7.2, 2H), 3.59 (d, J = 17.2, 1H), 3.52 (t, J = 7.2, 1H),
3.47 (d, J = 17.2, 1H), 3.03−2.87 (m, 2H), 2.86−2.67 (m, 2H), 2.48
(dt, J = 14.4, 7.2, 1H), 2.39 (dt, J = 14.4, 7.2, 1H), 1.27 (t, J = 7.2, 3H),
1.26 (t, J = 7.2, 3H); 13C NMR (CDCl3, 75 MHz) δ 172.5 (C), 140.0
(C), 134.6 (CH), 128.9 (CH × 2), 128.4 (CH × 2), 126.2 (CH),
117.2 (CH2), 64.5 (CH), 60.6 (CH2), 60.4 (CH2), 54.8 (CH2), 52.6
(CH2), 35.3 (CH2), 35.0 (CH2), 14.5 (CH3), 14.3 (CH3); ESI-HRMS
m/z [M + Na]+ calcd for C19H27O4N1Na 356.1838; found 356.1834.
4703
dx.doi.org/10.1021/jo300484x | J. Org. Chem. 2012, 77, 4696−4705

The Journal of Organic Chemistry
Article
Alternative Preparation of 34. A mixture of ethyl glycinate
hydrochloride (279.0 mg, 2.00 mmol), 2-bromoethylbenzene (334 μL,
2.10 mmol), K2CO3 (415.0 mg, 3.00 mmol), and potassium iodide
(33.0 mg, 0.40 mmol) in acetonitrile (4 mL) was heated at reflux for
24 h. After being cooled to room temperature, the resulting mixture
was poured into NaHCO3 aq (10 mL) and extracted with ethyl acetate
(40 mL × 3). The combined organic layers were washed with brine
(50 mL), dried over K2CO3, and concentrated in vacuo. The residue
was briefly purified by silica gel column chromatography eluted with
hexane/ethyl acetate (1:2) to afford ethyl N-(2-phenylethyl)glycinate
(191.0 mg, 0.920 mmol, 46% yield), which was used in the next step
without further purification.
REFERENCES
■
(1) Stevens, T. S.; Creighton, E. M.; Gordon, A. B.; MacNicol, M. J.
Chem. Soc. 1928, 3193−3197.
(2) Vanecko, J. A.; Wan, H.; West, F. G. Tetrahedron 2006, 62,
1043−1072.
(3) Tayama, E.; Orihara, K.; Kimura, H. Org. Biomol. Chem. 2008, 6,
3673−1680.
(4) Sweeney, J. B. Chem. Soc. Rev. 2009, 38, 1027−1038.
(5) Pine, S. H. In Organic Reactions; Dauben, W. G., Ed.; John Wiley
& Sons: New York, 1970; Vol. 18, Chapter 4, pp 403−465.
(6) Lepley, A. R.; Giumanini, A. G. In Mechanisms of Molecular
Migrations; Thyagarajan, B. S., Ed.; John Wiley & Sons: New York,
1971; Vol. 3, pp 297−440.
The briefly purified ethyl N-(2-phenylethyl)glycinate (188.0 mg,
0.910 mmol) was treated with allyl bromide (942 μL, 10.9 mmol) and
K2CO3 (1132 mg, 8.19 mmol) in DMF (3 mL) at 60 °C for 48 h.
After being cooled to room temperature, the resulting mixture was
poured into NaHCO3 aq (10 mL) and extracted with ethyl acetate (40
mL × 3). The combined organic layers were washed with brine (50
mL), dried over K2CO3, and concentrated in vacuo. The residue was
purified by silica gel column chromatography eluted with hexane/ethyl
acetate (14:1) to afford a pale yellow oil (227 mg, 0.789 mmol, 87%
yield from ethyl N-(2-phenylethyl)glycinate). The Rf value of thin
́
(7) Marko, I. In Comparative Organic Synthesis; Trost, B. M., Ed.;
Pergamon Press: Oxford, 1991; Vol. 3, Chapter 3.10, pp 913−974.
(8) Honda, K.; Inoue, S.; Sato, K. J. Am. Chem. Soc. 1990, 112, 1999−
2001.
(9) Vanecko, J. A.; Wan, H.; West, F. G. Tetrahedron 2006, 62,
1043−1072.
(10) Nemoto, H.; Cai, J.; Yamamoto, Y. Tetrahedron Lett. 1996, 37,
539−542.
1
(11) Keana, J. F.; Mann, J. S. J. Org. Chem. 1990, 55, 2868−2871.
(12) Laurent, S.; Botteman, F.; Elst, L. V.; Muller, R. N. Helv. Chim.
Acta 2004, 87, 1077−1089.
layer chromatography and H NMR of the product were completely
identical with 34 obtained from 27. The FT-IR, 13C NMR, and HRMS
of 34 as described below were the data of the product obtained from
this alternative preparation procedure.
(13) Nemoto, H.; Kawamura, T; Yatsuzuka, K.; Kamiya, M. PCT-
appl. 2011, WO2011046007.
(14) Nemoto, H.; Yatsuzuka, K.; Tamagawa, S.-y.; Hirao, N.; Kamiya,
34, a pale yellow oil: FT-IR (neat) 3443, 3078, 3027, 2979, 2935,
2840, 2360, 2341, 1943, 1844, 1801, 1730, 1642, 1604, 1558, 1541,
1497, 1453, 1417, 1369, 1340, 1271, 1225, 1180, 1155, 1028, 996, 969,
M.; Kawamura, T. Synlett 2011, 5, 615−618.
1
917, 747, 699 cm−1; H NMR (CDCl3, 400 MHz) δ 7.27−7.21 (m,
(15) Independent work on the mono-C-allylation of DTPA penta-
tert-butyl ester was reported, although the selectivity was unclear
because they did not mention any details about the byproducts:
Giovenzana, G. B.; Imperio, D.; Lattuadab, L.; Uggerib, F. Org. Biomol.
Chem. 2011, 9, 679−681.
2H), 7.18−7.12 (m, 3H), 5.83−5.65 (m, 2H), 5.18 (bd, J = 17.2, 1H),
5.08 (bd, J = 10.0, 1H), 5.04 (dd, J = 1.6, 17.2, 1H), 4.99 (bd, J = 10.0,
1H), 4.19−4.06 (m, 2H), 3.47 (t, J = 7.2, 1H), 3.44−3.35 (m, 1H),
3.16 (dd, J = 14.4, 7.2, 1H), 2.96−2.86 (m, 1H), 2.81−2.61 (m, 3H),
2.46 (dt, J = 14.4, 7.2, 1H), 2.31 (dt, J = 14.4, 7.2, 1H), 1.24 (t, J = 7.2,
3H); 13C NMR (CDCl3, 75 MHz) δ 172.6 (C), 140.5 (C), 136.9
(CH), 135.1 (CH), 128.9 (CH × 2), 128.3 (CH × 2), 126.0 (CH),
116.9 (CH2), 116.8 (CH2), 63.1 (CH), 60.1 (CH2), 54.5 (CH2), 52.8
(CH2), 35.5 (CH2), 34.5 (CH2), 14.6 (CH3); ESI-HRMS m/z [M +
H]+ calcd for C18H26O2N1 288.1964; found 288.1959.
(16) Details of the mathematical combination for multiallylation of 7
are available in the Supporting Information.
(17) The 13C NMR spectrum of the crude mixture of this one-pot
procedure did not show any significant peaks around 75 ppm. This
value was estimated as the quaternary carbon of gem-diallylated
glycinyl residues using the 13C NMR predicting function of
ChemDraw version 11.
(18) As an example of tertiary amine bearing a secondary alkyl group,
30 in DMF (0.3 M) was treated at 60 °C for 72 h with an extremely
large excess of reagents (50 equiv of allyl bromide and 9 equiv of
K2CO3). Even under such an aggressive condition, the recovered 30
(70−80% yield) was the main ingredient in the crude mixture. The
low mass spectra (ESI) of the crude mixture showed several small
peaks, one of which appeared at 391 (molecular weight of the gem-
diallylated product). Isolation of the gem-diallylated product was,
however, unsuccessful.
ASSOCIATED CONTENT
■
S
* Supporting Information
1
Charts of H and 13C NMR of 8C, 8L, 24−26, 28, 30, 31, and
1
34; charts of H NMR of 29 and two of the roughly purified
compounds [nonseparable 8LR/8CL (also a diastereomeric
mixture), and 8CLR (a diastereomeric mixture)]; details of the
mathematical combinations for multiallylation of 7. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(19) Preparation of 23: Wathier, M.; Grinstaff, M. W. J. Am. Chem.
Soc. 2008, 130, 9648−9649.
(20) Preparation of 27: We introduced a large alkyl group to prepare
27 according to the known method in order to prepare a starting
material that would not be volatile and thus enable better calculation
of recovery yield. The reason for the choice of the large group on 26 is
AUTHOR INFORMATION
■
Corresponding Author
*Fax: +81 (0) 88 633 7284. E-mail: nem@ph.tokushima-u.ac.
the same. Lop
Alvarado, P.; Ramos, M. T.; Avendano, C.; Menen
2005, 3412−3422.
́
ez-Cobenas, A.; Cledera, P.; San
́
chez, J. D.; Lop
́
ez-
̃
jp.
́
dez, J. C. Synthesis
̃
Notes
(21) The byproduct 34 was also prepared via an alternative synthetic
route because it was obtained in a very small amount. See the detail in
the Experimental Section.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
(22) Although the alternative route to afford 34 via N-allylation of
31, followed by elimination of ethyl bromoacetate, can also be
considered, the route via 32 and 33 is preferable. In fact, several
attempts of the reaction of 31 in DMF with K2CO3 and allyl bromide
(50 equiv to 31) did not afford any C-allylated products. This result is
consistent with the phenomenon shown in Scheme 7.
■
We thank Professor M. Ochiai and Dr. K. Miyamoto of The
University of Tokushima for discussions about the conversion
of quaternary ammonium salts to tertiary amines. This work
was partly supported by a Grant-in-Aid for Scientific Research
from the Ministry of Education, Culture, Sports, Science and
Technology of Japan (Grant No. 22590101 in 2010−2012).
(23) Although the reaction from QAX to “TA + organic halides” is
reversible (Hofmann elimination is irreversible), the reported reaction
4704
dx.doi.org/10.1021/jo300484x | J. Org. Chem. 2012, 77, 4696−4705

The Journal of Organic Chemistry
Article
herein may be classified as a subsidiary of the Hofmann elimination.
Incidentally, we found a few old papers where the reaction from QAX
to TA due to a nucleophilic attack by halide anion was described:
Westaway, K. C.; Poirier, R. A. Can. J. Chem. 1975, 53, 3216−3226.
Westaway, K. C.; Ali, S. F. Can. J. Chem. 1979, 57, 1354−1367.
(24) An example from quaternary ammonium perchlorates or
tetrafluoroborates to TA due to an intermolecular nucleophilic attack
by an amine, a thiol, or a carboxylate: Knier, B. L.; Jencks, W. P. J. Am.
Chem. Soc. 1980, 102, 6789−6798.
(25) We have demonstrated an example of the synthetic applications
of this chemsitry.13,14 In our previously reported patent13 and paper,14
the terminal olefinic moiety of 8C and N-Boc-2-(4-iodophenyl)
ethanamine via Mizoroki−Heck reaction was described. We are
planning to synthesize a conjugate molecule possessing high chelating
ability for heavy metal cations and transporting ability to a specific
organ, nidus, or cells by using the resulting DTPA derivative bearing
primary amino group at the edge of the C-branched chain.
4705
dx.doi.org/10.1021/jo300484x | J. Org. Chem. 2012, 77, 4696−4705