Synthesis of Chiral Bicyclic Guanidinium Salts using Di(imidazole-1-yl)methanimine

Article abstract of DOI:10.1021/acs.joc.6b00283

A detailed account of the synthesis of chiral bicyclic guanidinium salts is presented. This work represents the first systematic investigation of an approach toward the challenging target molecules via a key guanylation step employing di(imidazole-1-yl)methanimine (6) followed by a two-fold cyclization, which resulted in guanidinium salts 8 and 10. Factors governing the regioselectivity of the final cyclization step are discussed based on further data obtained in the course of the attempted syntheses of two additional bicyclic guanidinium salts.

Full text of DOI:10.1021/acs.joc.6b00283

Article
pubs.acs.org/joc
Synthesis of Chiral Bicyclic Guanidinium Salts using Di(imidazole-1-
yl)methanimine
Aleksej Turockin,*,†,§ Roman Honeker,†,∥ William Raven,‡ and Philipp Selig*,†,⊥
̌
‡
†Institute of Organic Chemistry and Institute of Inorganic Chemistry, RWTH Aachen University, Landoltweg 1, 52074 Aachen,
Germany
S
* Supporting Information
ABSTRACT: A detailed account of the synthesis of chiral
bicyclic guanidinium salts is presented. This work represents
the first systematic investigation of an approach toward the
challenging target molecules via a key guanylation step
employing di(imidazole-1-yl)methanimine (6) followed by a
two-fold cyclization, which resulted in guanidinium salts 8 and 10. Factors governing the regioselectivity of the final cyclization
step are discussed based on further data obtained in the course of the attempted syntheses of two additional bicyclic guanidinium
salts.
compound 3 developed by Schmidtchen18 and TBO derivative
INTRODUCTION
It is of common knowledge that guanidines belong to the
■
4 reported by Corey (Figure 2).19 Misaki−Sugimura catalyst
(5), on the other hand, represents a rare exception, as it is the
only successful non-C2-symmetric bicyclic guanidine organo-
catalyst known to date.20−22
strongest organic bases and are notable for their utility as
catalysts and reagents in organic synthesis.1 Among the great
variety of members of the guanidine family, species containing
all three nitrogen atoms embedded into a bicyclic molecular
framework (e.g., compounds 1 and 2 being the simplest
representatives, see Figure 1) deserve special attention: Not
Figure 2. Bicyclic chiral guanidine bases used in molecular recognition
and organocatalysis.
The vast majority of bicyclic guanidines have been
synthesized according to the strategy presented in Figure 3.
Figure 1. Simplest bicyclic guanidine bases 1,4,6-triazabicyclo[3.3.0]-
oct-4-ene (TBO, 1) and 1,5,7-triazabicyclo[4.4.0]dec-5-ene (TBD, 2).
only do these molecules exhibit much stronger Brønsted
basicities as compared to their acyclic counterparts,2 but
bicyclic guanidines have also been recognized as exceptionally
strong nucleophiles.3 Furthermore, their ability to act as
potentially bidentate hydrogen-bond donors has rendered
these molecules into unique multifunctional organocatalysts.4,5
Our first efforts in the field of guanidine organocatalysis
consisted of systematic investigations with the commercially
available base 1,5,7-triazabicyclo[4.4.0]dec-5-ene (TBD, 2) and
provided ample precedence of the multifunctional nature of this
reagent.6−8 These studies have led to the discoveries of TBD-
mediated complex multistep reaction cascades and enabled new
entries to highly substituted heterocycles. The next objective of
our research was the development of asymmetric versions for
those transformations by using chiral bicyclic guanidines.
Chiral bicyclic guanidines have been known for more than
two decades and have found applications in the fields of
molecular recognition9,10 and asymmetric organocatalysis.11−17
Among the structures of interest, one can easily recognize the
striking prevalence of the C2-symmetric bicyclic systems such as
Figure 3. Retrosynthetic analysis of bicyclic guanidines via cyclic
thiourea intermediates.
In every case, an acyclic triamine precursor was used as a key
intermediate which was subsequently converted to a cyclic
thiourea followed by a condensation reaction leading to the
target structure. Special note should be taken of the fact that
late cyclization stages included variations of the thiocarbonyl
24
equivalent (e.g., Cl2CS,23 Me2CS3 or Im2CS25) and were
reported to proceed both as a multistep sequence or a one-pot
reaction.
Among the studies dealing with the syntheses of bicyclic
guanidines, the contributions of Franz Schmidtchen deserve
special attention: In one of his very last published works,26 he
Received: February 7, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
used an approach which made use of guanylating reagent 6,
which had previously been developed by Wu et al. in the
laboratories of Guilford Pharmaceuticals Inc. (Figure 4).27 The
RESULTS AND DISCUSSION
Preparation of Starting Materials. The protected chiral
aminoalcohol 19 was an obvious choice as a key intermediate
for the syntheses of guanidinium salts 8 and 10 and is readily
available from L-methionine following the sequence described
by Davis et al. (Scheme 1).29 It is noteworthy that, after some
■
Scheme 1. Synthesis of the Protected Aminoalcohol 19
Figure 4. Retrosynthetic analysis of bicyclic guanidines by
Schmidtchen approach.
key step was the reaction between a primary amine and reagent
6: The imidazole leaving groups were displaced by 2 equiv of
the amine, and an open-chain guanidinium salt was formed.
Final steps included two-fold intramolecular nucleophilic
substitution affording the desired structure. In contrast to the
approach via triamine precursors, the significant advantage of
this new strategy was the introduction of the critical
guanylation reaction in the early stage of the synthesis.
Subsequent end-game chemistry allowed for robust and reliable
assembly of the final bicyclic guanidine scaffold. As a matter of
fact, this approach was successfully applied to the synthesis of
guanidine 3 (Figure 2). Despite the obvious elegance of this
route, it has remained neglected and was never applied for the
syntheses of other bicyclic guanidines.
In order to investigate the practical applicability and
limitations of the guanylation approach reported by Schmidtch-
en, we set out to prepare guanidinium salts presented in Figure
5. Besides our efforts to access the hitherto unknown
experimentation, the synthesis of this useful building block was
easily scaled up, and we can now provide detailed experimental
procedures for the preparation of gram quantities of chiral
amine 19 (see Experimental Section). The first steps involved
the transformation of L-methionine (11) into L-homoserine
(13) by S-methylation followed by hydrolysis of the
methylsylfonium salt 12 under basic conditions. In agreement
with the work of Davis, precise pH control was vital for the
success of the reaction, since racemization readily took place at
pH values above 6. A further observation was attributed to the
temperature control of the heating medium in the event that
the use of heating mantles, despite their practical convenience,
happened to cause the destruction of the reaction components
affording homoserine as a dark brown solid. Due to insufficient
transparency of the corresponding solutions of compound 13,
using the material obtained this way was troublesome for the
determination of its enantiopurity by polarimetry. We found
that using an oil bath at 140 °C was absolutely sufficient to
circumvent any destruction processes and to obtain enantio-
pure homoserine (13) in decagram quantities as a beautiful
white solid. Subsequent cyclization under acidic conditions and
lactone opening using excess of the phenyl Grignard reagent
delivered the chiral aminodiol 15. Primary alcohol and amine
functionalities were protected by acetylation in order to enable
the removal of the tertiary hydroxyl group by transfer
hydrogenolysis in the next step. After the subsequent
hydrogenylitic reduction, hydrolysis, and TBDPS-protection,
the key target intermediate 19 was obtained.
Figure 5. Target chiral bicyclic guanidinium salts.
guanidinium salts 7 and 8, we also investigated the syntheses
of TBO-derivative 928 and TBD-derivative 10.24 With regard to
the latter two structures, there is no doubt that the planned
study was of special interest for the development of a new and
convenient approach to C2-symmetric guanidinium salts. For
instance, despite a brilliant and a very detailed report by Davis
on guanidine 10,24 it was evident that the synthesis was
anything but simple, as it required careful choices of protecting
groups and included several unstable intermediates. Since the
only known entries to guanidines 9 and 10 had relied
exclusively on the approach rationalized on Figure 3, we were
eager to investigate the applicability of the Schmidtchen’s
strategy as a more robust alternative. We looked upon the
syntheses of those structures as an exercise in guanidine
chemistry allowing us to address the following synthetic
questions: (a) applicability of the chosen approach to the
modular syntheses of non C2-symmetric bicyclic guanidines;
(b) syntheses of C2-symmetric guanidines; and (c) syntheses of
[4.4.0]-, [3.3.0]-, and [4.3.0]-bicyclic systems.
The tert-butyl-substituted protected aminoalcohol 22 was
prepared from L-tert-leucine (20) in two steps by the reduction
30
using NaBH4/I2 followed by O-silylation (Scheme 2).
Substrate 24 was prepared from 3-amino-1-propanol using
the same TBDPS-protection conditions.
Attempted Synthesis of the Chiral [4.3.0]-Bicyclic
Guanidinium Chloride 7. The next objective of our study
was the synthesis of the non-C2-symmetric bicyclic guanidinium
B
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 2. Syntheses of the Protected Aminoalcohols 22 and
24
Scheme 3. Synthesis of the Nonsymmetrical Guanylating
Reagent 26
The subsequent step was the synthesis of the key chiral
guanidine 27 by the attachment of the aminoalcohol 22
through the substitution of the second imidazole moiety (Table
2). No product was formed when we reacted the components
salt 7 from the aminoalcohols 22 and 24 and the guanylating
reagent 6. We chose as a first step the reaction of compound 6
with the protected tert-leucinol 22, leading to the unsymmetric
guanylating reagent 25 (Table 1). The desired intermediate was
Table 2. Synthesis of the Open-Chain Guanidine
Intermediate 27
Table 1. Synthesis of the Chiral Guanylating Reagent 25
entry
solvent
T (°C)
yield of 27 (%)
1
2
3
4
1.4-dioxane
neat
90
−
−
−
−
110
150
200
115
neat
entry
solvent
yield of 25 (%)
neat
a
b
a
5
neat
55
1
2
3
4
THF
THF
THF
neat
62
b
a
b
15
1.0 equiv TFA was added. Product isolated as a trifluoroacetate salt
(27·TFA).
c
22
traces
a
Reaction was carried out on 337 μmol scale in an open round-
b
22 and 26 at 90 °C in dioxane (Table 2, entry 1). Even
performing the reaction at elevated temperatures under solvent
free conditions resulted in no conversion of starting materials
(Table 2, entries 2−4). The elaboration of the guanidine
intermediate 27 was finally possible by running this reaction at
115 °C in the presence of 1 equiv of TFA (Table 2, entry 5).26
It should be noted that the key guanylation step was performed
under solvent-free conditions and the use of stirring bars
became impossible due to an extremely high viscosity of the
reaction mixture. In search for a solution to this practical
obstacle, we found that using a Kugelrohr oven as a reactor was
a very convenient alternative to magnetic stirring and enabled
even mixing of the highly viscous reaction mixture. At this
point, we also faced a significant purification problem, as the
guanidinium salt 27·TFA could not be sufficiently separated
from the unreacted amine 22 using standard eluents (MeOH/
CH2Cl2 or MeOH/EtOAc) for column chromatography. We
found that two chromatographic purifications were necessary to
obtain the product 27·TFA in acceptable purity. For the first
chromatography, we employed silica gel as a stationary phase
and were able to separate the mixture of the product and amine
22 from more polar and apolar impurities, while the second
chromatography was performed using neutral aluminum oxide
to separate the guanidinium salt 27·TFA from the unreacted
amine 22.
bottomed flask. Reaction was carried out on 2.36 mmol scale in an
open round-bottomed flask. Reaction was carried out in a closed
c
round-bottomed flask.
isolated after reacting the starting materials at 40 °C in THF
(Table 1, entry 1). Despite the synthetically useful yield of 62%,
we faced significant obstacles at this stage of the synthesis. First,
steric hindrance of tert-butyl group resulted in a rather low
reactivity of compound 22, and, as follows, this step suffered
from very long reaction times (120 h). Furthermore,
concentration dependence of the one-fold guanylation
represented another problem for the elaboration of the
intermediate 25. This reaction was performed in an open
reaction vessel enabling slow evaporation of the solvent and
leading to gradual increase of the concentrations of the reaction
components. However, we were not able to reproduce this
initially promising result on a seven-fold scale (2.36 mmol
aminoalcohol 22) and obtained the intermediate 25 in only
15% yield (Table 1, entry 2). Conversely, when the reaction
was carried out in a closed round-bottom flask, the yield of the
isolated product was 22% (Table 1, entry 3). As we attempted
this transformation under solvent-free conditions, decomposi-
tion of the guanylating reagent 6 took place to a significant
degree, so that only traces of compound 25 could be detected
(Table 1, entry 4).
Cleavage of the silyl protecting groups was carried out to
afford diol 28 at the next stage (Scheme 4). This deprotected
guanidinium species possessed an extremely high polarity, and
any conventional purification techniques such as extraction or
column chromatography were not applicable at this stage. For
that reason, fluoride on polymer carrier was used instead of
TBAF for the deprotection step, since the latter would produce
In an alternative approach, we reacted the protected
aminoalcohol 24 with the reagent 6. Without any steric
hindrance from the bulky tert-butyl group, the displacement of
imidazole took place significantly faster (23 h) and did not
show the above-mentioned sensitivity to the concentrations of
the reacting components (Scheme 3).
C
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 4. Final Steps in the Attempted Synthesis of the
Guanidinium Salt 7
Figure 6. Key HMBC correlations for the intermediate 31. Formation
of the product with alternative structure 31A was ruled out due to
unlikely correlation over five bonds through two nitrogen atoms.
possible bicycle possessing the structure 30. Indeed, upon
reacting the monocyclic guanidinium salt 31 with DBU at 75
°C, a bicyclic product 30 was obtained, which was spectroscopi-
cally identical to the main product of the two-fold cyclization of
the intermediate 29.
Regioselectivities of the Final Cyclization Steps. A
possible mechanistic rationale for the formation of compound
30 is presented in Scheme 5 and can be generalized to explain
the formation of undesired regioisomers of other target
guanidines mentioned in this manuscript. We initially expected
that the desired guanidinium salt 7 would be formed by a two-
fold intramolecular cyclization at the less substituted nitrogen
atom (Pathways c → h and b → g). These reaction pathways
were supposed to be the preferred ones, because alkyl-
substituted nitrogen atoms were sterically more hindered,
especially those containing the bulky tert-butyl group. After the
isolation of the species 31 and 30, it became obvious that we
were dealing with the alternative reaction pathway: The
reactivity was governed by electronic rather than steric factors,
resulting in preferential nucleophilic attack by a secondary
nitrogen atom (Pathway a → e). Thus, we suggest that a strong
electron-donating effect of the tert-butyl group led to enhanced
nucleophilicity of the neighboring nitrogen atom resulting in
the observed regioselectivity. As a result, it is likely that the first
cyclization preferentially took place on the more nucleophilic
nitrogen atom leading to the intermediate 31. Subsequently,
the second nucleophilic substitution led to the bicyclic
guanidinium salt 30.
Synthesis of the Chiral [4.4.0] Bicyclic Guanidinium
Chloride 8. We now turned to the synthesis of the chiral
benzhydryl-substituted guanidinium chloride 8 (Scheme 6).
The nonsymmetric guanylating reagent 26 was reacted with the
aminoalcohol 19 under the same conditions we applied for the
synthesis of the compound 27·TFA. The key guanidinium salt
32 was isolated in 39% yield after two consecutive chromato-
graphic purifications over silica gel and aluminum oxide. After
deprotection of the alcohol functionalities, the guanidinium diol
33 was chlorinated with thionyl chloride to afford dichloride
34. After the final cyclization step and chromatographic
purification, we were able to obtain the target guanidinium
salt 8 in 44% yield. The structure of this compound was
established by X-ray crystallography and is presented in Figure
7.32 In addition, we isolated a fraction consisting of the desired
product 8 and regioisomer 35 (34% total yield). Thus, by
contrast with the attempted cyclization toward guanidium salt
7, the total yield of bicycle 8 was 69% and clearly prevailed over
the undesired side product 35.
an inseparable mixture with the desired intermediate. Indeed,
the guanidinium salt 28 could be conveniently purified by
simple filtration of the polymer beads and subsequent
precipitation. The intermediate 29 was obtained by reacting
the diol 28 with excess thionyl chloride at 60 °C (when this
reaction was attempted at room temperature (rt) overnight, we
observed incomplete conversion of the starting material). The
final two-fold cyclization step was performed under conditions
analogous to those reported by Schmidtchen et al.: Dichloride
29 was suspended in MeCN and, after addition of an excess
DBU, complete conversion of the starting material was
observed after 18 h.26 Subsequent work up and chromato-
graphic purification afforded guanidinium salts 30 and 31. The
main bicyclic product was the undesired regioisomer and was
isolated as an inseparable 92:8 mixture with another guanidine
species.31 We assume, that the minor component of this
mixture was the desired compound possessing structure 7.
The structure determination of the obtained bicyclic product
30 deserves special comment. It should first be noted that the
standard two-dimensional NMR experiments (NOESY, HSQC,
HMBC) did not provide any conclusion whether the isolated
compound possessed structure 30 or 7. Attempts to obtain
crystalline material suitable for an X-ray study were fruitless.
First, the usual aggregate state of the main product was an oil.
Second, we did observe formation of beautiful clear prisms in
acetonitrile after storing the corresponding solution of the
compound 30 for several days at rt. Nevertheless, after we
removed the acetonitrile supernatant, crystalline material
“melted” to a colorless oil within several minutes. Significant
information regarding the real structure of the product 30 was
obtained after studying the monocyclic guanidinium salt 31.
HMBC spectra of this species showed a clear three-bond
through-nitrogen correlation between the methine proton and a
methylene carbon attached to one of the nitrogen atoms
(Figure 6). This observation enabled us to assume that the
isolated compound indeed possessed the structure 31 rather
than the alternative connectivity 31A: Whereas the long-range
correlation between the nuclei of the compound 31 is
transferred over three bonds, the latter structure would require
a very unlikely correlation across five bonds including two
heteroatoms. After having determined the structure of
monocyclic guanidinium salt 31, we were now able to deduce
the connectivity of the species 30. We envisioned that
cyclization of the guanidinium salt 31 would lead to only one
Attempted Synthesis of the C2-Symmetric Chiral
[3.3.0]-Bicyclic Guanidinium Chloride 9 (Tan’s Guani-
dine). Protected aminoalcohol 22 was used in the key
guanylation step leading to chiral acyclic guanidinium salt 36
(Table 3). Initial reaction conditions were adopted from the
D
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 5. Possible Cyclization Pathways of 29 Leading to Various Bicyclic Regioisomers
Scheme 6. Synthesis of the Bicyclic Guanidinium Salt 8
Table 3. Synthesis of the Open-Chain Guanidinium Salt 36
entry
t (h)
T (°C)
yield of 36 (%)
1
2
3
2
2
4
115
105
105
26
20
40
above-mentioned syntheses of the non C2-symmetric guanidi-
nium salts: We ran the two-fold substitution at 115 °C for 2 h
in the presence of 1 equiv of TFA. Due to the fact that this
initial attempt afforded the desired product in only 26% yield
(Table 3, entry 1), the reaction temperature was reduced to
avoid eventual decomposition of the guanylating reagent 6.33
As a result, we obtained the guanidinium trifluoroacetate 36 in
Figure 7. X-ray crystal structure of guanidinium salt 8.32 Ellipsoids are
drawn at 80% probability. A water molecule has been omitted for
clarity.
E
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
an acceptable yield of 40% after prolonged reaction time at 105
°C (Table 3, entries 2−3).
the final product. Interestingly, no traces of the desired
regioisomer 9 were found in the reaction mixture, as species 39
was the only isolated bicyclic product.
This key step suffered from the same purification problem
described for related intermediates 27·TFA and 32, as the
protected aminoalcohol 22 could not be separated from the
product 36 by column chromatography when standard eluents
consisting of CH2Cl2 and MeOH were used. Owing to the fact
that our synthetic studies also revealed a great sensitivity of this
problematic separation to the activity grade of aluminum oxide,
we also faced problems with the reproducibility of our own
purification procedures. In search for more reliable separation
methods, it was found that guanidinium salt 36 could be easily
separated from unreacted aminoalcohol 22 by a single column
chromatography over silica gel when a small amount of AcOH
or formic acid was added to the eluent. We assume that using
acidic additives led to the formation of the ammonium salt of
the compound 22, which was significantly more polar than the
guanidinium salt 36. As a result, the desired key intermediate
36 could be isolated as a less polar fraction in acceptable purity
free of any starting amine.
Synthesis of the C2-Symmetric Chiral [4.4.0]-Bicyclic
Guanidinium Chloride 10 (Davis’s Guanidine). Our final
endeavor was targeted toward the synthesis of the chiral TBD
derivative 10 first described by Davis et al.24 The guanylation
step was performed under the same reaction conditions we
used for the synthesis of the compound 36, and the open-chain
guanidinium trifluoroacetate 40 was obtained in 38% yield
(Scheme 8). At the next stage, cleavage of the silyl groups was
carried out to give diol 41, which was isolated as a mixture of
guanidinium fluoride and trifluoroacetate salts. Subsequent
reactions with thionyl chloride and DBU-mediated double
cyclization afforded the desired guanidinium salt 10 in 14%
yield along with its regioisomer 43, which was formed in 49%
yield. After column chromatographic purification, product 10
was isolated in 8% yield in pure form.
An X-ray crystallographic study on the bicyclic salt 1032
delivered a very surprising structural insight contradicting the
computational data obtained by Davis et al. In contrast to the
initially proposed preference to equatorial positioning of the
benzhydryl groups,34 we could clearly observe that the bulky
residues occupy the axial positions (Figure 8)! Moreover, the
After subsequent desilylation and chlorination steps,
guanidinium salt 3832 was isolated as a crude mixture and
subjected to the final cyclization step without additional
purification (Scheme 7). Guanidinium salt 39 was obtained as
Scheme 7. Final Steps in the Attempted Synthesis of the
a
Guanidinium Salt 9
Figure 8. X-ray crystal structure of guanidinium salt 10.32 Ellipsoids
are drawn at 80% probability. Water molecules have been omitted for
clarity.
a
Ellipsoids of 38 are drawn at 80% probability.
Scheme 8. Synthesis of the Bicyclic Guanidinium Salt 10
F
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
prepared by adding 0.04 g bromocresol green to 100 mL EtOH. Then
a 0.1 M NaOH solution was slowly dripped in until the solution just
turned pale blue.35 The staining reagent ninhydrin was prepared by
dissolving 1.5 g ninhydrin and 3 mL acetic acid in 100 mL butanol.
NMR spectra were recorded using 300, 400, and 600 MHz NMR
spectrometers. Chemical shifts δ are given in ppm. Calibration of the
spectra was based on solvent residual peaks for the samples recorded
in CDCl3, methanol-d4, DMSO-d6. Coupling constants J are given in
Hz. The following abbreviations are used to describe the signals in the
NMR spectra: s, singlet; d, doublet; t, triplet; q, quartet; quint, quintet;
and m, complex multiplet. The abbreviation “virt x” is used to show
that the detected signal seemed to look like x (e.g., “virt t” stands for
“virtual triplet”). Signal and structure assignments for the compound
31 were performed using 2D NMR methods (HSQC and HMBC).
The representative 2D NMR spectra can be found in the Supporting
Information. HRMS were measured with an LTQ Orbitrap XL
spectrometer.
orientation of the two benzhydryl groups around the
guanidinium active center creates a well-defined groove for
substrate binding which could indeed make compound 10 a
viable chiral organocatalyst. Although we are aware that X-ray
crystallography does not always represent the real conformation
present in the solution, we do hope that this observation
provides a useful foundation for further studies on Davis’s
guanidine 10.
CONCLUSION
■
We have carried out the first systematic study toward chiral
bicyclic guanidinium salts using guanylating reagent 6 with
subsequent two-fold intramolecular nucleophilic substitution.
In course of this work, we have demonstrated the broad
robustness and utility of the key guanylation step, as it was
applicable to the syntheses of both symmetrically and
nonsymmetrically substituted open-chain guanidinium salts.
Most importantly, this approach demonstrated a high tolerance
toward steric bulk of alkylated amines, so that even the strongly
hindered tert-leucinol 22 was a viable substrate for the key
transformation. A special feature of this work is also the
strategic employment of TBDPS protecting groups enabling
convenient purification of guanidinium intermediates. Although
we have demonstrated the applicability of the selected approach
to the modular syntheses of [4.4.0]-, [4.3.0]-, and [3.3.0]-
bicyclic guanidinium salts, regioselectivity in the final
cyclization step remains a yet to be solved problem. Our data
strongly suggest that the preference for the observed reaction
pathways is governed by electronic rather than steric factors.
Our further synthetic studies dealing with chiral bicyclic
guanidinium salts will be reported elsewhere in due course.
Di(1H-imidazol-1-yl)methanimine (6).26,27 The following procedure
was performed under exclusion of moisture. Due to high toxicity and
volatility of bromocyanide, all operations should be performed in a properly
functioning fume hood. Imidazole (3.2 g, 50 mmol, 3.0 equiv) was
dissolved in dry CH2Cl2 (250 mL). Subsequently, a solution of
bromocyanide (1.85 g, 17.5 mmol) in CH2Cl2 (10 mL) was added by
syringe, and the solution heated to reflux for 45 min. After cooling to
rt, the white precipitate was collected by decanting the supernatant
and washing with the cold CH2Cl2. Mother liquor was then
concentrated to ca. 20 mL and cooled to −20 °C for 4 d. Precipitated
solid was filtered and washed with cold CH2Cl2 to give a second crop
of the product. Combined solids weighed 1.92 g (11.9 mmol, 68%). Rf
= 0.34 (CH2Cl2/MeOH = 9/1, [UV]). 1H NMR (400 MHz, DMSO-
d6): δ 10.20 (s, 1H), 8.11/8.05 (bs, 2H), 7.60/7.53 (bs, 2H), 7.13/
7.11 (bs, 2H). 13C NMR (101 MHz, DMSO-d6): δ 140.9, 137.4,
129.85/129.62, 118.95/118.92. This is a known compound.
Spectroscopic data agree with those reported in the literature.26,27
(S)-(3-Amino-3-carboxypropyl)dimethylsulfonium Iodide (12).29
To a flask containing a suspension of methionine (11, 60 g, 402
mmol) in distilled water (390 mL), iodomethane (64.0 mL, 145.9 g,
1.03 mol, 2.55 equiv) was added all at once. The resulting mixture was
vigorously stirred at 35−40 °C overnight. After 21.5 h, no more solid
material was left, and the biphasic mixture was evaporated under
reduced pressure to afford a yellow oil. This material was dissolved in
distilled water (160 mL). Addition of EtOH (580 mL) and a gentle
mixing resulted in a rapid formation of a white precipitate. This
suspension was allowed to stand overnight at rt, and the solid was
collected by filtration. After removing solvent residues under reduced
pressure, the product 12 was obtained as white solid (95.48 g, 327
mmol, 81%). It is recommended to store the obtained product in
darkness, as it develops a brownish color upon exposure to the light. Rf
= 0.19 (EtOH/AcOH/H2O = 3/1/1, [ninhydrin]). Mp 163−166 °C
(Lit.:29 163−165 C). 1H NMR (600 MHz, D2O): δ 3.92 (t, J = 6.5 Hz,
1H), 3.59−3.53 (m, 1H), 3.50−3.44 (m, 1H), 2.99 (s, 6H), 2.43−2.38
(m, 2H). 13C NMR (151 MHz, D2O): δ 172.4, 52.7, 39.4, 25.0, 24.8,
24.7. Mp 163−166 °C (Lit.:29 163−165 C). This is a known
compound. Spectroscopic data agree with those reported in the
literature.36
EXPERIMENTAL SECTION
■
General Experimental Methods. Solvents for column chroma-
tography and TLC (EtOAc, MeOH, CH2Cl2, pentane, Et2O) were of
technical grade. EtOAc, MeOH, and CH2Cl2 were distilled before use.
THF was of technical grade, purified by distillation and stored over
KOH. Afterward, it was distilled over SOLVONA and stored over
molecular sieves (4 Å). Dry CH2Cl2 was distilled over SOLVONA and
stored over molecular sieves (4 Å). MeCN (HPLC grade or extra dry)
and DMF (HPLC grade) were obtained commercially and used
without further purification.
Working Techniques. Reactions with moisture- or air-sensitive
compounds were carried out using the standard Schlenk technique. If
the reactions were carried out below rt, the following cooling baths
were employed: Dry ice/acetone (−20 to −30 °C), crushed ice/water
(0 to 2 °C). Glass stoppers and PTFE sleeves were used to seal
reaction vessels. The temperature of each reaction indicates the
temperature of the cooling or heating medium. Yields of the reactions
were calculated from the mass of purified and dried products. In the
cases of problematic purifications, yields and purities of intermediates
were estimated. For column chromatography, silica gel 60 (230−400
mesh) and neutral aluminum oxide 90 (70−230 mesh, activity stage I)
were used. In the advanced phase of our study, we found that
chromatographic purifications using aluminum oxide were not
necessary when we used eluents consisting of AcOH/MeOH/
CH2Cl2 and silica gel as stationary phase.
Analysis. Thin-layer chromatography was carried out with ready-
to-use glass plates coated with silica gel from a commerical source.
Detection was done both by irradiation of the plate with UV-light at
254 nm and by application of specific staining reagents. The ceric
ammonium molybdate stain (CAM) was prepared by dissolving 10 g
(NH4)6Mo7O24·4H2O and 0.2 g Ce(SO4)2·4H2O in 200 mL of 10%
H2SO4. The staining reagent KMnO4 was prepared by dissolving 3 g
KMnO4 and 10 g K2CO3 in 300 mL water followed by addition of 5
mL 5% aqueous NaOH. The staining reagent bromocresol green was
(S)-Homoserine (13).29 We recommend using an oil bath as a
heating medium in order to avoid extensive overheating of the reaction
mixture. We found an optimal temperature of the oil bath to be 140
°C. Our initial attempts to use a heating mantle without well-defined
temperature control led to dissatisfactory results: The reaction mixture
quickly developed a dark brown color, and the final product could not
be obtained in pure form. (S)-(3-Amino-3-carboxypropyl)dimethyl
sulfonium iodide (12, 51.4 g, 176 mmol) dissolved in distilled water
(81 mL) was heated to gentle reflux (140 °C) for 30 min. Within this
time, the pH dropped from 5.8 to 3.8, and dropwise addition of
NaHCO3 in distilled water (152 mL) was started. Addition was
performed under constant control of the pH value with the
appropriate pH indicator test paper (pH range 3.8 to 5.8). In order
to avoid the racemization, NaHCO3 solution was added until the pH
rose to ∼5.0. Reaction mixture was further refluxed until the pH
G
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
dropped to values 3.5−4.5, whereupon the base was added dropwise
again. Once the reaction was finished, water was removed under
reduced pressure to give an oil which solidified upon standing at rt.
The solid was dissolved in distilled water (45 mL) by gentle mixing
and careful heating with a heat gun. When acetone (103 mL) and
EtOH (1.28 L) were added to this solution, a white solid precipitated.
After removing all solvents under reduced pressure, the solid was
suspended in EtOH (500 mL), and after refluxing for 10 min filtered
off again. The precipitate was washed with cold EtOH (100 mL) and
dried in vacuo to give homoserine (13, 14.4 g, 121 mmol, 69%) as a
white solid. The product contained an unidentified impurity. Rf = 0.67
(EtOH/AcOH/H2O = 3/1/1, [ninhydrin]). 1H NMR (600 MHz,
D2O): δ 3.85 (dd, J = 7.5, 4.8 Hz, 1H), 3.81−3.74 (m, 2H), 2.19−2.12
(m, 1H), 2.06−1.99 (m, 1H). 13C NMR (151 MHz, D2O): δ 174.3,
58.6, 53.3, 32.0. [α]D21 °C = −8.3 (H2O, c = 1) (Lit.:37 − 8.0 [H2O, c =
1]). Mp 195 °C (dec) (Lit.:29 201−202 °C [dec]). This is a known
compound. Spectroscopic data agree with those reported in the
literature.29
d and at −20 °C for 2 d. The formed precipitate was collected by
filtration and washed with hexane (20 mL). After drying under
reduced pressure, the final product (9.84 g, 28.8 mmol, 83%) was
1
obtained as a white solid. H NMR (400 MHz, CDCl3) δ 7.47−7.44
(m, 4H), 7.32−7.14 (m, 6H), 6.00 (br d, J = 8.0 Hz, 1H), 5.00−4.94
(m, 1H), 4.09 (virt t, J = 6.4 Hz, 2H), 2.02 (s, 3H), 1.85−1.79 (m, 2
H), 1.76 (s, 3H). 13C NMR (101 MHz, CDCl3) δ 171.3, 170.9, 145.3,
144.8, 128.7, 128.5, 127.3, 127.2, 125.7, 125.6, 81.0, 61.9, 53.5, 29.5,
23.3, 21.2. This is a known compound. Spectroscopic data agree with
those reported in the literature.29
(3S)-3-(Acetylamino)-4,4-diphenylbutyl Acetate (17).29 To a
solution containing (3S)-3-(acetylamino)-4-hydroxy-4,4-diphenylbutyl
acetate (16) (9.84 g, 28.8 mmol) in glacial acetic acid (99 mL),
palladium on charcoal (10%, 1.33 g) and ammonium formate (9.18 g,
146 mmol, 5.0 equiv) were added. The resulting mixture was stirred at
110 °C for 32 h. It is highly recommended to monitor the reaction
progress by NMR spectroscopy, since complete conversion of the
starting material could not be sufficiently detected using TLC. The
reaction mixture was cooled to rt and filtered through a plug of Celite
which was additionally washed with a portion of CH2Cl2 (150 mL).
The solution was washed with saturated aqueous Na2CO3 (3 × 600
mL), brine (1 × 100 mL), and dried over MgSO4. After evaporation of
the solvents under reduced pressure, the product 17 (8.0 g, 24.6
mmol, 85%) was obtained as a yellow solid and used for subsequent
reactions without additional purification. Rf = 0.40 (Pentante/EtOAc =
(S)-2-Oxotetrahydrofuran-3-aminium Chloride (14).29 Homoser-
ine (13, 14.4 g, 121 mmol) was refluxed in an aqueous HCl (2 M, 235
mL) for 3 h (temperature of oil bath was kept at ca. 140 °C). The
mixture was stirred for further 17 h at rt. Subsequently, water was
removed by azeotropic distillation with ethanol. Boiling temperature of
the azeotrope was constantly kept below 79 °C by adding ethanol to
the mixture. After approximately 4−5 L of EtOH were distilled off and
the volume of the remaining brown solution was approximately 100
mL, it was cooled to rt and left to stand overnight. Filtration of the
brown precipitate and subsequent washing with ethanol afforded the
1
1/1, [UV, CAM]). H NMR (600 MHz, CDCl3) δ 7.26−7.22 (m,
8H), 7.19−7.12 (m, 2H), 5.26 (br d, J = 9.0 Hz, 1H), 4.89 (virt qd, J =
9.0, 3.0 Hz, 1H), 4.11−4.02 (m, 2H), 3.95 (d, J = 10.2 Hz, 1H), 1.99
(s, 3H) 1.92−1.86 (m, 1 H), 1.74 (s, 3H), 1.63−1.56 (m, 1 H). 13C
NMR (151 MHz, CDCl3) δ 171.2, 170.0, 142.0, 141.8, 129.0, 128.8,
128.32, 128.28, 127.1, 126.9, 61.7, 56.9, 49.2, 32.9, 23.4, 21.2. This is a
known compound. Spectroscopic data agree with those reported in the
literature.29
1
product 14 (11.83 g, 86 mmol, 71%) as a white powder. H NMR
(600 MHz, D2O): δ 4.64 (td, J = 9.2, 1.0 Hz, 1H), 4.51−4.44 (m, 2H),
2.86−2.80 (m, 1H), 2.51−2.41 (m, 1H). 13C NMR (151 MHz, D2O):
δ 174.4, 67.3, 48.5, 26.7. Mp 227 °C (dec) (Lit.:38 210−212 °C). This
is a known compound. Spectroscopic data agree with those reported in
the literature.38
(3S)-3-Amino-4,4-diphenyl-1-butanol (18).29 A suspension of
(3S)-3-(acetylamino)-4,4-diphenylbutyl acetate (17) (2.65 g, 8.14
mmol) in aqueous HCl (33 mL, 1.4 M) was stirred at 100 °C for 7 h.
The resulting solution was then cooled to rt and basified with solid
NaOH until pH 14 was reached. Upon addition of NaOH, an oily
precipitate was formed. The resulting emulsion was extracted with
CH2Cl2 (3 × 200 mL), and the combined organic layers were dried
over MgSO4. The solvent was removed under reduced pressure, and
the product 18 (1.81 g, 7.5 mmol, 92%) was obtained as brownish oil
which solidified upon standing at rt. Rf = 0.28 (CH2Cl2/MeOH = 9/1,
[UV, ninhydrin]). 1H NMR (600 MHz, CDCl3) δ 7.36−7.30 (m, 4H),
7.28−7.25 (m, 4H), 7.22−7.15 (m, 2H), 3.82−3.69 (m, 4 H), 2.55 (s
br, 3 H), 1.71−1.67 (m, 1 H), 1.49−1.43 (m, 1 H). 13C NMR (101
(S)-2-Amino-1,1-diphenylbutane-1,4-diol (15).29 Magnesium turn-
ings (4.34 g, 179 mmol, 4.9 equiv) were suspended in dry THF (70
mL) under Ar. To this suspension, a solution of bromobenzene (25.0
g, 159 mmol, 4.4 equiv) in dry THF (70 mL) was added dropwise
within 40 min. The reaction mixture was subsequently heated to reflux
for 1 h and then cooled to 2 °C. (S)-3-Aminodihydrofurane-2(3H)-on
hydrochloride (14, 5.0 g, 36.6 mmol) was added in small portions over
12 min. The mixture was stirred for another 18 min at the same
temperature and was then warmed to rt. After 19 h, the excess of
Grignard reagent was destroyed by the careful addition of ice. The
resulting mixture was cooled to 2 °C and aqueous HCl (2 M, 350 mL)
was carefully added. Subsequently, the solution was stirred at rt for 1 h.
After addition of aqueous ammonia (25%, 150 mL), the resulting
suspension was stirred for another 1 h at rt. Precipiate was filtered off
and washed with precooled EtOH (−20 °C) to afford the desired
product (5.97 g, 23.2 mmol, 64%) as an off-white solid. 1H NMR (300
MHz, DMSO-d6) δ 7.58 (d, J = 7.4 Hz, 2H), 7.49 (d, J = 7.4 Hz, 2H),
7.37−7.06 (m, 6H), 5.33 (bs, 1H), 3.89 (dd, J = 9.3, 2.9 Hz, 1H),
3.64−3.43 (m, 2H), 1.77−1.03 (m, 3H). 13C NMR (151 MHz,
DMSO-d6) δ 147.3, 146.4, 128.0, 127.8, 126.02, 125.98, 125.8, 125.5,
80.0, 59.6, 55.0, 33.7. [α]D22 °C = −28.1 (EtOH, c = 0.26) (Lit.:29 −26.0
[EtOH, c = 0.26]). Mp 198−199 °C (Lit.:29 200 °C). This is a known
compound. Spectroscopic data agree with those reported in the
literature.29
MHz, CDCl3) δ 142.7, 142.0, 129.3, 129.0, 128.4, 128.1, 127.2, 126.9,
62.7, 60.9, 56.0, 35.4. [α]D21
°
= +15.5 (CH2Cl2, c = 1). This is a
C
known compound. Spectroscopic data agree with those reported in the
literature.29
(2S)-4-[tert-Butyl(diphenyl)silyl]oxy-1,1-diphenyl-2-butanamine
(19). Aminoalcohol 18 (4.42 g, 16.4 mmol) and imidazole (3.74 g,
54.9 mmol, 3.3 equiv) were dissolved in MeCN (93 mL, HPLC-
garde). After addition of TBDPSCl (9.6 mL, 10.1 g, 37 mmol, 2.2
equiv), the resulting solution was stirred at rt under Ar for 2.5 h. The
solvent was removed under reduced pressure, and the crude product
(gummy mass) was purified by flash column chromatography (Si, 7.0
× 24 cm, CH2Cl2/MeOH = 99/1 → 98/2 → 97/3 → 95/5). The
product 19 (7.41 g, 15.4 mmol, 94%) was obtained as brownish
(3S)-3-(Acetylamino)-4-hydroxy-4,4-diphenylbutyl Acetate
(16).29 Acetic anhydride (15.3 g, 14.2 mL, 149.9 mmol, 4.3 equiv)
was slowly added to a suspension containing the diol 15 (8.88 g, 34.5
mmol) and pyridine (69.7 g, 71 mL, 881.4 mmol, 25.5 equiv), and the
resulting solution was stirred at rt for 23 h. The reaction mixture was
poured into a separatory funnel containing aqueous HCl (2 M, 560
mL), and the resulting emulsion was extracted with CH2Cl2 (3 × 450
mL). The combined organic layers were dried over MgSO4 and
concentrated under reduced pressure. The resulting crude product was
suspended in hexane (132 mL) and heated to reflux. EtOAc (ca. 70
mL) was added in small portions until the solid material dissolved.
After the solution was slowly cooled to rt, it was stored at +7 °C for 8
1
viscous oil. Rf = 0.42 (CH2Cl2/MeOH = 9/1, [UV, ninhydrin]). H
NMR (600 MHz, CDCl3) δ 7.68 (dd, J = 8.4, 1.2 Hz, 2H), 7.64 (dd, J
= 8.4, 1.2 Hz, 2H), 7.45−7.41 (m, 2H), 7.40−7.35 (m, 6H), 7.33−7.24
(m, 6H), 7.21−7.16 (m, 2H), 3.89−3.83 (m, 2 H), 3.78−3.75 (m,
1H), 3.72 (d, J = 10.2 Hz, 1H), 1.83−31.78 (m, 1H), 1.55 (br s, 2H),
1.44−1.37 (m, 1H), 1.08 (s, 9H). 13C NMR (151 MHz, CDCl3) δ
143.3, 143.1, 135.78, 135.76, 133.9, 133.8, 129.84, 129.80, 128.9,
128.8, 128.5, 128.4, 127.9, 127.8, 126.7, 126.5, 62.3, 60.3, 52.1, 37.8,
̃
27.1, 19.4. [α]2D1 °C = +3.9 (CHCl3, c = 1). IR (ATR) ν = 3386 cm−1
(w), 3065 (w), 2931 (m), 1593 (w), 1491 (w), 1427 (w), 1256 (w),
1104 (s), 936 (w), 722 (s), 740 (s). MS (EI, 70 eV) m/z (%) 480.6
H
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
(11) [M+], 422.4 (76) [(M − t-Bu)+], 312.4 (82) [(M − CHPh2)+],
(4 × 90 mL), and the combined organic layers were washed with H2O
(2 × 150 mL) and brine (150 mL). After drying over Na2SO4 and
evaporation of the solvent under reduced pressure, the desired product
24 (7.93 g, 25.3 mmol, 87%) was obtained as yellowish oil. Rf = 0.26
(CH2Cl2/MeOH = 9/1, [UV, ninhydrin]). 1H NMR (400 MHz,
CDCl3): δ 7.69−7.64 (m, 4H), 7.45−7.35 (m, 6H), 3.75 (t, J = 6.0 Hz,
2H), 2.84 (t, J = 6.7 Hz, 2H), 1.74−1.66 (m, 2H), 1.30 (bs, 2H), 1.05
(s, 9H). 13C NMR (101 MHz, CDCl3): δ 135.5, 133.8, 129.6, 127.6,
+
234.3 (100), 198.2 (54), 167.2 (43) [CHPh2 ], 135.2 (40). MS (ESI,
Orbitrap) m/z 502.2 (M + Na+), 480.3 (M + H+). HRMS (ESI,
Orbitrap) calcd for C32H38NOSi+ (M + H+): 480.27171; found:
480.27063.
(S)-tert-Leucinol (21).30 A 2 L three-necked flask was charged with
dry THF (400 mL) and NaBH4 (19.0 g, 502.4 mmol, 2.5 equiv).
Subsequently, L-tert-leucine (20, 27.4 g, 209 mmol) was added in small
portions, and weak gas evolution was observed. After cooling the
resulting suspension in an ice/H2O bath, a solution of iodine (53.0 g,
209 mmol, 1.0 equiv) in dry THF (100 mL) was added dropwise
resulting in vigorous gas evolution. When the addition of iodine was
complete, the resulting white suspension was stirred at 100 °C for 19 h
(overnight). Afterward, the reaction mixture was cooled to rt, and
MeOH was added dropwise until no more gas evolution was observed
and clear solution was obtained. All solvents were removed under
reduced pressure to afford a white paste. The resulting crude material
was taken up with 20% aqueous KOH solution (415 mL) and stirred
at rt for 4 h. This mixture was subsequently extracted with CH2Cl2 (4
× 250 mL). Combined organic layers were dried over Na2SO4 and
concentrated under reduced pressure. The resulting oily residue was
purified by Kugelrohr distillation affording the final product 21 (22.57
g, 193 mmol, 92%) as a colorless jelly-like substance solidifying at +7
°C. The product contained an unknown impurity. Rf = 0.28 (CH2Cl2/
MeOH = 9/1, [ninhydrin]). 1H NMR (600 MHz, CDCl3): δ 3.64 (dd,
J = 10.5, 3.6 Hz, 1H), 3.16 (dd, J = 10.5, 9.8 Hz, 1H), 2.45 (dd, J = 9.8,
̃
61.9, 39.3, 36.2, 26.9, 19.1.IR: ν = 3059 (m), 2937 (vs), 2864 (s), 1589
(w), 1468 (m), 1429 (m), 1385 (m), 1102 (vs), 818 (m), 704 (s), 612
(m), 503 (s). MS (ESI): m/z = 314.19 [M + H+], 236.15 [(M −
Ph)+]. HRMS (ESI, Orbitrap): calcd for C19H28NOSi+ [M + H+]:
314.19347; found: 314.19357. Spectroscopic data do not agree with
those reported in the literature.40 We assume that the compound
reported in the literature is the hydrochloride salt of amine 24.
(S)-N-(1-((tert-butyldiphenylsilyl)oxy)-3,3-dimethylbutan-2-yl)-
1H-imidazole-1-carboximidamide (25). Guanylating reagent 6 (45.3
mg, 281 μmol) was added to a flask containing a solution of
aminoalcohol 22 (120 mg, 337 μmol, 1.2 equiv) in THF (reagent
grade, 1.8 mL). The reaction mixture was stirred at 40 °C for 120 h in
an open flask, so that a slow evaporation of the solvent and
concentration of reaction mixture was achieved. Afterward, the
remaining solvent was removed in vacuo, and the crude product was
purified by flash column chromatography (Si, 1.8 × 19 cm, pentane/
EtOAc = 1/4) affording the product 25 (78.8 mg, 175 μnol, 62%) as a
white foam. The final product contained the residual ethyl acetate and
some unidentified impurities (estimated purity ca. 90%). Rf = 0.45 (P/
EtOAc = 1/4 [UV, bromocresol green]). Due to the very low
concentration of the sample, some signals cannot be seen in the 13C
NMR spectrum. 1H NMR (600 MHz, CD3OD) δ 8.16 (bs, 1H), 7.69
(dd, J = 8.0, 1.4 Hz, 2H), 7.67 (dd, J = 8.0, 1.4 Hz, 2H), 7.58 (bs, 1H),
7.45−7.32 (m, 6H), 7.05 (bs, 1H), 3.99 (dd, J = 10.0, 2.9 Hz, 1H),
3.69−3.63 (bm, 1H), 3.49−3.37 (bm, 1H), 1.01 (s, 9H), 0.89 (s, 9H).
13C NMR (151 MHz, CD3OD) δ 136.72, 136.70, 130.9, 130.8, 128.8,
3.6 Hz, 1H), 2.30 (bs, 3H), 0.82 (s, 9H). 13C NMR (151 MHz,
CDCl3): δ 62.3, 61.6, 33.0, 26.2. [α]D25
°
= +36.2 (CH3Cl, c = 1)
C
(Lit.:39 [α]D25 °C = +38.2 [CH3Cl, c = 1]). This is a known compound.
Spectroscopic data agree with those reported in the literature.39
(2S)-1-[tert-Butyl(diphenyl)silyl]oxy-3,3-dimethyl-2-butanamine
(22). Aminoalcohol 21 (3.05 g, 26 mmol) and imidazole (5.31 g, 78
mmol, 3.0 equiv) were dissolved in MeCN (109 mL, HPLC-grade).
After addition of TBDPSCl (13.5 mL, 14.3 g, 52 mmol, 2.0 equiv), the
resulting solution was stirred at rt under Ar for 2 h. The solvent was
removed under reduced pressure, and the crude product (gummy
mass) was purified by flash column chromatography (Si, 5.0 × 20 cm,
CH2Cl2/MeOH = 97/3 → 95/5). The final product 22 (4.08 g, 11.5
mmol, 44%) was obtained as a colorless oil. The corresponding
hydrochloride salt was isolated as a slightly more polar fraction. This
HCl salt was dissolved in CH2Cl2 (150 mL) and washed with saturated
aq. NaHCO3 (3 × 150 mL) solution. The combined aqueous phases
were extracted with CH2Cl2 (50 mL). The combined organic phases
were dried over MgSO4, and the solvent was removed under reduced
pressure to give an additional portion of the product 22 (3.29 g, 9.2
mmol, 36%). The total yield of 22 was 80% (7.37 g, 20.7 mmol). Rf =
128.7, 118.8, 35.2, 27.3, 19.9. MS (ESI): m/z = 471.25 [M + Na+],
449.27 [M + H+], 371.22 [(M − Ph)+]. HRMS (ESI, Orbitrap): calcd
for C26H37N4OSi+ [M + H+]: 449.27311; found: 449.27155. Due to
the fact that this intermediate was not used for the synthesis of the
target guanidine, IR spectroscopic data and optical rotation were not
collected.
N-(3-((tert-Butyldiphenylsilyl)oxy)propyl)-1H-imidazole-1-carbox-
imidamide (26). Guanylating reagent 6 (1.27 g, 7.87 mmol) was
added to a flask containing a solution of aminoalcohol 24 (2.96 g, 9.44
mmol, 1.2 equiv) in THF (reagent grade, 36 mL). The reaction
mixture was stirred at 40 °C for 23 h in an open flask to allow a slow
evaporation of the solvent and concentration of the reaction mixture.
Afterward, the remaining solvent was removed in vacuo, and the crude
product was purified by column chromatography (Si, 4.8 × 18.5 cm,
CH2Cl2/MeOH = 92/8) affording the desired product (2.46 g, 6.05
mmol, 77%) as a white foam. The final product contained some
unidentified impurities (estimated purity ca. 95%). Rf = 0.37 (CH2Cl2/
MeOH = 9/1 [UV, bromocresol green]). 1H NMR (600 MHz,
CD3OD) δ 8.08 (bs, 1H), 7.71−7.64 (m, 4H), 7.49 (bs, 1H), 7.40−
7.30 (m, 6H), 7.01 (bs, 1H), 3.81 (t, J = 6.3 Hz, 2H), 3.38 (t, J = 6.3
Hz, 2H), 1.91 (quint, J = 6.3 Hz, 2H), 1.05 (s, 9H). 13C NMR (151
MHz, CD3OD) δ 149.3, 136.7, 136.6, 134.7, 130.8, 129.3, 128.7, 118.5,
1
0.5 (CH2Cl2/MeOH = 9/1, [UV, ninhydrin]). H NMR (400 MHz,
CDCl3) δ 7.70−7.66 (m, 4H), 7.44−7.35 (m, 6H), 3.79 (dd, J = 9.6,
3.2 Hz, 1H), 3.45 (virt t, J = 9.6 Hz, 1H), 2.63 (dd, J = 8.8, 2.8 Hz,
1H), 1.53 (br s, 2H), 1.06 (s, 9H), 0.82 (s, 9H). 13C NMR (151 MHz,
CDCl3) δ 135.74, 135.72, 133.8, 133.7, 129.83, 129.80, 127.85, 127.83,
66.0, 61.8, 33.0, 27.0, 26.7, 19.4. [α]2D1 °C = +18.4 (CH3Cl, c = 1). IR
(ATR) ν
̃
= 3069 cm−1 (w), 2952 (s), 2861 (s), 2329 (w), 2111 (w),
1588 (w), 1470 (s), 1427 (m), 1391 (m), 1361 (m), 1189 (w), 1106
(s), 1171 (s), 999 (m), 938 (w), 820 (s), 735 (s), 700 (s). MS (EI, 70
eV) m/z (%) 356.4 (30) [M+], 298.3 (100) [(M − t-Bu)+], 198.2
(77), 178.2 (12), 135.1 (15), 86.2 (52), 57.3 (23). MS (ESI, Orbitrap)
m/z 733.4 [2M + Na+], 711.5 [2M + H+], 378.2 [M + Na+], 356.2 [M
+ H+], 278.2 [(M − Ph)+]. HRMS (ESI, Orbitrap) calcd for
C22H34NOSi+ [M + H+]: 356.24042; found: 356.24023.
̃
62.4, 41.9, 33.4, 27.5, 20.0. IR: ν = 3015 (m), 2968 (m), 2168 (w),
1979 (w), 1739 (vs), 1437 (m), 1366 (s), 1217 (s), 1097 (m), 909
(w), 739 (w), 696 (w). MS (ESI): m/z = 407.22 [M + H+], 329.18
[(M − Ph)+]. HRMS (ESI, Orbitrap): calcd for C23H31N4OSi+ [M +
H+]: 407.22616; found: 407.22485.
3-((tert-Butyldiphenylsilyl)oxy)propan-1-amine (24). TBDPSCl
(7.57 mL, 8.0 g, 29.1 mmol) was added to a solution of 3-amino-1-
propanol 23 (2.66 mL, 2.62 g, 34.9 mmol, 1.2 equiv) in MeCN
(HPLC-Grade, 160 mL), and the resulting solution was stirred under
Ar for 45 h. The solvent was removed in vacuo and the resulting crude
mixture purified by flash column chromatography (Si, 7.5 × 16 cm,
CH2Cl2/MeOH = 9/1), affording the hydrochloride salt of the desired
product as a yellowish solid. This material was dissolved in H2O (200
mL), and the resulting solution titrated with aqueous NaOH (1 M)
until pH > 10 was reached. This solution was extracted with CH2Cl2
(S)-6-(tert-Butyl)-2,2,15,15-tetramethyl-3,3,14,14-tetraphenyl-
4,13-dioxa-7,9-diaza-3,14-disilahexadecan-8-iminium Trifluoroace-
tate/Hydroxide (27·TFA/27·H2O).41 To a flask containing a solution
of aminoalcohol 22 (550 mg, 1.54 mmol) in 20 mL CH2Cl2, the
unsymmetrical guanylating reagent 26 (1.26 g, 3.09 mmol, 2.0 equiv)
was added. After stirring the resulting mixture for 5 min, the solvent
was removed under reduced pressure, and TFA (115 μL, 176 mg, 1.54
mmol, 1.0 equiv) was added. The resulting mass was mixed in a
Kugelrohr oven at 115 °C for 2 h under Ar. The resulting crude
I
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
gummy mass was purified by column chromatography (Si, 3.0 × 19
cm, CH2Cl2/MeOH = 97/3) to remove the impurities which were
more polar than the obtained product. This prepurified mixture was
subjected to another chromatographic purification (Al neutral, 3.0 ×
19 cm, CH2Cl2/MeOH = 97/3) to separate the final product from
unreacted aminoalcohol 22. The product 27·TFA (687 mg, 850 μmol,
55%) was obtained as an off-white foam and contained residual
CH2Cl2 and some unidentified impurities (estimated purity ca. 95%).
The following analytical data are given for the trifluoroacetate salt
27·TFA. Rf = 0.18 (Si, CH2Cl2/MeOH = 97/3 [UV, bromocresol
green]) Rf = 0.23 (Al, CH2Cl2/MeOH = 97/3 [UV, bromocresol
CD3OD) δ = 3.97 (dd, J = 11.0, 2.8 Hz, 1H), 3.72 (dd, J = 11.0, 2.8
Hz, 1H), 3.67 (t, J = 6.3 Hz, 2H), 3.59 (dd, J = 11.0 Hz, 1H), 3.47−
3.39 (m, 2H), 2.07 (quint, J = 6.3 Hz, 2H), 1.01 (s, 9H). 13C NMR
(151 MHz, CD3OD) δ 158.7, 63.8, 46.1, 42.5, 40.0, 36.6, 32.7, 26.6.
19F NMR (564 MHz, CD3OD) δ −154.2 (s, 1F). [α]2D9
°
= −9.2
C
̃
(MeOH, c = 1). IR: ν = 2963 (m), 1633 (vs), 1446 (m), 1372 (m),
1284 (w), 1219 (w), 1001 (w), 728 (m). MS (ESI): m/z = 254.12 [M
+
− F−]. HRMS (ESI, Orbitrap): calcd for C10H22N3Cl2 [M − F−]:
254.11853; found: 254.11768.
(S)-3-(tert-Butyl)-1,2,3,5,6,7-hexahydroimidazo[1,2-a]pyrimidin-
8-ium Chloride (30).45 To a flask containing a suspension of crude
guanidinium salt 29 (855 μmol considering quantitative yield of the
previous step) in dry MeCN (9.8 mL), DBU (511 μL, 521 mg, 3.4
mmol, 4 equiv) was added. The resulting dark brown solution was
stirred under Ar at rt overnight. The reaction mixture was acidified
with HCl solution (1.25 M in MeOH) to pH 1−2, and the solvents
were removed under reduced pressure. The resulting crude product
was purified by column chromatography (Si, 3 × 16.5 cm, EtOAc/
MeOH = 7/3). The isolated material (product contaminated with
silica particles) was suspended in MeCN, and the silica was filtered off.
Solvent was evaporated to give a yellowish oil (125 mg, 574 μmol, 67%
over 2 steps) which consisted of two regioisomers 30 and 7 (30/7 =
1
green]) H NMR (600 MHz, CD3OD) δ 7.71−7.64 (m, 8H), 7.45−
7.36 (m, 12H), 3.92 (dd, J = 10.2, 2.7 Hz, 1H), 3.76 (t, J = 5.7 Hz,
2H), 3.63−3.59 (m, 1H), 3.58−3.55 (m, 1H), 3.45−3.36 (m, 2H),
1.85 (m, 2H), 1.05 (s, 9H), 1.04 (s, 9H), 0.92 (s, 9H). 13C NMR (151
MHz, CD3OD) δ 158.7, 136.63, 136.61, 136.57, 134.5, 133.98, 133.95,
131.1, 131.0, 128.99, 128.95, 128.87, 64.8, 63.8, 62.0, 39.7, 34.7, 32.9,
27.4, 27.3, 26.9, 20.0, 19.9. 19F NMR (564 MHz, CD3OD) δ −76.7
(s).
The optical rotation is given for the corresponding hydroxide salt of
27 (27·H2O).42 [α]D29 °C = −46.7 (CH3Cl, c = 1). IR: ν
̃
= 3060 (w),
2948 (m), 2864 (m), 1739 (m), 1640 (m), 1468 (m), 1427 (m), 1366
(m), 1202 (m), 1102 (vs), 1001 (m), 818 (m), 697 (vs). MS (ESI):
m/z = 694.42 [(M − CF3COO−)+], 616.38 [(M − CF3COO− − H+ −
1
92/8, determined by H NMR). Rf = 0.18 (EtOAc/MeOH = 7/3
[bromocresol green, KMnO4]). [α]2D9 °C = −12.6 (MeOH, c = 0.27).
+
Ph)+]. HRMS (ESI, Orbitrap): calcd for C42H60N3O2Si2 [(M −
̃
IR: ν = 2953 (m), 1668 (vs), 1612 (vs), 1472 (m), 1373 (w), 1323
CF3COO−)+]: 694.42186; found: 694.42377.
(w), 1285 (w), 1190 (w). MS (ESI): m/z = 182.16 [M − Cl−], 124.09
+
[(M − Cl−−C(CH3)3)+]. HRMS (ESI, Orbitrap): calcd for C10H20N3
(S)-((1-Hydroxy-3,3-dimethylbutan-2-yl)amino)((3-hydroxy-
propyl)amino)methaniminium Fluoride (28).43 To a flask containing
a solution of protected guanidine 27·TFA (685 mg, 848 μmol) in
absolute MeCN (16 mL), fluoride on polymer (848 mg, 2.54 mmol, 3
equiv; 3.0 mmol/g loading on Amberlite IRA-900 resin; obtained from
a commerical source) was added, and the resulting mixture was stirred
under an Ar atmosphere at rt overnight. During the course of this
reaction, temporary clotting of the polymer beads was observed.
Subsequently, fluoride on polymer beads were filtered off and washed
with MeOH, and all solvents were removed under reduced pressure.
The residue was dissolved in MeOH (3 mL). Upon dropwise addition
of Et2O to this solution, the product 28 precipitated as a heavy
colorless oil. Addition of Et2O (ca. 3−4 mL) was continued until no
more oil formation was observed. The resulting mixture was gently
mixed on rotary evaporator for 1 h resulting in even distribution of the
precipitate on the wall of the flask. After subsequent decantation of the
solvent, washing of the oily residue with Et2O (4 × 20 mL) and drying
under reduced pressure, the guanidinium salt 28 was obtained as a
colorless syrup (191 mg, 805 μmol, 95%). Isolated product was
obtained in ca. 95% purity and contained traces of solvents and small
amounts of the starting material 27·TFA or the corresponding
monodeprotected intermediates. Rf = 0.00 (CH2Cl2/MeOH = 9/1
[M − Cl−]: 182.16517; found: 182.16415.
1
NMR data of the regioisomer 30 (major isomer): H NMR (600
MHz, CD3OD) δ 3.78−3.68 (m, 2H), 3.51 (dd, J = 9.2, 6.7 Hz, 1H),
3.36−3.33 (m, 4H), 2.09−1.97 (m, 2H), 0.92 (s, 9H). 13C NMR (151
MHz, CD3OD) δ 156.8, 63.9, 51.1, 42.8, 39.2, 34.4, 25.3, 21.4.
NMR data of the minor regioisomer, which is supposed to be the
salt 7, is presented. This regioisomer was not isolated in pure form,
and therefore only characteristic signals of the t-butyl rest are given. 1H
NMR (600 MHz, CD3OD) δ 1.02 (s, 9H). 13C NMR (151 MHz,
CD3OD) δ 26.3.
(S)-1-(1-Chloro-3,3-dimethylbutan-2-yl)tetrahydropyrimidin-
2(1H)-iminium Chloride (31). Monocyclized guanidinium salt 31 (63
mg, 248 μmol, 29% over 2 steps) was isolated as a minor reaction
product as an off-white solid. The structure and relevant signal
assignments were performed using HSQC and HMBC (see
Supporting Information). Rf = 0.37 (EtOAc/MeOH = 7/3
[bromocresol green, KMnO4]). 1H NMR (600 MHz, CD3OD) δ
3.95 (dd, J = 10.7, 2.0 Hz, 1H, CHHCl), 3.60−3.51 (m, 2H, NCHtBu
and CHHCl), 3.40−3.37 (m, 4H, 2 × CH2N), 1.99−1.93 (m, 2H,
CH2CH2CH2), 1.00 (s, 9H). 13C NMR (151 MHz, CD3OD) δ 155.5,
63.2 (CH), 46.2 (CH2), 39.8 (CH2), 36.6 (CH2), 26.6, 21.3 (CH2).
1
[bromocresol green]). H NMR (600 MHz, CD3OD) δ 3.86 (dd, J =
̃
IR: ν = 2952 (m), 2865 (m), 1644 (vs), 1460 (m), 1376 (w), 1318
11.2, 3.1 Hz, 1H), 3.67−3.63 (m, 2H), 3.47 (dd, J = 11.2, 9.1 Hz, 1H),
(w), 1203 (m), 1118 (w), 725 (m). MS (ESI): m/z = 220.14 [M
(37Cl) − Cl−], 218.14 [M (35Cl) − Cl−], 182.16 [(M − 2Cl− − H+)+].
HRMS (ESI, Orbitrap): calcd for C10H21N3Cl+ [M (35Cl) − Cl−]:
218.14185; found: 218.14130.
3.38−3.33 (m, 3H), 1.83−1.77 (m, 2H), 0.98 (s, 9H). 13C NMR (151
MHz, CD3OD) δ 159.13, 64.3, 62.6, 59.5, 39.6, 35.0, 32.6, 27.0. 19F
C
NMR (564 MHz, CD3OD) δ −139.6 (s, 1F). [α]D29
°
= −46.7
̃
(MeOH, c = 1). IR: ν = 3432 (b, vs), 2922 (s), 2856 (m), 1640 (s),
(S)-7-Benzhydryl-2,2,16,16-tetramethyl-3,3,15,15-tetraphenyl-
4,14-dioxa-8,10-diaza-3,15-disilaheptadecan-9-iminium 2,2,2-tri-
fluoroacetate (32).41 To a flask containing a solution of aminoalcohol
19 (470 mg, 980 μmol) in CH2Cl2 (20 mL), the unsymmetrical
guanylating reagent 26 (797 mg, 1.96 mmol, 2.0 equiv) was added.
After stirring the resulting mixture for 5 min, the solvent was removed
under reduced pressure, and TFA (72.8 μL, 112 mg, 980 μmol, 1.0
equiv) was added. The resulting mass was mixed in a Kugelrohr oven
at 115 °C for 2 h under Ar. The resulting crude gummy mass was
purified by column chromatography (Si, 3.0 × 18.5 cm, CH2Cl2/
MeOH = 97/3) to remove the impurities, which were more polar than
the obtained product. This prepurified mixture was subjected to
another chromatographic purification (Al neutral, 3.0 × 19.5 cm,
CH2Cl2/MeOH = 97/3) to separate the final product from unreacted
aminoalcohol 19. The product 32 (355 mg, 381 μmol, 39%) was
obtained as a white foam. Rf = 0.37 (Si, CH2Cl2/MeOH = 97/3 [UV,
bromocresol green]) Rf = 0.43 (Al, CH2Cl2/MeOH = 97/3 [UV,
1448 (vs), 1384 (vs), 1234 (m), 1059 (s), 875 (w), 668 (m). MS
(ESI): m/z = 240.17 [(M − F− − H+ + Na+)+], 218.19 [(M − F−)+].
+
HRMS (ESI, Orbitrap): calcd for C10H24N3O2 [(M − F−)+]:
218.18630; found: 218.18666.
(S)-((1-Chloro-3-methylbutan-2-yl)amino)((3-chloropropyl)-
amino)methaniminium Fluoride (29).44 To a flask containing a
solution of guanidinium salt 28 (203 mg, 855 μmol) in absolute
MeCN (16 mL), thionyl chloride (374 μL, 611 mg, 5.1 mmol, 6
equiv) was added. The solution was stirred at 60 °C for 21 h. After
addition of MeOH (20 mL), the resulting mixture was stirred for 40
min at rt. The solvents were removed under reduced pressure to give
the crude product 29 as a yellowish solid. The crude product
contained unknown impurities. It was characterized and subjected to
the next reaction step without purification. The two-step yield was
determined after the subsequent cyclization step. Rf = 0.18 (CH2Cl2/
1
MeOH = 9/1 [bromocresol green, KMnO4]) H NMR (600 MHz,
J
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
bromocresol green]). A significant signal broadening was observed in
the H NMR spectrum. A possible explanation for this effect is the
(101 MHz, CD3OD): δ 157.4, 143.0, 142.4, 130.1, 129.6, 129.5, 129.2,
1
C
128.2, 128.1, 59.4, 54.8, 42.4, 42.0, 39.6, 37.8, 32.6. [α]D31
°
= +57.2
aggregate formation or hindered rotation due to steric bulk of
̃
(MeOH, c = 2). IR: ν = 3162 (m), 2960 (w), 2876 (w), 1634 (vs),
1
protecting groups. H NMR (400 MHz, CD3OD): δ 7.75−7.02 (m,
1494 (m), 1449 (m), 1290 (m), 1123 (m), 750 (m), 700 (vs). MS
(ESI): m/z = 382.14 [M (2 × 37Cl) − Cl−], 380.15 [M (35Cl, 37Cl) −
Cl−], 378.15 [M (2 × 35Cl) − Cl−]. HRMS (ESI, Orbitrap): calcd for
30H), 4.61−4.50 (bm, 1H), 4.07 (bd, J = 10.8 Hz, 1H), 3.83−3.73
(bm, 1H), 3.72−3.64 (bm, 1H), 3.59−3.51 (bm, 2H), 3.17−2.97 (bm,
2H), 1.90−1.78 (bm, 1H), 1.65−1.42 (bm, 3H), 1.03 (bs, 18H). 13C
NMR (151 MHz, CD3OD): δ 157.7, 143.1, 142.7, 136.5, 136.44,
136.36, 134.4, 133.9, 133.8, 131.0, 130.9, 130.0, 129.6, 129.14, 129.09,
128.9, 128.83, 128.81, 128.0, 127.8, 62.6, 61.8, 58.6, 54.3, 39.2, 37.2,
+
C20H26N3Cl2 [M (2 × 35Cl) − Cl−]: 378.14983; found: 378.14871.
(S)-2-Benzhydryl-2,3,4,6,7,8-hexahydro-1H-pyrimido[1,2-a]-
pyrimidin-9-ium Chloride (8).45 To a flask containing a solution of
crude dichloride intermediate (61 mg, ca. 147 μmol) in dry CH2Cl2
(4.7 mL), DBU (91.3 μL, 93.1 mg, 612 μmol, 4.2 equiv) was added.
The resulting solution was stirred at rt for 3 h. The reaction mixture
was acidified with HCl solution (1.25 M in MeOH), and the solvents
were removed under reduced pressure. The resulting crude product
was purified by column chromatography (Si, 1.8 cm × 14.5 cm,
EtOAc/MeOH = 7/3). After evaporation of solvents, the isolated
material (product containing silica particles) was suspended in MeCN,
and silica was filtered off. Solvent was evaporated to give the bicyclic
guanidinium salt 8 (22 mg, 64 μmol, 44%) as a white foam. Crystals
suitable for X-ray analysis were obtained by dissolving this material in
MeCN followed by slow evaporation at rt over several days.32 Rf = 0.32
32.4, 27.5, 27.4, 19.92, 19.87. 19F NMR (564 MHz, CD3OD): δ −76.3
(s, 3F). [α]2D9
°
= +28.1 (CHCl3, c = 2.5). IR: ν
̃
= 3058 (w), 2938
C
(m), 2863 (m), 1677 (m), 1593 (m), 1465 (m), 1427 (m), 1379 (w),
1189 (m), 1100 (vs), 999 (m), 820 (m), 697 (m). MS (ESI): m/z =
818.45 [M − CF3COO−]. HRMS (ESI, Orbitrap): calcd for
+
C52H64N3O2Si2 [M − CF3COO−]: 818.45316; found: 818.44891.
(S)-((4-Hydroxy-1,1-diphenylbutan-2-yl)amino)((3-hydroxy-
propyl)amino)methaniminium Fluoride (33).43 To a flask containing
a solution of protected guanidine 32 (344 mg, 370 μmol) in absolute
MeCN (7.2 mL), fluoride on polymer (370 mg, 1.11 mmol, 3 equiv;
3.0 mmol/g loading on Amberlite IRA-900 resin; obtained from a
commerical source) was added, and the resulting mixture was stirred
under an Ar atmosphere at rt overnight. During the course of this
reaction, temporary clotting of the polymer beads was observed.
Subsequently, fluoride on polymer beads was filtered off and washed
with MeOH (10 mL), and all solvents were removed under reduced
pressure. The residue was dissolved in 1.5 mL MeOH (1.5 mL). Upon
dropwise addition of Et2O to this solution, the product 33 precipitated
as a heavy colorless oil. Addition of Et2O was continued until no more
oil formation was observed. The resulting mixture was gently mixed on
rotary evaporator for 1 h resulting in even distribution of the
precipitate on the wall of the flask. After subsequent decantation of the
solvent, washing of the oily residue with Et2O (3 × 15 mL) and drying
under reduced pressure, the guanidinium salt 33 was obtained as a
white foam (132 mg). Isolated product was obtained in ca. 95% purity
and contained traces of solvents and small amounts of the starting
material 32 or the corresponding monodeprotected intermediates.
Taking this into account, the estimated amount of the product was
approximately 125 mg (347 μmol, 95%). Rf = 0.00 (CH2Cl2/MeOH =
9/1 [UV, bromocresol green]). A significant signal broadening was
1
(EtOAc/MeOH = 7/3 [UV, bromocresol green, KMnO4]). H NMR
(600 MHz, CD3OD): δ 7.47 (d, J = 7.3 Hz, 2H), 7.39−7.34 (m, 4H),
7.31 (t, J = 7.7 Hz, 2H), 7.26 (t, J = 7.3 Hz, 1H), 7.21 (t, J = 7.3 Hz,
1H), 4.38 (ddd, J = 11.2, 7.8, 3.6 Hz, 1H), 3.96 (d, J = 11.2 Hz, 1H),
3.39−3.36 (m, 4H), 3.29−3.21 (m, 2H), 2.03−1.96 (m, 2H), 1.89−
1.83 (m, 1H), 1.79−1.72 (m, 1H). 13C NMR (151 MHz, CD3OD): δ
152.3, 142.5, 141.9, 130.2, 130.0, 129.4, 129.0, 128.4, 128.2, 58.0, 52.6,
47.7, 46.6, 39.0, 26.3, 21.5. [α]2D1 °C = +5.5 (MeOH, c = 0.47). IR: ν
̃
=
3029 (s), 2943 (s), 1630 (vs), 1520 (m), 1495 (m), 1449 (m), 1320
(m), 1082 (m), 755 (m), 704 (s). MS (ESI): m/z = 306.20 [M −
Cl−], 138.10 [(M − H+ − Cl− − CHPh2)+]. HRMS (ESI, Orbitrap):
+
calcd for C20H24N3 [M − Cl−]: 306.19647; found: 306.19604.
Additionally, a mixture consisting of the desired product 8 and
regioisomer 35 (8/35 = 3/1 based on 1H NMR analysis) was isolated.
Compound 35 could not be obtained in pure form, and thus only
partial assignment of its signals in NMR spectra was possible. Rf = 0.40
1
(EtOAc/MeOH = 7/3 [UV, bromocresol green, KMnO4]) H NMR
(600 MHz, CD3OD): δ 4.19 (d, J = 11.5 Hz, 1H), 3.19−3.13 (m, 2H),
3.03−2.97 (m, 1H), 2.26 (ddd, J = 12.2, 7.5, 11.5 Hz, 1H), 1.63−1.56
(m, 1H), 1.40−1.32 (m, 1H). 13C NMR (151 MHz, CD3OD): δ
152.2, 143.0, 141.9, 130.2, 130.0, 129.9, 129.2, 128.5, 128.3, 61.8, 55.1,
49.5, 39.3, 36.0, 24.0, 21.7.
1
observed in the H NMR spectrum. A possible explanation for this
effect is the aggregate formation or hindered rotation due to steric bulk
of protecting groups. 1H NMR (600 MHz, CD3OD): δ 7.40 (d, J = 7.6
Hz, 2H), 7.38 (d, J = 7.6 Hz, 2H), 7.32 (t, J = 7.6 Hz, 2H), 7.28 (t, J =
7.6 Hz, 2H), 7.20 (t, J = 7.6 Hz, 1H), 7.17 (t, J = 7.6 Hz, 1H), 4.63−
4.54 (bm, 1H), 4.03 (bd, J = 9.7 Hz, 1H), 3.65−3.59 (bm, 1H), 3.55
(td, J = 10.6, 4.3 Hz, 1H), 3.52−3.41 (bm, 2H), 3.19−2.87 (bm, 2H),
1.90−1.82 (m, 1H), 1.62−1.40 (bm, 3H). 13C NMR (151 MHz,
CD3OD): δ 158.1, 143.6, 143.0, 129.9, 129.6, 129.2, 127.9, 127.8, 59.5,
(6S,10S)-6,10-Di(tert-butyl)-2,2,14,14-tetramethyl-3,3,13,13-tet-
raphenyl-4,12-dioxa-7,9-diaza-3,13-disilapentadecan-8-iminium
2,2,2-trifluoroacetate (36).41 To a flask containing protected
aminoalcohol 22 (2.01 g, 5.65 mmol, 2.0 equiv) and guanylating
reagent 6 (456 mg, 2.83 mmol) of CH2Cl2 (1 mL) was added, and the
resulting mixture was concentrated using a rotary evaporator under
reduced pressure at rt. Upon removal of CH2Cl2, a suspension
containing finely distributed guanylating reagent 6 was obtained. After
the addition of TFA (217 μL, 323 mg, 2.83 mmol), the resulting
suspension was mixed in a Kugelrohr oven at 105 °C for 4 h. Upon
heating, the heterogeneous mixture became a very viscous brownish
solution. The resulting crude product was purified by column
chromatography (Si, 30 × 5 cm, CH2Cl2/MeOH/AcOH = 100/1/
0.5 → 98/2/0.5) to give the symmetrically substituted guanidinium
salt 36 (1.2 g) as an off-white foam. Isolated product contained
approximately 10 to 20 mol % of unknown impurities, which were
visible in the 1H NMR spectrum. Taking this into account, the
estimated amount of the pure product was approximately 980 mg
(1.13 mmol, 40%). Rf = 0.31 (CH2Cl2/MeOH/AcOH = 100/8/0.5,
[UV, bromocresol green]). 1H NMR (600 MHz, CDCl3) δ 9.16 (br s,
2H), 7.61 (d, J = 6.6 Hz, 4H), 7.58 (d, J = 6.6 Hz, 4H), 7.46−7.34 (m,
12H), 6.55 (br s, 2H), 3.94 (d, J = 10.2, 2H), 3.68 (virt t, J = 10.2 Hz,
2H), 3.20 (virt t, J = 8.4 Hz, 2H), 1.05 (s, 18H), 1.01 (s, 18H). 13C
NMR (151 MHz, CDCl3) δ 159.6, 135.51, 135.46, 132.1, 131.8,
130.49, 130.45, 128.33, 128.26, 66.8, 60.6, 34.0, 27.13, 27.06, 19.2. 19F
NMR (564 MHz, CDCl3) δ −75.7 (s, CF3). [α]2D1 °C = −80.2 (CHCl3,
59.3, 58.7, 54.0, 39.3, 37.2, 32.3. 19F NMR (564 MHz, CD3OD): δ
−144.2 (s, 1F). [α]D31
°
= +53.2 (MeOH, c = 2). IR: ν
̃
= 3060 (w),
C
3029 (w), 2946 (w), 1635 (vs), 1492 (m), 1451 (m), 1367 (w), 1063
(s), 749 (s), 698 (vs). MS (ESI): m/z = 364.20 [M − F− − H+ + Na+],
342.22 [M − F−]. HRMS (ESI, Orbitrap): calcd for C20H28N3O2+ [M
− F−]: 342.21760; found: 342.21649.
(S)-((4-Chloro-1,1-diphenylbutan-2-yl)amino)((3-chloropropyl)-
amino)methaniminium Chloride (34).44 To a flask containing a
mixture of guanidinium salt 33 (133 mg, 368 μmol) in CHCl3 (6.9
mL, HPLC grade), thionyl chloride (161 μL, 263 mg, 2.2 mmol, 6.0
equiv) was added. The mixture was stirred at 60 °C for 2.5 h. After
cooling to rt and addition of MeOH (12 mL), the resulting mixture
was stirred for 25 min at rt. The solvents were removed under reduced
pressure to give the intermediate 34 as a yellowish foam (150 mg).
This crude material was characterized and used in the next step
without additional purification. Rf = 0.30 (CH2Cl2/MeOH = 9/1 [UV,
1
bromocresol green]). H NMR (400 MHz, CD3OD): δ 7.47 (d, J =
7.4 Hz, 2H), 7.41 (d, J = 7.4 Hz, 2H), 7.35 (t, J = 7.4 Hz, 2H), 7.31−
7.16 (m, 4H), 4.70−4.59 (m, 1H), 4.03 (d, J = 10.9 Hz, 1H), 3.68−
3.54 (m, 2H), 3.45 (t, J = 5.7 Hz, 2H), 3.16 (t, J = 6.6 Hz, 2H), 2.12−
2.00 (m, 1H), 1.99−1.87 (m, 1H), 1.84−1.70 (bm, 2H). 13C NMR
c = 1). IR (ATR) ν
̃
= 3317 cm−1 (w), 2956 (m), 2170 (w), 1897 (w),
K
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
1640 (s), 1471 (m), 1427 (m), 1367 (w), 1309 (w), 1245 (w), 1197
(m), 1107 (s), 1000 (m), 938 (w), 819 (s), 738 (s), 700 (s). MS (EI,
70 eV) m/z (%) 736.9 (2) [(M − CF3COO−)+], 678.7 [(M −
CF3COO− − t-Bu)+] (45), 466.5 (100), 340.3 (12), 224.1 (27), 135.1
(33). MS (ESI, Orbitrap) m/z 736.5 [(M − CF3COO−)+]. HRMS
μL, 469 mg, 3.1 mmol, 4 equiv) was added. After addition of DBU, all
solids slowly dissolved, and the resulting brown solution was stirred
under an Ar atmosphere at rt for 28 h. The reaction mixture was
acidified with HCl solution (1 M in MeOH) to pH 1−2, and the
solvents were removed under reduced pressure. The resulting crude
product (brown sticky mass) was purified by column chromatography
(Si, 23 × 3 cm, EtOAc/MeOH = 20/1 → 7/1 → 7/3). After
evaporation of solvents, the isolated material (product containing silica
particles) was suspended in MeCN, and silica was filtered off. Solvent
was evaporated to give the final product as a yellowish solid (97 mg,
373 μmol, 48%). Rf = 0.41 (EtOAc/MeOH = 8/2, [bromocresol
+
(ESI, Orbitrap) calcd for C45H66N3O2Si2 [(M − CF3COO−)+]:
736.46881; found: 736.46930.
Bis[(1S)-1-(hydroxymethyl)-2,2-dimethylpropyl]aminometh-
animinium Fluoride (37).43 To a flask containing a solution of
protected guanidine 36 (844 mg; due to impurities, estimated amount
of the starting material ca. 950 μmol) in MeCN (HPLC-grade, 19
mL), fluoride on polymer (1.2 g, 3.58 mmol, 4 equiv; 3.0 mmol/g
loading on Amberlyst A-26 resin; obtained from a commerical source)
was added, and the resulting mixture was stirred under Ar atmosphere
at rt. During this time, a white precipitate was formed. After 25 h,
MeOH was added to the resulting suspension until the formed
precipitate completely dissolved. Fluoride on polymer beads were
filtered off and washed with additional MeOH (10 mL), and the
solvents were removed under reduced pressure. When the resulting
crude product (oily mass) was washed with an Et2O/pentane = 1/5
mixture (3 × 5 mL), the product 37 was obtained as a white solid (261
mg, 934 μmol, 98%: estimated value due to impurities in the starting
1
green]) H NMR (600 MHz, CD3OD) δ 4.13 (dd, J = 9.6, 3.6 Hz,
1H), 3.96 (virt t, J = 9.6 Hz, 1H), 3.79 (virt t, J = 9.6 Hz, 1H), 3.64
(virt t, J = 9.6 Hz, 1H), 3.62 (dd, J = 9.6, 3.6 Hz, 1H), 3.56 (virt t, J =
9.6, Hz, 1H), 0.99 (s, 9H), 0.96 (s, 9H). 13C NMR (101 MHz,
1
CD3OD) δ 167.4, 73.0, 69.8, 51.8, 49.8, 35.3, 34.0, 26.4, 25.5. H
NMR (400 MHz, CDCl3) δ 9.14 (br s, 1H), 8.85 (br s, 1H), 4.06 (dd,
J = 10.0, 3.6 Hz, 1H), 3.93 (virt t, J = 10.0 Hz, 1H), 3.72 (virt t, J =
10.0 Hz, 1H), 3.44 (virt t, J = 9.6 Hz, 1H), 3.43 (dd, J = 9.6, 3.6 Hz,
1H), 3.38 (virt t, J = 9.6, Hz, 1H), 0.92 (s, 9H), 0.91 (s, 9H). 13C
NMR (101 MHz, CDCl3) δ 166.2, 71.7, 69.1, 50.9, 49.0, 34.5, 33.3,
26.2, 25.3. [α]2D1 °C = −18.5 (MeOH, c = 0.1). Mp decomposes above
1
200 °C. IR (ATR) ν
̃
= 2957 cm−1 (s), 1699 (s), 1559 (s), 1471 (s),
material). Rf = 0.0 (CH2Cl2/MeOH = 9/1, [bromocresol green]). H
NMR (400 MHz, CD3OD) δ 3.91 (dd, J = 10.8, 2.8 Hz, 2H), 3.51 (dd,
J = 9.2, 10.8 Hz, 2H), 3.35 (dd, J = 8.8, 2.8 Hz, 2H), 1.01 (s, 18H). 13C
NMR (151 MHz, CD3OD) δ 160.4, 65.1, 63.3, 35.0, 27.2. 19F NMR
(376 MHz, CD3OD) δ −150.6 (s, F−). [α]D21 °C = −97.3 (MeOH, c =
1399 (m), 1366 (m), 1272 (s), 1136 (m), 1067 (m), 1024 (m), 934
(w), 843 (m), 714 (m). MS (EI, 70 eV) m/z (%) 224.3 (4) [(M −
Cl−)+], 166.2 (100) [(M − HCl − t-Bu)+], 96.1 (39), 57.2 (28). MS
(ESI, Orbitrap) m/z 224.2 [(M − Cl−)+]. HRMS (ESI, Orbitrap)
1). Mp 161−164 °C. IR (ATR) ν
̃
= 3386 cm−1 (s), 3230 (s), 2956 (s),
+
calcd for C13H26N3 [(M − Cl−)+]: 224.21212; found: 224.21201.
2869 (s), 2325 (w), 2103 (w), 1689 (s), 1630 (s), 1523 (m), 1470 (s),
1362 (m), 1303 (m), 1251 (m), 1090 (m), 1056 (s), 1015 (m), 930
(m), 857 (m), 738 (m), 673 (m). MS (EI, 70 eV) m/z (%) 260.4
(100) [(M − F−)+], 228.3 (34), 202.3 (25) [(M − F− − t-Bu)+], 57.3
(12) [t-Bu+]. MS (ESI, Orbitrap) m/z 260.2 [(M − F−)+]. HRMS
(7S,11S)-7,11-Dibenzhydryl-2,2,16,16-tetramethyl-3,3,15,15-tet-
raphenyl-4,14-dioxa-8,10-diaza-3,15-disilaheptadecan-9-iminium
2,2,2-trifluoroacetate (40).41 To a flask containing TBDPS-protected
aminoalcohol 19 (1.8 g, 3.75 mmol, 2.0 equiv), guanylating reagent 6
(303 mg, 1.88 mmol) and CH2Cl2 (1 mL) were added, and the
resulting mixture was concentrated on a rotary evaporator under
reduced pressure at rt. Upon removal of CH2Cl2, a suspension
containing finely distributed guanylating reagent 6 was obtained. After
the addition of TFA (144 μL, 214 mg, 1.88 mmol), the resulting
suspension was mixed in a Kugelrohr oven at 105 °C for 4 h. Upon
heating, the heterogeneous mixture became a very viscous brownish
solution. The resulting crude product was purified by column
chromatography (Si, 30 × 5 cm, CH2Cl2/MeOH/AcOH = 100/2/
0.5 → 97/3/0.5) to give the symmetrically substituted guanidinium
salt 40 (758 mg, 690 μmol, 38%) as a brownish oil. Rf = 0.56
+
(ESI, Orbitrap) calcd for C13H30N3O2 [(M − F−)+]: 260.23325;
found: 260.23312.
Bis[(1S)-1-(chloromethyl)-2,2-dimethylpropyl]aminometh-
animinium Chloride (38).44 To a flask containing a suspension of
guanidinium salt 37 (110 mg, 394 μmol) in MeCN (HPLC grade, 12
mL), thionyl chloride (343 μL, 563 mg, 4.7 mmol, 12 equiv) was
added. Upon addition of SOCl2, all solids immediately dissolved, and
the resulting mixture was stirred under an Ar atmosphere at 65−70 °C
for 24 h. It is highly recommended to monitor the reaction progress by
TLC using KMnO4 staining reagent (a very characteristic and
intensive staining of the product), since starting material and product
are not UV active and using bromocresol green gave a very weak
staining. After addition of MeOH (15 mL), the resulting mixture was
stirred for 1 h at rt. The solvents were removed under reduced
pressure to give the crude product as a yellow solid. The crude product
was used without further purification.
1
(CH2Cl2/MeOH/AcOH = 100/8/0.5, [UV, bromocresol green]). H
NMR (600 MHz, CDCl3) δ 9.39 (br s, 2H), 7.62 (d, J = 6.6 Hz, 4H),
7.47 (t, J = 7.2 Hz, 8H), 7.41−7.38 (m, 2H), 7.36 (d, J = 7.8 Hz, 4H),
7.33−7.28 (m, 10H), 7.27−7.25 (m, 8H), 7.19 (t, J = 7.2 Hz, 4H),
7.04 (t, J = 7.2 Hz, 2H), 6.25 (br s, 2H), 4.47 (virt q, J = 10.8 Hz, 2H),
4.05 (d, J = 11.4 Hz, 2H), 3.52 (dd, J = 4.2, 10.8 Hz, 2H), 3.34 (virt t, J
= 11.4 Hz, 2H), 1.67−1.62 (m, 2H), 1.18−1.12 (m, 2H), 1.03 (s, 9H).
13C NMR (151 MHz, CDCl3) δ 158.1, 142.2, 141.6, 135.5, 135.4,
133.0, 132.1, 130.47, 130.46, 129.1, 128.8, 128.2, 128.1, 127.9, 127.1,
126.9, 60.3, 56.8, 53.7, 35.1, 27.3, 19.4. 19F NMR (282 MHz, CDCl3) δ
A small amount of the crude product (ca. 120 mg) was purified by
column chromatography (Si, 2.0 × 20 cm, CH2Cl2/MeOH = 98/2 →
95/5) to give a sample for spectrosopic analyses (66 mg, 198 μmol) as
a yellowish solid. This sample still contained unidentified impurities.
Crystals suitable for X-ray analysis were obtained by dissolving this
material in MeOH followed by slow evaporation at rt over several
days.32 Rf = 0.33 (CH2Cl2/MeOH = 9/1, [KMnO4, bromocresol
−75.6 (s, CF3). [α]D21 °C = +2.5 (CHCl3, c = 1). IR (ATR) ν
̃
= 3335
cm−1 (w), 3187 (w), 3065 (w), 2935 (w), 2860 (w), 1805 (w), 1685
(s), 1428 (m), 1379 (w), 1305 (w), 1250 (w), 1194 (s), 1107 (s), 936
(m), 823 (m), 747 (s), 698 (s). MS (EI, 70 eV) m/z (%) 984.1 (1)
[(M − CF3COO−)+], 816.9 (29) [(M − CF3COO−− CHPh2)+],
465.4 (35), 422.4 (80), 405.3 (100), 312.3 (71), 234.3 (49), 224.2
1
green]) H NMR (600 MHz, CD3OD) δ 3.95 (dd, J = 12.0, 3.0 Hz,
2H), 3.79−3.69 (br m, 2H), 3.64−3.55 (br m, 2H), 1.03 (s, 18H). 13C
NMR (151 MHz, CD3OD) δ 159.6, 63.9, 46.3, 36.9, 26.7. [α]2D1 °C =
+
̃
+0.5 (MeOH, c = 1). Mp decomposes above 220 °C IR (ATR) ν =
(37), 199.2 (43), 167.2 (37) [CHPh2 ]. MS (ESI, Orbitrap) m/z
3156 cm−1 (s), 2963 (s), 2411 (m), 2326 (m), 1661 (s), 1620 (s),
1473 (s), 1439 (m), 1368 (m), 1270 (m), 1213 (m), 1095 (m), 1053
(m), 1016 (m), 915 (m), 852 (w), 791 (w), 734 (s). MS (EI, 70 eV)
m/z (%) 297.4 (2) [(M − Cl−)+], 236.1 (100), 200.6 (50). MS (ESI,
Orbitrap) m/z 296.2 [(M − Cl−)+]. HRMS (ESI, Orbitrap) calcd for
984.5 [(M − CF3COO−)+]. HRMS (ESI, Orbitrap) calcd for
+
C65H74N3O2Si2 [(M − CF3COO−)+]: 984.53141; found: 984.52673.
Bis[(1S)-1-benzhydryl-3-hydroxypropyl]aminomethaniminium
Fluoride and Trifluoroacetate (41).43 To a flask containing a solution
of protected guanidinium salt 40 (1.33 g, 1.21 mmol) in MeCN
(HPLC-grade, 80 mL), fluoride on polymer (2.0 g, 6.0 mmol, 5.0
equiv; 3.0 mmol/g loading on Amberlyst A-26 resin; obtained from a
commerical source) was added, and the resulting mixture was stirred
under Ar at rt. During this time, a white precipitate was formed. After
43 h, MeOH (ca. 30 mL) was added to the resulting suspension until
+
C13H28Cl2N3 [(M − Cl−)+]: 296.16548; found: 296.16486.
(2S,5S)-2,5-Di(tert-butyl)-1H,2H,3H,5H,6H,7H-imidazo[1,2-a]-
imidazol-4-ium Chloride (39).45 To a flask containing a suspension of
crude guanidinium salt 38 (769 μmol considering quantitative yield of
the previous reaction) in MeCN (HPLC grade, 13 mL), DBU (460
L
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
the formed precipitate completely dissolved. Fluoride on polymer
beads were filtered off and washed with additional MeOH (10 mL),
and the solvents were removed under reduced pressure. When the
resulting crude product (oily mass) was washed with an Et2O/pentane
= 1/1 mixture (20 mL), a precipitate was formed. It was washed with
pentane (10 mL) and dried under reduced pressure, and the product
41 (640 mg) was obtained as an off-white solid. The isolated product
contained unidentified impurities and residual Et2O. Quantitative yield
was assumed.
Crystals suitable for X-ray analysis were obtained by dissolving this
material in MeOH/CHCl3 followed by slow evaporation at rt over
several days.32 Rf = 0.46 (CH2Cl2/MeOH = 9/1, [UV, bromocresol
1
green]). H NMR (600 MHz, CD3OD) 7.41 (d, J = 7.2 Hz, 4H),
7.32−7.27 (m, 12H), 7.22−7.17 (m, 4H), 4.33 (ddd, J = 9.0, 10.8, 3.6
Hz, 2H), 3.87 (br d, J = 10.8 Hz, 2H), 3.42−3.35 (m, 4H), 1.83−1.79
(m, 2H), 1.74−1.67 (m, 2H). 13C NMR (151 MHz, CD3OD) δ 152.3,
142.6, 141.8, 130.3, 130.1, 129.4, 129.1, 128.6, 128.3, 58.4, 52.7, 46.9,
̃
26.7. [α]2D1 °C = −23.0 (MeOH, c = 1). IR (ATR) ν = 3645 cm−1 (w),
3263 (w), 3138 (w), 2938 (w), 2873 (w), 2081 (w), 1620 (s), 1526
(w), 1493 (m), 1448 (m), 1360 (m), 1314 (s), 1106 (m), 1029 (w),
919 (s), 898 (s). MS (EI, 70 eV) m/z (%) 472.6 (1) [(M − Cl−)+],
304.4 (100), 137.2 (7). MS (ESI, Orbitrap) m/z 472.3 [(M − Cl−)+].
Severe signal broadening was observed in NMR spectra, probably,
due to aggregate formation. For this reason, only approximate
chemical shifts and integral values are presented. Rf = 0.0 (CH2Cl2/
MeOH = 9/1, [UV, bromocresol green]). 1H NMR (600 MHz,
CD3OD) 7.75−6.40 (m, 22 H), 4.75−3.80 (br m, 4H), 3.60−3.10 (br
m, 2H), 2.00−1.55 (br m, 2H), 1.50−1.20 (br m, 2H). 13C NMR (151
MHz, CD3OD) δ 158.3, 143.7, 143.1, 129.9, 129.7, 129.1, 129.0, 128.7,
127.9, 59.3, 53.6, 41.8, 37.7. One signal was not included due to
overlaps resulting from low resolution. 19F NMR (282 MHz, CD3OD)
δ −76.9 (s, CF3, minor signal), −150.5 (s, F−, major signal). [α]D24 °C =
+
HRMS (ESI, Orbitrap) calcd for C33H34N3 [(M − Cl−)+]:
472.27472; found: 472.27390.
(2S,6S)-2,6-Dibenzhydryl-1H,2H,3H,4H,6H,7H,8H,9H-pyrimido-
[1,2-a]pyrimidin-5-ium Chloride (43). The undesired regioisomer 43
(250 mg, 492 μmol, 31%) was the main reaction product and obtained
as a slightly less polar fraction (white foam). Furthermore, a mixture
containing both regioisomers isomers (10/43 = 1/3) was obtained in a
total yield of 24%. The total yield of the desired guanidinium salt 10 is
therefore ca. 14%. Rf = 0.46 (CH2Cl2/MeOH = 9/1, [UV,
−4.7 (MeOH, c = 1). Mp 115−118 °C IR (ATR) ν
̃
= 3309 cm−1 (w),
3176 (w), 3060 (w), 3029 (w), 2929 (w), 1780 (w), 1646 (s), 1551
(w), 1493 (m), 1330 (w), 1251 (w), 1184 (w), 1106 (m), 1044 (m),
912 (w), 822 (m), 747 (s), 698 (s). MS (EI, 70 eV) m/z (%) 508.1
(1) [(M − F−)+], 340.5 (6) [(M − CF3COO− − CHPh2)+], 167.2
1
bromocresol green]). H NMR (400 MHz, CD3OD) 7.59 (d, J =
7.2 Hz, 2H), 7.52 (d, J = 7.2 Hz, 2H), 7.49−7.35 (m, 9H), 7.29−7.22
(m, 4H), 7.18−7.14 (m, 1H), 7.06 (d, J = 7.2 Hz, 2H), 4.50 (d, J =
11.6 Hz, 1H), 4.30−4.25 (m, 1H), 4.22 (d, J = 11.6 Hz, 1H), 3.52 (d, J
= 11.2 Hz, 1H), 3.44−3.36 (m, 1H), 3.27−3.16 (m, 2H), 2.54−2.47
(m, 1H), 2.03−1.93 (m, 1H), 1.87 (dd, J = 14.0, 5.6 Hz, 1H), 1.61−
1.53 (m, 1H), 1.18−1.11 (m, 1H). 13C NMR (101 MHz, CD3OD) δ
151.8, 143.5, 142.4, 142.3, 142.2, 130.5, 130.3, 130.2, 130.0, 129.4,
129.2, 128.9, 128.5, 128.4, 128.3, 62.0, 57.1, 55.4, 52.5, 47.5, 36.1, 25.3,
+
(100) [CHPh2 ], 152.2 (32), 99.2 (8). MS (ESI, Orbitrap) m/z 508.3
+
[(M − F−)+]. HRMS (ESI, Orbitrap) calcd for C33H38N3O2 [(M −
F−)+]: 508.29585; found: 508.29373.
Bis[(1S)-1-benzhydryl-3-chloropropyl]aminomethaniminium
Chloride (42).44 To a flask containing a suspension of guanidinium salt
41 (914 mg, 1.7 mmol) in MeCN (HPLC grade, 106 mL), thionyl
chloride (1.3 mL, 2.1 g, 17.6 mmol, 12 equiv) was added. Upon
addition of SOCl2, all solids immediately dissolved, and the resulting
mixture was stirred under Ar at rt for 24 h. After addition of MeOH
(20 mL), the resulting mixture was stirred for 1 h at rt. The solvents
were removed under reduced pressure to give the crude product 42
(1.41 g) as a yellowish solid. The crude product was used without
further purification. A small amount of the crude product (ca. 100 mg)
was purified by column chromatography (Si, 2.0 × 20 cm, CH2Cl2/
MeOH = 99/1 → 97/3) to give an analytically pure sample for
spectrosopic analyses as an off-white solid. Rf = 0.71 (CH2Cl2/MeOH
̃
24.0. [α]2D1 °C = +111.2 (MeOH, c = 1). IR (ATR) ν = 3636 cm−1 (w),
3278 (w), 3028 (w), 2945 (w), 2884 (w), 2059 (w), 1630 (s), 1527
(w), 1493 (m), 1449 (m), 1363 (m), 1316 (s), 1205 (w), 1101 (m),
1027 (m), 749 (s), 699 (s). MS (EI, 70 eV) m/z (%) 472.6 (5) [(M −
Cl−)+], 304.4 (100), 137.1 (19). MS (ESI, Orbitrap) m/z 472.3 [(M −
+
CF3COO−)+]. HRMS (ESI, Orbitrap) calcd for C33H34N3 [(M −
CF3COO−)+]: 472.27472; found: 472.27481.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.joc.6b00283.
■
1
S
= 9/1, [bromocresol green]). H NMR (300 MHz, CD3OD) 7.41−
7.18 (m, 20H), 4.45 (virt t, J = 9.6 Hz, 2H), 3.87 (br d, J = 10.8 Hz,
2H), 3.24−3.12 (br m, 2H), 3.09−2.94 (br m, 2H), 1.95−1.80 (br m,
2H), 1.74−1.57 (br m, 2H). 13C NMR (151 MHz, CD3OD) δ 157.0,
143.2, 142.5, 130.2, 129.7, 129.6, 129.2, 128.3, 128.2, 59.7, 54.5, 41.8,
Crystallographic data for compounds 8, 10 and 38 (CIF)
NMR spectra of all intermediates and final products.
Relevant HSQC and HMBC correlations of the
compound 31 (PDF)
37.9. [α]2D1 °C = +61.6 (MeOH, c = 1). Mp 133−136 °C IR (ATR) ν
̃
=
3633 cm−1 (w), 3031 (m), 2317 (m), 2087 (w), 1961 (w), 1716 (w),
1582 (s), 1493 (m), 1449 (m), 1361 (m), 1251 (w), 1169 (w), 1083
(m), 1039 (m), 918 (w), 814 (w), 749 (s), 699 (s). MS (EI, 70 eV)
m/z (%) 471.6 (4), 304.4 (100), 137.3 (7). MS (ESI, Orbitrap) m/z
AUTHOR INFORMATION
Corresponding Authors
*E-mail: aleksej.turockin@org.chem.ethz.ch.
*E-mail: philipp.selig@patheon.com.
+
■
544.2 [(M − Cl−)+]. HRMS (ESI, Orbitrap) calcd for C33H36Cl2N3
[(M − Cl−)+]: 544.22808; found: 544.22595.
(2S,8S)-2,8-Dibenzhydryl-1H,2H,3H,4H,6H,7H,8H,9H-pyrimido-
[1,2-a]pyrimidin-5-ium Chloride (10).45 To a flask containing a
suspension of crude guanidinium salt 42 (1.6 mmol assuming
quantitative yield of the previous reaction) in dry MeCN (30 mL),
DBU (955 μL, 974 mg, 6.4 mmol, 4.0 equiv) was added. The resulting
brown suspension was stirred under Ar at rt for 4 h. After addition of
one more portion of DBU (470 μL, 479 mg, 3.2 mmol, 2.0 equiv), the
mixture was stirred at rt for another 16 h. The reaction mixture was
acidified with HCl (1 M in MeOH) to pH 1−2, and the solvents were
removed under reduced pressure. The resulting crude product (brown
sticky mass) was purified by column chromatography (Si, 26 × 3.5 cm,
CH2Cl2/MeOH = 97/3 → 92/8). Due to the very similar polarities of
the formed regioisomers 10 and 43, one more chromatographic
purification was necessary to afford separation (Si, 20 × 2.0 cm,
toluene/CH2Cl2/MeOH = 1/98/1 → 1/97/2 → 1/96/3 → 1/95/4→
CH2Cl2/MeOH = 96/4). The desired regioisomer 10 (69 mg, 136
μmol, 8%) was obtained as yellowish foam.46
Present Addresses
§Laboratorium fur Organische Chemie, ETH Zurich, HCI
̈
̈
E306, Vladimir-Prelog-Weg 3, 8093 Zurich, Switzerland
∥Organisch-Chemisches Institut, Westfalische Wilhelms-Uni-
̈
versitat Munster, Corrensstraße 40, 48149 Munster, Germany
̈
̈
̈
⊥Patheon, St.-Peter-Str. 25, 4020 Linz, Austria
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Generous financial support by the Fonds der Chemischen
Industrie (Liebig-Scholarship to P.S. no. Li 186/09, PhD
scholarship to A.T. no. Do-L 190/06) and Evonik Industries
M
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/
cif.
(33) At elevated temperatures, the reagent 6 is known to decompose
to the corresponding cyanamide by elimination of imidazole:
McCallum, P. B. W.; Grimmett, M. R.; Blackman, A. G.; Weavers,
R. T. Aust. J. Chem. 1999, 52, 159.
AG (Germany Grant to R.H.) is gratefully acknowledged. We
thank Prof. Dr. U. Englert (Institute of Inorganic Chemistry,
RWTH Aachen University) for X-ray structure determinations.
Special thanks go to Prof. Dr. D. Enders, Institute of Organic
Chemistry, RWTH Aachen University, for the generous
provision of laboratory space and chemicals.
(34) Boyle, P. H.; Davis, A. P.; Dempsey, K. J.; Hosken, G. D. J.
Chem. Soc., Chem. Commun. 1994, 1875.
(35) Bromocresol green staining reagent has become for us one of
the most valuable qualitative tools for convenient detection of
guanidines. In course of this work, we have noticed that developing
of TLC plates containing guanidines and guanidinium intermediates
with bromocresol green showed a very specific turquoise color clearly
distinct from the blue staining of amines.
(36) Sciuto, S.; Piatelli, M.; Chillemi, R. Phytochemistry 1982, 21, 227.
(37) Staron, T.; Allard, C.; Xuong, N. D.; Chambre, M.-M.;
Grabowski, H. C. R. Acad. Sci. 1964, 259, 3114.
(38) Kumaraswamy, G.; Jayaprakash, N. Tetrahedron Lett. 2010, 51,
6500.
(39) Cariou, C. A. M.; Kariuki, B. M.; Snaith, J. S. Org. Biomol. Chem.
2008, 6, 3337.
(40) Baliani, A.; Peal, V.; Gros, L.; Brun, R.; Kaiser, M.; Barrett, M.
P.; Gilbert, I. H. Org. Biomol. Chem. 2009, 7, 1154.
(41) This procedure was based on the guanylation protocol described
by Schmidtchen in ref 26.
(42) 27·H2O was obtained after additional column chromatography
of 27·TFA over activated alumina as stationary phase. We assume that
anion exchange occurred due to different activity grade of that batch of
aluminium oxide compared to the one used for the first chromato-
graphic purification.
(43) This deprotection was based in the procedure described by
Schmidtchen in ref 26.
(44) This chlorination procedure was based on the protocol
described by: McKay, A. F.; Kreling, M.-E. Can. J. Chem. 1957, 35,
1438.
(45) The final cyclization protocol was based on the procedure
described by Schmidtchen in ref 26.
(46) The corresponding hydroiodide salt of this guanidine was
reported by Davis et al. (see ref 24). Our spectroscopic data essentially
match those reported by Davis. The only disagreement is that we
observed eight aromatic signals in the 13C NMR spectrum, whereas
Davis reported 10 carbon signals in the aromatic region.
REFERENCES
■
(1) Ishikawa, T. In Superbases for Organic Synthesis; Ishikawa, T., Ed.;
John Wiley & Sons Ltd.: Chichester, 2009; p 93.
(2) For instance, pKa[80% DMSO] of guanidinium hydrochloride is
13.6; see: Baldini, L.; Cacciapaglia, R.; Casnati, A.; Mandolini, L.;
Salvio, R.; Sansone, F.; Ungaro, R. J. Org. Chem. 2012, 77, 3381.
Whereas pKa[MeCN] of protonated TBD (2) is 28.0; see: Kaljurand,
I.; Kutt, A.; Soovali, L.; Rodima, T.; Maemets, V.; Leito, I.; Koppel, I.
̈
̈
̈
A. J. Org. Chem. 2005, 70, 1019.
(3) Maji, B.; Stephenson, D. S.; Mayr, H. ChemCatChem 2012, 4,
993.
(4) Pratt, R. C.; Lohmeijer, B. G. G.; Long, D. A.; Waymouth, R. M.;
Hedrick, J. L. J. Am. Chem. Soc. 2006, 128, 4556.
(5) Kiesewetter, M. K.; Scholten, M. D.; Kirn, N.; Weber, R. L.;
Hedrick, J. L.; Waymouth, R. M. J. Org. Chem. 2009, 74, 9490.
(6) Selig, P.; Turock
(7) Selig, P.; Turock
2930.
(8) Selig, P.; Turock
297.
(9) Blondeau, P.; Segura, M.; Per
Chem. Soc. Rev. 2007, 36, 198.
̌
in, A.; Raven, W. Synlett 2013, 24, 2535.
in, A.; Raven, W. Chem. Commun. 2013, 49,
̌
̌
in, A.; Raven, W. Adv. Synth. Catal. 2013, 355,
́
́
ez-Fernandez, R.; de Mendoza, J.
(10) Dieng, P. S.; Sirlin, C. Int. J. Mol. Sci. 2010, 11, 3334.
(11) Fu, X.; Tan, C.-H. Chem. Commun. 2011, 47, 8210.
(12) Selig, P. Synthesis 2013, 45, 703.
(13) Taylor, J. E.; Bull, S. D.; Williams, J. M. J. Chem. Soc. Rev. 2012,
41, 2109.
(14) Leow, D.; Tan, C.-H. Chem. - Asian J. 2009, 4, 488.
(15) Terada, M. Yuki Gosei Kagaku Kyokaishi 2010, 68, 1159.
(16) Ishikawa, T. Chem. Pharm. Bull. 2010, 58, 1555.
(17) Shubina, T. E.; Freund, M.; Schenker, S.; Clark, T.; Tsogoeva, S.
B. Beilstein J. Org. Chem. 2012, 8, 1485.
(18) Gleich, A.; Schmidtchen, F. Chem. Ber. 1990, 123, 907.
(19) Corey, E. J.; Grogan, M. J. Org. Lett. 1999, 1, 157.
(20) Misaki, T.; Takimoto, G.; Sugimura, T. J. Am. Chem. Soc. 2010,
132, 6286.
(21) Misaki, T.; Kawano, K.; Sugimura, T. J. Am. Chem. Soc. 2011,
133, 5695.
(22) Dai, Q.; Huang, H.; Zhao, J. C.-G. J. Org. Chem. 2013, 78, 4153.
(23) Corey, E. J.; Ohtani, M. Tetrahedron Lett. 1989, 30, 5227.
(24) Davis, A. P.; Dempsey, K. J. Tetrahedron: Asymmetry 1995, 6,
2829.
(25) Echavarren, A.; Galan
J.-M. Helv. Chim. Acta 1988, 71, 685.
́ ́
, A.; de Mendoza, J.; Salmeron, A.; Lehn,
(26) Jadhav, V. D.; Schmidtchen, F. P. J. Org. Chem. 2008, 73, 1077.
(27) Wu, Y.-Q.; Hamilton, S. K.; Wilkinson, D. E.; Hamilton, G. S. J.
Org. Chem. 2002, 67, 7553.
(28) For detailed account on applications of Tan’s guanidine, see:
Leow, D.; Tan, C.-H. Synlett 2010, 11, 1589.
(29) Boyle, P. H.; Davis, A. P.; Dempsey, K. J.; Hosken, G. L.
Tetrahedron: Asymmetry 1995, 6, 2819.
(30) McKennon, M. J.; Meyers, A. I. J. Org. Chem. 1993, 58, 3568.
(31) This ratio was determined using NMR spectroscopy by
integration of the signals corresponding to tert-butyl groups (see
Experimental Section). However, precise structure assignment of the
minor species was not possible due to its extremely low quantity and
our inability to isolate it in pure form.
(32) See the Supporting Information for the X-ray crystal structures
of 8, 10, and 38. CCDC-1449940 (8)/1449941 (10)/1449942 (38)
contain the supplementary crystallographic data for this paper. These
data can be obtained free of charge from The Cambridge
N
DOI: 10.1021/acs.joc.6b00283
J. Org. Chem. XXXX, XXX, XXX−XXX