Separation of fluorinated amino acids and oligopeptides from their non-fluorinated counterparts using high-performance liquid chromatography

Article abstract of DOI:10.1016/j.jfluchem.2009.11.025

Chromatographic conditions for the separation of fluorinated amino acids and oligopeptides from their non-fluorinated counterparts were explored. The separation of six pairs of analytes, including both aromatic and aliphatic fluorocarbons, was investigate

Full text of DOI:10.1016/j.jfluchem.2009.11.025

Journal of Fluorine Chemistry 131 (2010) 439–445
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Short communication
Separation of fluorinated amino acids and oligopeptides from their
non-fluorinated counterparts using high-performance liquid chromatography
Nu Xiao a, Y. Bruce Yu a,b,
*
a Department of Pharmaceutical Sciences, University of Maryland, Baltimore, MD 21201, United States
b Fischell Department of Bioengineering, University of Maryland, College Park, MD 20742, United States
A R T I C L E I N F O
A B S T R A C T
Article history:
Chromatographic conditions for the separation of fluorinated amino acids and oligopeptides from their
non-fluorinated counterparts were explored. The separation of six pairs of analytes, including both
aromatic and aliphatic fluorocarbons, was investigated at various temperatures using both hydrocarbon
and fluorocarbon columns and eluents. Our results show that when hydrocarbon eluents are used,
fluorocarbon column provides better separation of fluorinated amino acids or oligopeptides from their
non-fluorinated counterparts; when fluorocarbon eluents are used, hydrocarbon column provides better
separation of fluorinated amino acids or oligopeptides from their non-fluorinated counterparts. These
chromatographic behaviors reflect the fluorophilicity possessed by fluorinated amino acids and
oligopeptides.
Received 19 October 2009
Received in revised form 3 November 2009
Accepted 30 November 2009
Available online 4 December 2009
Keywords:
Fluorinated amino acids
Fluorinated peptides
Fluorocarbon eluent
Fluorocarbon column
HPLC
ß 2009 Elsevier B.V. All rights reserved.
Fluorophilicity
The vast majority of fluorinated drugs contain just a few
fluorine atoms [3,4]. In this work, we explore the separation of
lightly fluorinated amino acids and oligopeptides from their non-
1. Introduction
Although rarely existing in nature [1], organofluorine com-
pounds are finding increasing applications in a wide range of
biological and medical sciences, such as biochemistry [2],
medicinal chemistry [3], pharmaceutical chemistry [4], green
chemistry [5], biotechnology [6], drug delivery [7] and diagnostic
imaging [8]. As is the case with any class of organic molecules,
separation is an important consideration for organofluorine
compounds [9]. Heavily fluorinated molecules have unique
partition properties between fluorocarbon solvents and hydrocar-
bon solvents [10]. This feature has been exploited for the
extraction and separation of compounds with multiple fluorine
atoms, using either perfluorocarbon solvent extraction [11] or
fluorous silica-gel chromatography [12].
fluorinated counterparts using high-performance liquid chroma-
tography (HPLC). This problem arises during our work on
developing fluorinated analogs of peptide drugs. We investigated
several aspects of the separation of fluorinated amino acids and
oligopeptides from their non-fluorinated counterparts, including:
aromatic vs. aliphatic fluorocarbons, fluorocarbon (F-) vs. hydro-
carbon (H-) columns, fluorocarbon (F-) vs. hydrocarbon (H-)
eluents, and temperature. Both fluorocarbon columns and
fluorocarbon eluents have been investigated for the separation
of lightly fluorinated compounds [13–15]. However, to the best of
knowledge, there has been no prior report in which fluorinated
eluents, columns and analytes are compared side-by-side with
their non-fluorinated counterparts.
We investigated six pairs of analytes, 2/1, 4/3, 6/5, 8/7, 10/9 and
12/11, which can be divided into three groups: the aromatic
fluorocarbon group (C6H5 ! C6H4F substitution), which includes
the 2/1 and the 4/3 pairs; the aliphatic fluorocarbon group
(CH3 ! CF3 substitution), which includes the 6/5 and the 8/7 pairs;
and the hydrocarbon control group (H ! CH3 substitution), which
includes the 10/9 and the 12/11 pairs. 8 stems from our effort on
making fluorinated analogs of the peptide drug octreotide
(Sandostatin1). Fig. 1 shows the structures of the 12 analytes.
For the separation of each pair of analytes, we used an F-column
that contains the n-C8F17 group and an H-column that contains the
n-C8H17 group. For each column, we used two fluorocarbon
eluents, trifluoroethanol (TFE) and hexafluoroisopropanol (HFIP),
Abbreviations: Cys, cysteine; DCM, dichloromethane; DIC, N,N0-diisopropylcarbo-
diimide; DIPEA, N,N-diisopropylethylamine; DMF, N,N-dimethylformamide; EtOH,
ethanol; F%wt, fluorine weight percentage in an analyte or a solvent; Fmoc,
fluorenylmethoxycarbonyl; HFIP, hexafluoroisopropanol; HOBt, N-hydroxybenzo-
triazole; HPLC, high-performance liquid chromatography; ISP, isopropanol; Lys,
lysine; MS, mass spectrometry; MW, molecular weight; Nle, norleucine; NMR,
nuclear magnetic resonance; Nva, norvaline; Phe, phenylalanine; RPLC, reversed-
phase liquid chromatography; tBu, tert-butyl; TFA, trifluoroacetic acid; TFE,
trifluoroethanol; tfT, trifluorothreonine; Thrallo, allo-
L
-threonine; tR, retention time;
Trp, tryptophan; Trt, trityl; Tyr, tyrosine;
difference.
d
, chemical shift; tR, retention time
D
* Corresponding author at: Department of Pharmaceutical Sciences, University of
Maryland, Baltimore, MD 21201, United States. Tel.: +1 410 705 7514.
E-mail address: byu@rx.umaryland.edu (Y.B. Yu).
0022-1139/$ – see front matter ß 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2009.11.025

440
N. Xiao, Y.B. Yu / Journal of Fluorine Chemistry 131 (2010) 439–445
Fig. 1. Structures of analytes. The difference between a fluorinated analyte and its non-fluorinated analyte is highlighted in red. The 12 analytes can be divided into three groups:
the C6H5 ! C6H4F group (including the 2/1 pair and the 4/3 pair); the CH3 ! CF3 group (including the 6/5 pair and the 8/7 pair); and the H ! CH3 group (including the 12/11 pair
and the 10/9 pair). We used protected version of the amino acids because free amino acids were not sufficiently retentive on the HPLC columns we used. Peptide sequences are:
Trp-Phe (3), Trp-Phe(4-F) (4),
molecular disulfide bond, as in octreotide.(For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
D-Phe-c[Cys-Tyr-D D-Phe-c[Cys-Tyr-D-Trp-Lys-tfT-Cys]-tfT-amide (8). 7 and 8 are cyclized through intra-
-Trp-Lys-Thrallo-Cys]-Thrallo-amide (7) and

N. Xiao, Y.B. Yu / Journal of Fluorine Chemistry 131 (2010) 439–445
441
Fig. 2. Structures of stationary phase (columns) and mobile phase (eluents).
and their respective hydrocarbon counterparts, ethanol (EtOH) and
isopropanol (ISP). For each eluent, chromatographic runs were
conducted at temperatures ranging from 5 8C to 60 8C, at 5 8C
increments. Fig. 2 shows structures of column stationary phases
and structures of the eluents.
According to this criterion, all four fluorinated analytes are more
hydrophobic than their non-fluorinated counterparts. However,
the difference in hydrophobicity caused by fluorination, gauged
by retention time difference
D
tR = tR(fluorinated) ꢀ tR(non-fluo-
rinated), varied considerably among the analytes even after
normalization by the number of fluorine atoms or fluorine weight
percentage (Table 1). This demonstrates that the extent of
hydrophobicity elevation via fluorination depends on molecular
context, which can outweigh the number of H ! F substitutions.
This is shown by the 6/5 and 8/7 pairs: even though 6 contains only
one tfT residue (one CH3 ! CF3 substitution) while 8 contains two
2. Results and discussion
2.1. Hydrophobicity analysis
Fig. 3 shows the retention behavior of analytes 1–12 in
analytical RPLC. All four fluorinated compounds (2, 4, 6 and 8)
were more retentive than their respective non-fluorinated
counterparts (1, 3, 5 and 7). The retention time, tR, provides a
measure of the relative hydrophobicity of the analytes [16], as
confirmed by the two control hydrocarbon pairs (10/9 and 12/11).
tfT residues (two CH3 ! CF3 substitutions),
DtR is larger for the 6/5
pair than for the 8/7 pair, with or without fluorine content
normalization (Table 1). Clearly, hydrophobicity elevation
through fluorination is not determined solely by the number of
H ! F substitutions.
Fig. 3. Hydrophobicity analysis via RPLC. Analytes 1–10 were co-injected in one batch and analytes 9–12 were co-injected in another batch. Analytes 1 and 5 co-eluted under
this condition.

442
N. Xiao, Y.B. Yu / Journal of Fluorine Chemistry 131 (2010) 439–445
Table 1
Characteristics of analytes.
n(F)a
F%wtb
D
tR (min)
D
tR/n(F) (min)
DtR/F%wt (min)
c
Analytes
MW (Da)
1
2
387
405
0
1
0
3.91
6.13
3.91
6.13
7.37
2.45
N/A
N/A
0.83
1.20
1.75
1.48
N/A
N/A
4.7
3
4
351
369
0
1
0
5.1
5
6
397
451
0
3
0
22.11
14.70
9.35
12.6
7
8
1048
1156
0
6
0
9.9
9
297
311
0
0
0
0
10
11
12
339
353
0
0
0
0
13.77
a
Number of fluorine atoms in an analyte.
b
c
Fluorine weight percentage in an analyte (=19 ꢁ n(F)/MW).
Retention time difference between an analyte pair in RPLC (Kromasil-C18 column).
2.2. Separation analysis
separation), the sign of DtR (order of elution), and the temperature
dependency of
The effect of columns. For a given pair of analytes, the choice of
column, H- or F-, has some effect on the magnitude of tR, but no
effect of the sign of tR. The only exception is the 8/7 pair (Fig. 7):
DtR.
Separation profiles of the six analytes pairs, 2/1, 4/3, 6/5, 8/7,
10/9 and 12/11 are shown in Figs. 4–9, respectively. Our analyses of
D
their separation profiles focus on the magnitude of
D
tR (degree of
D
Fig. 4. Separation analysis of the 2/1 pair.
D
tR = tR(2) ꢀ tR(1). Left panel: H-column; right panel: F-column. (*) Water/EtOH eluent system; (*) water/TFE eluent system; (&)
water/ISP eluent system; (&) water/HFIP eluent system; (~) water/TFE:HFIP (1:1) eluent system.
Fig. 5. Separation analysis of the 4/3 pair.
D
tR = tR(4) ꢀ tR(3). Left panel: H-column; right panel: F-column. (*) Water/EtOH eluent system; (*) water/TFE eluent system; (&)
water/ISP eluent system; (&) water/HFIP eluent system; (~) water/TFE:HFIP (1:1) eluent system.

N. Xiao, Y.B. Yu / Journal of Fluorine Chemistry 131 (2010) 439–445
443
Fig. 6. Separation analysis of the 6/5 pair.
D
tR = tR(6) ꢀ tR(5). Left panel: H-column; right panel: F-column. (*) Water/EtOH eluent system; (*) water/TFE eluent system; (&)
water/ISP eluent system; (&) water/HFIP eluent system; (~) water/TFE:HFIP (1:1) eluent system.
when TFE was used as eluent on H-column, the sign of
reversed as the temperature increased from 5 8C to 60 8C. Even in
D
tR was
dramatic effect is with the aromatic fluorocarbon group (2/1 and 4/
3): when the eluent was switched from H- to F-solvents, the sign of
this case (8/7, TFE), the magnitude of
D
tR remained small after sign
DtR was reversed. As a general trend, DtR decreased steadily with
reversal.
the MW and F%wt of eluents for the two fluorocarbon groups (2/1
and 4/3; 6/5 and 8/7), regardless of columns and temperature. In
contrast, for the hydrocarbon group (10/9 and 12/11), DtR
The effect of eluents. For a given pair of analytes, the choice of
eluents has significant, even dramatic, effects on tR. The most
D
Fig. 7. Separation analysis of the 8/7 pair.
D
tR = tR(8) ꢀ tR(7). Left panel: H-column; right panel: F-column. (*) Water/EtOH eluent system; (*) water/TFE eluent system; (&)
water/ISP eluent system; (&) water/HFIP eluent system; (~) water/TFE:HFIP (1:1) eluent system.
Fig. 8. Separation analysis of the 10/9 pair.
D
tR = tR(10) ꢀ tR(9). Left panel: H-column; right panel: F-column. (*) Water/EtOH eluent system; (*) water/TFE eluent system;
(&) water/ISP eluent system; (&) water/HFIP eluent system; (~) water/TFE:HFIP (1:1) eluent system.

444
N. Xiao, Y.B. Yu / Journal of Fluorine Chemistry 131 (2010) 439–445
Fig. 9. Separation analysis of the 12/11 pair.
D
tR = tR(12) ꢀ tR(11). Left panel: H-column; right panel: F-column. (*) Water/EtOH eluent system; (*) water/TFE eluent system;
(&) water/ISP eluent system; (&) water/HFIP eluent system; (~) water/TFE:HFIP (1:1) eluent system.
increased when the eluent is switched from H- to F-eluents when
H-column is used.
columns, compared with its non-fluorinated counterpart when F-
eluents are used. Third, the two aromatic fluorocarbon analytes, 2
and 4, demonstrate greater affinity toward F-eluents than toward
F-column. When the eluents were switched from H- to F-, the sign
The effect of temperature. Temperature has various effects on the
magnitude of
DtR with no clear pattern. However, with the
exception of the 8/7 pair on H-column with TFE as eluent,
of
DtR (elution order) of the 2/1 and the 4/3 pairs was reversed in
temperature has no impact on the sign of
we investigated.
D
tR for any analyte pair
most cases, even when F-column was used, indicating that 2 and 4
prefer F-eluents over F-column.
Fluorocarbons vs. hydrocarbons. Even though analytes 2, 4, 6, 8,
10 and 12, are all more hydrophobic than their respective
counterparts, 1, 3, 5, 7, 9 and 11, there are significant differences
between hydrocarbons and fluorocarbons. For the two hydrocar-
bon pairs (10/9 and 12/11), the more hydrophobic ones (10 and 12)
were always eluted later than their less hydrophobic counterparts
The control hydrocarbon compounds demonstrate the expected
aversion toward fluorocarbons (fluorophobicity). For these two
analyte pairs, 10/9 and 12/11, when H-eluents were used,
DtR
decreased as the column was switched from H- to F-, showing
aversion toward F-column by the extra –CH2–; when H-column
was used, DtR increased as the eluents were switched from H- to F-,
(9 and 11), i.e.,
D
tR > 0 under all conditions. However, for the four
showing aversion toward F-eluents by the extra –CH2–.
fluorocarbon pairs (2/1, 4/3, 6/5 and 8/7), the more hydrophobic
ones (2, 4, 6 and 8) sometimes were eluted earlier than their less
3. Conclusion
hydrophobic counterparts (1, 3, 5 and 7), i.e.,
DtR < 0 under certain
conditions. Such reversal of elution order happened to every
fluorinated analyte pair on both H- and F-columns. The common
feature is that in each case where the elution order was reversed, F-
eluents were used.
Aromatic vs. aliphatic fluorocarbons. Among fluorocarbons, there
are also differences between aromatic fluorocarbons and aliphatic
fluorocarbons. Elution order reversal happens much more readily
for aromatic fluorocarbons than for aliphatic fluorocarbons
Chromatographic separation is a balance of complex inter- and
intra-molecular interactions involving analytes, eluents, and
columns. Generally speaking, optimal separation requires judi-
cious choice of eluents, column and temperature. For a pair of
lightly fluorinated amino acid or oligopeptide and its non-
fluorinated counterpart, eluents have much more pronounced
effect on their separation than columns and temperature. When
hydrocarbon eluents are used, fluorocarbon column provides
better separation of fluorinated amino acids or oligopeptides from
their non-fluorinated counterparts; when fluorocarbon eluents are
used, hydrocarbon column provides better separation of fluorinat-
ed amino acids or oligopeptides from their non-fluorinated
counterparts. These chromatographic behaviors reflect the fluor-
ophilicity possessed by fluorinated amino acids and oligopeptides.
For the two pairs of hydrocarbon analytes (10/9 and 12/11), the
best separation condition is achieved with H-column and the
water/trifluoroethanol eluent system. This improved separation in
comparison with the water/ethanol eluent system is apparently
caused by the aversion of hydrocarbons toward F-eluents.
(Figs. 4–7). Further, after reversal, the magnitude of
DtR was
much larger for aromatic fluorocarbons than for aliphatic
fluorocarbons. This is an interesting observation considering that
F-eluents themselves are aliphatic fluorocarbons.
2.3. Fluorophilicity of lightly fluorinated compounds
It is well known that heavily fluorinated compounds have
affinity toward each other, the so-called fluorophilicity [17]. Our
work shows lightly fluorinated amino acids and oligopeptides also
possess fluorophilicity. First, fluorinated analytes 2, 4, 6 and 8
demonstrate affinity toward F-column. When H-eluents were
used, the magnitude of
D
tR increased as the column was switched
4. Experimental
from H- to F- (Figs. 4–7), indicating that a fluorinated analyte
becomes more retentive on the F-column compared with its non-
fluorinated counterpart. Second, fluorinated analytes 2, 4, 6 and 8
demonstrate affinity toward F-eluents. When the eluents were
4.1. Chemicals, materials and instrumentation
Amino acids and resin. All amino acids have the L-configuration
(i.e., 2S) unless otherwise specified. (2S, 3S)-N-Fmoc-O-tert-butyl-
threonine (5) was purchased from BACHEM. (2S, 3R)-4,4,4-
Trifluoro-N-Fmoc-O-tert-butyl-threonine (6) was synthesized
switched from H- to F-,
eluents (Figs. 4–7), regardless of column, indicating that
fluorinated analyte becomes less retentive on both H- and F-
DtR decreased steadily with the F%wt of the
a

N. Xiao, Y.B. Yu / Journal of Fluorine Chemistry 131 (2010) 439–445
445
and purified by us as described in a previous publication [18].
Fmoc-norvaline (11) and Fmoc-norleucine (12) were purchased
from Aapptec. Other Fmoc-protected amino acids and rink amide
MBHA resin (0.65 mmol/g, 100–200 mesh) were purchased from
Novabiochem. All purchased amino acids were used without
further purification.
Eluents. EtOH was from Sigma–Aldrich (spectrophotometric
grade); ISP was from EMD (HPLC grade); TFE and HFIP were from
Oakwood Products (reagent grade). EtOH and ISP were used as
purchased. TFE and HFIP were distilled before usage.
hydrophobicity of peptides and amino acids [14]. The chro-
matographic conditions were: eluent A: 0.2% TFA in water; eluent
B: 0.2% TFA in CH3CN; gradient: 0.25% B/min, starting from 10% B;
flow rate: 0.3 mL/min; column temperature: 25 8C. Room temper-
ature was set at 20 8C. To avoid overcrowding the chromatogram,
injections were conducted in two batches: analytes 1–10 were co-
injected and analytes 9–12 were co-injected.
For separation analysis, the chromatographic conditions were:
eluent A: 0.1% TFA in water; eluent B: 0.1% TFA in EtOH, or TFE, or
ISP, or HFIP, or TFE/HFIP mixture (1:1); gradient: 1% B/min; flow
rate: 0.5 mL/min; column temperature: 5–60 8C. For column
temperatures below 20 8C, room temperature was set at 5 8C.
For column temperatures between 20 8C and 60 8C, room
temperature was set at 20 8C. Each pair of analytes was co-
injected. For detailed gradient conditions of each pair, see Table S1
of Supporting Information.
Columns. For separation studies, H-column, Zorbax Eclipse XDB-
C8 (4.6 mm ꢁ 150 mm, 5
m
m); F-column, FluoroFlash1 (4.6 mm ꢁ
150 mm, mm). For hydrophobicity analysis, Kromasil-C18
5
(2.1 mm ꢁ 150 mm, 5
(4.6 mm ꢁ 250 mm, 5
(4.6 mm ꢁ 250 mm, 5
m
m
m
m). For purity analysis, Zorbax XDB-C18
m). For chiral purity analysis, ChiraDex
m). For peptide purification, Zorbax C8
m).
preparative column (21.2 mm ꢁ 250 mm, 5
m
Instrumentation. HP1200 liquid chromatography system
Acknowledgements
(Agilent Technologies); JEOL ECX 9.4T NMR spectrometer
(
19F 367 MHz); LCQ Man-O.2.2 mass spectrometer.
This research was supported by NIH grant EB 004416 and the
Kimmel Foundation. YBY was a Kimmel scholar.
4.2. Peptide synthesis and purification
Peptides (3, 4, 7 and 8) were made using Fmoc solid-phase
chemistryonrinkamideMBHAresin. ExceptCys andtfT, allcoupling
reactions were conducted using the Liberty microwave peptide
synthesizer(5eq. Fmoc-AA-OH, 4.5eq.DIC, 5eq. HOBt, 5eq. DIPEAin
DMF, 75 8C, 200 s). Cys and tfT are prone to racemization and their
manual incorporation into peptides used the following conditions:
Cys, 5 eq. Fmoc-Cys(Trt)-OH, 4.5 eq. DIC, 5 eq. HOBt, no base, r.t., 3 h;
tfT, 3 eq. Fmoc-tfT(OtBu)-OH, 2.7 eq. DIC, 3.3 eq. HOBt, no base, 0.5
eq. CuCl2ꢂ2H2O in DMF/DCM (1/1), 0 8C, 16 h. CuCl2 was added to
reduce racemization [18]. Cyclic peptides (7 and 8) were made by
crosslinking the two cysteine residues in each peptide (intra-
molecular S-S bond) in 10 mM ammonium acetate aqueous solution
containing 17% DMSO (pH 7.0).
The peptides were purified using preparative reversed-phase
liquid chromatography (RPLC). The purity of each peptide was
verified using analytical reversed-phase and chiral liquid chroma-
tography. The molecular weight (MW) of each peptide was verified
using mass spectrometry (MS). The fluorinated octapeptide (8) was
further characterized by 19F NMR spectroscopy. For HPLC, MS and
NMR data on 3, 4, 7 and 8, see Supporting Information.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jfluchem.2009.11.025.
References
[1] D. O’Hagan, D.B. Harper, J. Fluorine Chem. 100 (1999) 127–133.
[2] B.J. Stockman, J. Am. Chem. Soc. 130 (2008) 5870–5871.
[3] S. Purser, P.M. Moore, S. Swallow, V. Gouverneur, Chem. Soc. Rev. 37 (2008) 320–
330.
[4] K. Mu¨ ller, C. Faeh, F. Diederich, Science 317 (2007) 1881–1886.
[5] I.T. Horva´th, Green Chem. 10 (2008) 1024–1028.
[6] S.M. Brittain, S.B. Ficarro, A. Brock, E.C. Peters, Nat. Biotechnol. 4 (2005) 463–468.
[7] H.-J. Lehmler, Expert Opin. Drug Deliv. 4 (2007) 247–262.
[8] E.G. Schutt, D.H. Klein, R.M. Mattrey, J.G. Reiss, Angew. Chem. Int. Ed. 42 (2003)
3218–3235.
[9] D.P. Curran, Synth. Lett. 9 (2001) 1488–1496.
[10] D.P. Curran, Z. Lee, Green Chem. 3 (2001) G3–G7.
[11] A. Studer, S. Hadida, R. Ferritto, S.-Y. Kim, P. Jeger, P. Wipf, D.P. Curran, Science 275
(1997) 823–826.
[12] Z. Luo, Q. Zhang, Y. Oderaotoshi, D.P. Curran, Science 291 (2001) 1766–1769.
[13] M. Przybyciel, LC–GC 23 (2005) 554–565.
[14] W. Zhang, J. Fluorine Chem. 129 (2008) 910–919.
[15] K. Valko´, S. Espinosa, C.M. Du, E. Bosch, M. Rose´s, C. Bevan, M.H. Abraham, J.
Chromatogr. A 933 (2001) 73–81.
4.3. HPLC
[16] J.M. Kovacs, C.T. Mant, R.S. Hodges, Biopolymers (Pept. Sci.) 84 (2006) 283–297.
[17] F.T.T. Huque, K. Jones, R.A. Saunders, J.A. Platts, J. Fluorine Chem. 115 (2002)
119–128.
For hydrophobicity analysis, we followed the method devel-
oped by Hodges and co-workers for measuring the relative
[18] N. Xiao, Z.-X. Jiang, Y.B. Yu, Biopolymers (Pept. Sci.) 88 (2007) 781–796.