An optimized protocol for the multigram synthesis of 3-(trifluoromethyl)bicyclo[1.1.1]pent-1-ylglycine (CF3-Bpg)

Article abstract of DOI:10.1016/j.jfluchem.2009.10.004

An optimized procedure for the multigram synthesis of 3-(trifluoromethyl)bicyclo[1.1.1]pent-1-ylglycine (CF₃-Bpg) has been developed. The overall yield of the synthesis for the optimized up-scaling was increased from 35% to 53%. Moreover, conditions for separating the key isomeric aminonitriles 7 and 8 by crystallization were found, which greatly facilitated the isolation of 8 on a multigram scale. Following this optimized protocol, 100 g of optically pure CF₃-Bpg have been synthesized. Crown Copyright

Full text of DOI:10.1016/j.jfluchem.2009.10.004

Journal of Fluorine Chemistry 131 (2010) 217–220
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
An optimized protocol for the multigram synthesis of
3-(trifluoromethyl)bicyclo[1.1.1]pent-1-ylglycine (CF3-Bpg)
Pavel K. Mykhailiuk a,b, , Nataliia M. Voievoda a,b, Sergii Afonin c, Anne S. Ulrich c,d, Igor V. Komarov a,b
*
a Department of Chemistry, Kyiv National Taras Shevchenko University, Vul. Volodymyrska 64, 01033 Kyiv, Ukraine
b Enamine Ltd., Vul. Oleksandra Matrosova 23, 01103 Kyiv, Ukraine
c Karlsruhe Institute of Technology (KIT), Forschungszentrum Karlsruhe, Institut fu¨r Biologische Grenzfla¨chen, POB 3640, 76021 Karlsruhe, Germany
d KIT, Universita¨t Karlsruhe, Institut fu¨r Organische Chemie, Fritz-Haber-Weg 6, 76133 Karlsruhe, Germany
A R T I C L E I N F O
A B S T R A C T
Article history:
An optimized procedure for the multigram synthesis of 3-(trifluoromethyl)bicyclo[1.1.1]pent-1-
ylglycine (CF3-Bpg) has been developed. The overall yield of the synthesis for the optimized up-scaling
was increased from 35% to 53%. Moreover, conditions for separating the key isomeric aminonitriles 7 and
8 by crystallization were found, which greatly facilitated the isolation of 8 on a multigram scale.
Following this optimized protocol, 100 g of optically pure CF3-Bpg have been synthesized.
Crown Copyright ß 2009 Published by Elsevier B.V. All rights reserved.
Received 25 July 2009
Received in revised form 6 October 2009
Accepted 7 October 2009
Available online 15 October 2009
Keywords:
Fluorine
Amino acids
Conformational rigidity
Large-scale synthesis
1. Introduction
coil peptide pair VPE/VPK [10]. In fact, during the last 3 years, CF3-
Bpg proved to be the optimal label for solid-state 19F NMR
In 2006 we reported the synthesis of 3-(trifluoromethyl)bicy-
clo[1.1.1]pent-1-ylglycine (1, CF3-Bpg) (Fig. 1) [1]. This CF3-
structure analysis of membrane-bound peptides [11]. The main
advantage of CF3-Bpg is the stereochemical stability compared to
CF3-Phg, which racemizes extensively under basic conditions
employed during solid phase Fmoc peptide synthesis, and which
has led to difficulties in numerous cases where it was impossible to
separate the resulting peptide epimers by HPLC [12]. This problem
is absent when using CF3-Bpg, as it does not racemize upon
incorporation into peptides. Along with increasing the final
product yield and eliminating laborious epimer separation and
assignment, this makes CF3-Bpg an ideal 19F NMR label. The
growing interest in CF3-Bpg prompted us to find a practical
procedure for its large-scale production. Here, we report an
optimized synthesis of CF3-Bpg, which allows its preparation in
multigram quantities.
substituted conformationally rigid
a-amino acid was prepared
as a specific label to study membrane-bound peptides by solid-
state 19F NMR. Having been designed as an analogue of non-polar
aliphatic amino acids, it can be readily used to replace naturally
occurring residues of leucine, isoleucine, valine, alanine, and
methionine, which are highly abundant in membrane proteins and
peptides. The applicability of CF3-Bpg as a 19F NMR label was
evaluated by comparison with the established reporter 4-CF3-
phenylglycine (CF3-Phg) [2–4], and the structural compatibility of
both amino acids was proven [5,6]. CF3-Bpg was successfully
applied thereafter to elucidate the conformation, membrane
alignment, and the dynamic behaviour of several natural
antimicrobial peptides, namely PGLa [5], gramicidin
S [6],
temporin A [7], temporin L [7], as well as the cell penetrating
peptides SAP [6,8], and transportan-10 [9] in their functionally
relevant membrane-bound states. Moreover, CF3-Bpg has also
been used to investigate intermolecular interactions of the coiled-
2. Results and discussion
In our original work, the laboratory-scale synthesis of CF3-Bpg
[1] starting from 2 (Scheme 1) resulted in 1 g of the final product in
35% overall yield, which shall serve as a reference point for the up-
scaling process.
With the intention to lower the costs of the synthesis where
possible, each step was carefully checked. Thus, the synthetic step
‘‘b’’ (Scheme 1) had formerly been performed according to the
original procedure reported by Adcock and Gack [13], i.e. 2.2 equiv.
* Corresponding author at: Department of Chemistry, Kyiv National Taras
Shevchenko University, Vul. Volodymyrska 64, 01033 Kyiv, Ukraine.
Tel.: +380 44 2393315; fax: +380 44 2351273.
E-mail address: pashamk@gmx.de (P.K. Mykhailiuk).
0022-1139/$ – see front matter. Crown Copyright ß 2009 Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2009.10.004

218
P.K. Mykhailiuk et al. / Journal of Fluorine Chemistry 131 (2010) 217–220
Table 1
Isolated yields of 4 at different CF3I/propellane ratios, relative to 2. The value
marked grey was used in the large-scale synthesis.
CF3I/propellane (mol. ratio)
2.2
64
1.7
62
1.5
63
1.1
62
0.8
54
Isolated yield of 4 (%)
Fig. 1. Structures of the 19F NMR labels CF3-Bpg (1) and CF3-Phg.
of CF3I were added to the solution of 1 equiv. of propellane in Et2O.
Since CF3I is a rather expensive chemical (448.5 for 100 g of CF3I
s
in Sigma–Aldrich, 2009), a series of experiments were performed
using different CF3I/propellane ratios in order to diminish the
amount of CF3I (Table 1). Indeed, using only half as much CF3I
(1.1 equiv. instead of 2.2 equiv.) significantly lowered the cost
without significantly compromising the yield (62% instead of 64%).
The next challenge encountered on the large-scale synthesis of
aminonitrile 8. A 1:1 mixture of 7 and 8 is obtained in the Strecker
reaction of 5 and 6 with (R)-phenylglycinol (steps ‘‘e-g’’, Scheme 1)
[1]. The individual components of the mixture can be separated by
flash chromatography. However, the very small difference in the Rf
values (ꢀ0.05) made purification of
8 rather laborious and
inefficient: to completely separate 5 g of the mixture 7 and 8, more
than 1.5 kg of Silica gel Merck 60 was needed. Obviously, to separate
large quantity of 7 and 8 (>100 g), a much more efficient procedure
had to be applied. Therefore, extensive experimentation was
performed to find conditions to isolate 8 by means of crystallization.
First, the 1/1 mixture of 7 and 8 was isomerized in refluxing
MeOH, to give an isomer ratio (7/8) of about 1/4, as previously
observed [1]. Next, this mixture was crystallized from cyclohexane,
which increased the ratio to ꢀ1/5 (Scheme 2). Remarkably, further
crystallizations from cyclohexane did not improve this ratio.
However, a second crystallization from CCl4 provided 8 already as a
single diastereomer (Scheme 2). The mother liquors from the two
crystallization batches were combined, dried, and subjected to
isomerization in MeOH to re-establish the initial compound ratio
(7/8 = ꢀ1/4). Threefold repetition of the isomerization–crystal-
Scheme 1. Synthetic scheme used in the initial synthesis of CF3-Bpg [1]. Reagents and
conditions: (a) 2.4 equiv. MeLi, pentane, ꢁ78 8C, 0.5 h; (b) 2.2 equiv. CF3I, pentane, rt,
20 h; (c) 1.1 equiv. t-BuLi, Et2O, ꢁ78 8C, 1 h; (d) 4 equiv. HCO2Me, Et2O, ꢁ78 8C, rt, 3 h;
(e) 1 equiv. (R)-a-phenylglycinol, CH2Cl2, rt, 2 h; (f) 3 equiv. Me3SiCN, rt, 10 h; (g)
chromatographic separation; (h) MeOH, reflux, 3 h; (i) 1.4 equiv. Pb(OAc)4, CH2Cl2,
0 8C, 5 min; (j) 20% HCl, reflux, 2 h; (k) chromatography on Dowex-50.
Scheme 2. Optimized double crystallization procedure for the isolation of 8 from the mixture 7/8.
Scheme 3. Reagents and conditions: (a) original procedure: 1.4 equiv. Pb(OAc)4, CH2Cl2, 0 8C, 5 min; (b) 20% HCl, reflux, 2 h; (c) chromatography on ‘‘KU-2’’; (d) optimized
procedure: 1.5 equiv. Pb(OAc)4, MeOH/CH2Cl2 (1/1), 0 8C, 15 min.

P.K. Mykhailiuk et al. / Journal of Fluorine Chemistry 131 (2010) 217–220
219
lization cycle provided 8 in 90% overall yield (from 4), being even
4.3. Preparation of 2(S)-{[(1R)-2-hydroxy-1-phenylethyl]amino}-2-
higher than that obtained initially by the original isomerization-
chromatography approach (80%, Scheme 1). Notably, direct
crystallization of 7/8 (1/4) from CCl4 did not afford pure 8.
The last step of the synthesis was the conversion of 8 into the
target amino acid 1 via oxidation with Pb(OAc)4. At the onset we
failed to obtain 1 in a yield better than 70% [1]. Later, we reasoned,
that it may happen because of the incomplete oxidation of 8 into 9
(Scheme 3). This assumption was subsequently proven by isolating
the amino acid 10 (10% yield from 8) along with 1 during
purification of the reaction mixture. Obviously, 10 had formed as a
result of acidic hydrolysis of the CBB N group of the residual non-
oxidized 8 (Scheme 3). Indeed, the yield of 1 was significantly
improved to 95%, simply by using 1.5 equiv. of Pb(OAc)4 and
increasing the reaction time to 15 min.
[3-(trifluoromethyl)bicyclo[1.1.1]pent-1-yl]acetonitrile (8)
The crude mixture 7/8 (ꢀ1/4) was prepared from 4 as
previously described. Next, it was filtered through silica gel to
remove minor viscous impurities. The obtained isomers 7/8
(100 g) were dissolved in cyclohexane (400 mL) upon heating,
and then left to stand at rt for 2 h. The white solid was filtered to
provide 92 g of 7/8 (ꢀ1/5). The obtained amount of 7/8 was
dissolved in CCl4 (550 mL) upon heating, and left at room
temperature for ꢀ30 min. To the warm solution (at this moment
crystallization has not yet started), several crystals of pure 8
were added, and the mixture was left at rt overnight. White solid
was filtered to afford pure 8 (52 g). Mother liquors from two
crystallizations were combined and evaporated. The residue was
dissolved in MeOH (500 mL) and refluxed for 3 h. The solvent
was evaporated and the residue (7/8, ꢀ1/4) was submitted
again to double crystallization. Repetition of the isomerization–
crystallization procedure for three times provided 90 g of
pure 8.
3. Conclusions
We have optimized the initially reported synthetic procedure
for CF3-Bpg. Conditions for isolating the key intermediate 8 from
the mixture 7/8 by double crystallization were found. Using the
optimized protocol, 100 g of CF3-Bpg were obtained. The overall
yield of CF3-Bpg was increased from 35% to 53%.
4.4. Preparation of (2S)-2-amino-2-[3-
(trifluoromethyl)bicyclo[1.1.1]pent-1-yl]ethanoic acid (1)
4. Experimental
Pb(OAc)4 (214 g, 0.48 mol) was added to a solution of 8 (100 g,
0.32 mol) in CH2Cl2/MeOH (3 L, 1/1). After being stirred at 0 8C for
15 min, the reaction was poured into a saturated aq. solution of
NaHCO3 (4 L). The resulting insoluble material was removed by
filtration and washed with CH2Cl2 (2 L). Organic layer was
separated, and the aqueous layer was extracted with CH2Cl2
(3 ꢂ 1 L). The combined organic phases were evaporated in
vacuum to give Schiff base 9 as yellow oil. It was dissolved in
aq. HCl (6 M, 6 L), and refluxed for 6 h. After cooling, the reaction
mixture was washed with Et2O (3 ꢂ 1 L) and the aqueous layer was
evaporated to produce the white solid. Afterwards, the residue was
dissolved in H2O (ꢀ500 mL), neutralized with aq. NaOH (0.3 M)
and submitted to cation exchange resin chromatography on ‘‘KU-
2’’. The column was washed with water. Then, elution with aq. NH3
(10%) was performed. Evaporation of the eluate afforded 1 (64 g,
0.30 mol, 95%) as a white solid.
4.1. General
Solvents were purified according to standard procedures.
Compound 2 was prepared according to the literature procedure
[14]. All other materials were purchased from Fluka and Enamine.
Melting points are uncorrected. 1H-, 13C- and 19F NMR spectra were
recorded on a Varian Unity Plus 400 spectrometer (at 400.4, 100.7
and 376.7 MHz, respectively). Chemical shifts are reported in ppm
downfield from TMS (1H, 13C) or CFCl3 (19F) as internal standards.
IR spectra were obtained on a Hewlett Packard UR 20 spectro-
meter. Mass spectra were recorded on an Agilent 1100 LCMSD SL
instrument by chemical ionization (CI).
4.2. Preparation of 1-iodo-3-(trifluoromethyl)bicyclo[1.1.1]pentane
(4)
If the reaction is carried out for 5 min instead of 15 min, the
yield of 1 decreases to 70%, and another product 10 (10%) could be
isolated as well. It was obtained as a second fraction in the ion
exchange chromatography, following 1.
2 (160 g, 0.54 mol) and absolute pentane (200 mL) were placed
in a 2-L Favorsky apparatus in argon atmosphere. The suspension
was cooled to ꢁ78 8C and 1.6 M solution of MeLi in Et2O (800 mL,
1.28 mol) was added dropwise during 20 min under stirring. The
mixture was allowed to warm to 0 8C and stirred for 1 h at this
temperature. Thereafter, the mixture of Et2O, pentane and
propellane was distilled under reduced pressure into another 2-
L vessel (thick glass) cooled by liquid nitrogen. According to
Mondanaro and Dailey [15], the yield of propellane ranges from
75% to 88% (up to 0.48 mol). 2-L receiver, still cooled by liquid
nitrogen, was disconnected from the Favorsky apparatus, and filled
with CF3I (105 g, 0.54 mol). The vessel was closed with a septum, in
which a syringe with a rubber balloon was injected. Then, the flask
was allowed to warm up to ꢁ20 8C, whereby the balloon was
inflated. The syringe was disconnected, and the septum was
additionally fixed on the flask by a metal clamp. The flask was left
in the dark for 3 days at rt. Next, the solvent was slowly removed in
vacuum on the rotary evaporator without heating (the product is
very volatile). Upon evaporating, the temperature inside the flask
decreased to ꢁ10 8C. It normally took 2–3 h to remove the solvent.
4 was obtained as white solid (87 g, 0.33 mol, 62% yield calculated
on 2). It had to be consumed right away, due to decomposition
upon storage (after 2 weeks in the dark at 0 8C, 4 is completely
decomposed). We normally used 4 the next day after isolation.
4.5. (2S)-{[(1R)-2-hydroxy-1-phenylethyl]amino}[3-
(trifluoromethyl)bicyclo[1.1.1]pent-1-yl]acetic acid (10)
m.p. 217–219 8C. ½a D20
ꢃ
= ꢁ4.8 (c = 11 mg/mL, CH3OH).
: 7.51 (m, 5H, Ph), 4.32 (dd, J = 7.6,
1H NMR (400 MHz, D2O),
d
5.0 Hz, 1H, CHCH2), 3.84 (dd, J = 12.0, 5.0 Hz, 1H, CHCHH), 3.78 (dd,
J = 12.0, 7.6 Hz, 1H, CHCHH), 3.73 (s, 1H, CHCOOH), 1.95 (s, 6H,
(CH2)3).
19F NMR (377 MHz, D2O),
13C NMR (100 MHz, D2O),
d
: ꢁ75.50 (s, CF3).
d
: 170.52 (s, COOH), 132.79 (s, tert-C,
Ph), 128.07 (s, CH, Ph), 127.96 (s, CH, Ph), 125.79 (s, CH, Ph), 121.34
(q, 1JC–F = 271.0 Hz, CF3), 61.44 (s, CH2Ph), 55.05 (s, CHCOOH), 53.41
(s, CH2OH), 46.39 (q, 3JC–F = 2.0 Hz, (CH2)3), 36.09 (q, 4JC–F = 2.0 Hz,
2
CH2CCH), 34.59 (q, JC–F = 39.0 Hz, CCF3).
MS (m/z): 330 (M+1).
Acknowledgements
PM is very grateful to Dr. Magda Tsapko for NMR measure-
ments, and all authors thank the Alexander von Humboldt
Foundation for supporting their institute partnership.

220
P.K. Mykhailiuk et al. / Journal of Fluorine Chemistry 131 (2010) 217–220
[7] D. Tiltak, PhD Thesis, University of Karlsruhe, 2009.
References
[8] P.K. Mykhailiuk, S. Afonin, G.V. Palamarchuk, O.V. Shishkin, A.S. Ulrich, I.V.
Komarov, Angew. Chem. 120 (2008) 5849–5851; Angew. Chem. Int. Ed. 45
(2008) 5765–5767.
[1] P.K. Mykhailiuk, S. Afonin, A.N. Chernega, E.B. Rusanov, M.O. Platonov, G.G.
Dubinina, M. Berditsch, A.S. Ulrich, I.V. Komarov, Angew. Chem. 118 (2006)
5787–5789; Angew. Chem. Int. Ed. 45 (2006) 5659–5661.
[2] R.W. Glaser, C. Sachse, U.H.N. Du¨rr, P. Wadhwani, A.S. Ulrich, J. Magn. Reson. 168
(2004) 153–163.
[3] R.S.W. Gla¨zer, C. Sachse, U.H.N. Du¨ rr, P. Wadhwani, S. Afonin, E. Strandberg, A.S.
Ulrich, Biophys. J. 88 (2005) 3392–3397.
[4] S.L. Grage, A.V. Suleymanova, S. Afonin, P. Wadhwani, A.S. Ulrich, J. Magn. Reson.
203 (2006) 77–86.
[9] B. Ehni, Diploma Thesis, University of Karlsruhe, 2009.
[10] M. Salwiczek, B. Koksch, manuscript in preparation.
[11] P. Wadhwani, E. Strandberg, in: I. Ojima (Ed.), Fluorine in Medicinal Chemistry
and Chemical Biology, Wiley-Blackwell, 2009, pp. 463–493.
[12] A. Eisele, Diploma Thesis, Karlsruhe University, 2007.
[13] J.L. Adcock, A.A. Gack, J. Org. Chem. 57 (1992) 6206–6210.
[14] K. Mondanaro Lynch, W.P. Dailey, J. Org. Chem. 60 (1995) 4666–4668.
[15] K.R. Mondanaro, W.P. Dailey, Org. Synth. 10 (2004) 658; Org. Synth. Coll. 75
(1998) 98.
[5] S. Afonin, P.K. Mykhailiuk, I.V. Komarov, A.S. Ulrich, J. Pept. Sci. 13 (2007) 614–623.
[6] P.K. Mykhailiuk, PhD Thesis, University of Karlsruhe, 2008.