Retention mechanisms of acidic and basic analytes on the Pentafluorophenyl stationary phase using fluorinated eluent additives

Journal of Chromatography A 1666 (2022) 462850

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Tutorial Article

Retention mechanisms of acidic and basic analytes on the
Pentafluorophenyl stationary phase using fluorinated eluent additives
Krit Lossmann a, Ruta Hecht a, Jaan Saame a, Agnes Heering a, Ivo Leito a, Karin Kipper a,b,c,∗
a

University of Tartu, Institute of Chemistry, 14a Ravila Street, 50411 Tartu, Estonia
Chalfont Centre for Epilepsy, Chesham Lane, Chalfont St Peter, Buckinghamshire, SL9 0RJ, United Kingdom
c
Department of Clinical and Experimental Epilepsy, Faculty of Brain Sciences, UCL Queen Square Institute of Neurology, Queen Square, London WC1N 3BG,
University College London, United Kingdom
b

a r t i c l e

i n f o

Article history:
Received 11 November 2021
Revised 20 January 2022
Accepted 23 January 2022
Available online 28 January 2022
Keywords:

Pentafluorophenyl column
HFTB, HFIP
Perfluoropinacol
LC-MS

a b s t r a c t
This work explores the effects of three selected fluoroalcohols - 1,1,1,3,3,3-hexafluoroisopropanol (HFIP),
1,1,1,3,3,3-hexafluorotert–butyl alcohol (HFTB) and hexafluoro-2,3-(trifluoromethyl)-2,3-butanediol (PP) as
novel eluent additives and their effect on the retention of basic and acidic analytes, using a reversed
phase (RP) column with a fluorophenyl (PFP) stationary phase. In order to observe the changes in the
model analytes’ retention, chromatograms were obtained at multiple (5.0; 6.0; 7.0; 8.5; 9.0 and 9.5) pH
values depending on the eluent. The retention observed with fluoroalcohols was compared with that of a
conventional eluent additive - ammonium acetate. When fluoroalcohols were used as eluent additives, a
decrease in the retention factors (compared with ammonium acetate) was generally observed for strong
acids. The retention factors of strong bases were generally higher when using HFIP and HFTB as eluent
additives. The behaviour of weak bases and weak acids was more nuanced, potentially enabling interesting selectivity. The extent of the effect regarding different fluoroalcohols also varied, with HFIP and
HFTB having a more significant effect on the retention of analytes than PP. The retention data were interpreted in terms of the hypothesis that four interactions are at play: (a) hydrophobic retention typical
to RP; (b) π -π interactions between the analytes containing an aromatic ring and the aromatic rings on
the stationary phase; (c) charge-charge or hydrogen bond interactions between the analytes and partially
deprotonated fluoroalcohols adsorbed on the stationary phase and (d) a hydrogen bond or charge-charge
interaction between the free silanol groups or their deprotonated forms on the stationary phase and the
analytes (either neutral or ionic). Alternative selectivity obtained through fluoroalcohols on the PFP stationary phase was compared with the C18 and biphenyl stationary phases. It was demonstrated that at
the same eluent pH but with a different buffer system and/or different RP stationary phases, very different selectivity and retention order can be obtained.
© 2022 The Author(s). Published by Elsevier B.V.
This is an open access article under the CC BY license ( />
1. Introduction
The most common stationary phase in reversed phase (RP) liquid chromatography, also in the field of bioanalytical applications,
continues to be C18. However, other non-polar stationary phases
are gaining popularity due to the offered alternative selectivity.
One such stationary phase is pentafluorophenyl (PFP) [1]. So far,

RP, reversed phase; PFP, pentafluorophenyl; HFIP, 1,1,1,3,3,3-hexafluoro-2propanol; HFTB, 1,1,1,3,3,3-hexafluoro-tert-butyl alcohol; PP, hexafluoro-2,3(trifluoromethyl)-2,3-butanediol, perfluoropinacol; MS, mass-spectrometric/massspectrometry; SST, system suitability test; LC, liquid chromatography; HB, hydrogen
bond; TOC, total organic carbon; PDA, photodiode array detector.
∗
Corresponding author.
E-mail address: (K. Kipper).

PFP has demonstrated usual RP retention patterns for neutral and
acidic analytes [2]. However, changes in the retention of the basic analytes indicate additional interactions. It has been speculated that these changes in analyte retention could be caused by
ionic interactions between the free silanol groups on the stationary phase and the basic analytes [1,2].
It is estimated that almost 95% of active pharmaceutical ingredients contain some ionisable functional group and that almost
75% of them are weakly basic [3]. Analytes with acidic or basic
properties may be partially or completely in a neutral or an ionic
form, depending on the pH of the mobile phase. If analytes are
partially in an ionic form, already a small change in the pH can
lead to a considerable change in the retention of analytes. Analytes
in a completely ionic form are polar and often elute too rapidly

/>0021-9673/© 2022 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license ( />

K. Lossmann, R. Hecht, J. Saame et al.

Journal of Chromatography A 1666 (2022) 462850
Table 1
Researched fluoroalcohols and their structures.

from the column [4]. In the case of basic analytes, this means that
mobile phases with basic pH might be preferred. For such mobile phases, one would need acidic and/or basic buffer components
with pKa values in the range of 7 to 10. [4]
Adding an organic solvent to the buffer solution can change the

pKa values of the analytes and buffer components, thus also changing the real pH of the mobile phase. In order to enable comparing
pH values of different mobile phases, an unified pH scale has been
established [5], which was also used in this work to estimate the
change of the aqueous pH values after adding methanol.
Using mass-spectrometric (MS) detectors allows to determine a
very small analyte quantity in a sample, but it also requires for
all of the eluent components to be volatile and not suppress analyte ionisation [6]. One such group of compounds, which can be
used as eluent additives and buffer components, are fluoroalcohols.
They are volatile, provide buffering capacity in the basic pH range
with ammonia as well as alternative selectivity [7,8]. 1,1,1,3,3,3hexafluoro-2-propanol (HFIP) and 1,1,1,3,3,3-hexafluoro–tert–butyl
alcohol (HFTB) have been shown to improve the separation and
ionisation of oligonucleotides [9,10], antibiotics [11] and sedative
drugs and their metabolites [7]. However, 1,1,1,4,4,4-hexafluoro2,3-bis(trifluoromethyl)butane-2,3-diol (perfluoropinacol, PP) has,
to the best of our knowledge, not been researched as an eluent
additive outside our group [12] and it has been possible to achieve
different retention from ammonium acetate when PP is used as
an eluent additive. Furthermore, using a buffer system of perfluoropinacol and ammonium ions provides a system with a wide
and continuous buffering range in the pH range 5.0–11.5. The proposed additional retention mechanism of using novel fluorinated
eluent additives is that the fluoroalcohols (which can partially deprotonate) are retained on the stationary phase (potentially forming a fluorous layer) and thus can form ion pairs with the analytes
[8]. Fluoroalcohols as eluent additives and PFP stationary phase
have both shown potential as alternatives for more common additives and stationary phases, respectively, to enhance selectivity.
This work aims to broaden the selection of LC-MS compatible eluent additives which could be used for basic analytes.
In this work, the authors explored the effect of three fluorinated
alcohols - HFIP, HFTB and perfluoropinacol - as novel eluent additives/buffer components on the retention of basic and acidic model
analytes at different mobile phase pH values using a column with
a PFP stationary phase. To the best of authors’ knowledge, this is
the first article where influence of fluoroalcohols as eluent additives is discussed on PFP stationary phase for small molecules. The
retention of the same analytes with a mobile phase based on ammonium acetate was used as a comparison. The results were also
compared with earlier results obtained when using fluorinated alcohols and columns with the C18 and biphenyl stationary phases.

Name

pKa

1,1,1,4,4,4-hexafluoro-2,3bis(trifluoromethyl)butane-2,3-diol,
perfluoropinacol

5.95;
10.43 [13]

1,1,1,3,3,3-hexafluoro-tert-butyl
alcohol

9.6 [14]

1,1,1,3,3,3-hexafluoroisopropanol

9.3 [15]

Structure

2-metoxypyridine, 2-methylpyridine, naphthylamine, urea, acetone, benzene, toluene and aniline were obtained from SigmaAldrich (Missouri, USA). N–hydroxy-6-bromobenzotriazole, N–
hydroxy-6-trifuoromethylbenzotriazole and N–hydroxy-5–chloro4-methylbenzotriazole were kindly donated by Dr W. Köning
(Hoechst AG, Frankfurt, Germany). The analytes were divided into
4 groups based on their acid-base behaviour: weak and strong
acids and weak and strong bases (Table 2).
The anaylte mixture for system suitability tests (SST) contained
acetone, benzene and toluene.
2.1. Preparation of the stock solution and working solution
Stock solutions were prepared at a concentration of 1 mg/mL in

a water–methanol 20:80 (v:v) mixture. All solutions were stored at
−20 °C. All working solutions were filtered using a 0.2 μm Sartorius Minisart RC 4 syringe filter before transferring the solution to
a vial. The samples were stored in an autosampler for up to 7 days
at 4 °C.
2.2. Chromatographic conditions
The analytical column Raptor Fluorophenyl 2.1 × 100 mm, particle size 2.7 μm, resistant in the pH range 2.0 to 8.0 [21] was
kindly donated by Restek (Pennsylvania, USA). The column was
equilibrated for 60 min, with the buffer used in the following experiments. The elution was isocratic at 25% MeOH with the flow
rate 0.5 mL/min for analytes and 0.4 mL/min for system suitability
test (SST). The column was thermostated at 40 °C. The injection
volume was 10 μL.
The results used for comparison were obtained from earlier
works by Veigure et al. [12,22].
The analysis was performed with the Shimadzu Nexera XR LC20AD HPLC system, PDA detector SPD-M20A and Shimadzu LCMS2020 MS detector. The detector choice depended on the analyte
(see Table S1). The UV/Vis detector was set to record between
190 nm and 700 nm, for chromatograms the extracted wavelength
was (254 ± 2) nm. The reference wavelength was 600 nm with
a bandwidth of 50 nm. The MS was operating in a positive and
negative scanning mode in the m/z range of 60–250. The chromatograms were processed using the Shimadzu LabSolutions version 5.75 SP2.
The eluent consisted of a buffer and methanol 75:25 (v:v) Eluent additive concentrations in the buffer solutions were 5 mM for
all additives. Aqueous buffer solutions were prepared at pH 5.0,
6.0, 7.0, 8.5, 9.0, 9.5 for PP and ammonium acetate, 8.5, 9.0, 9.5
for HFIP and 7.0, 8.5, 9.0, 9.5 for HFTB. Due to the pKa values of
HFTB and HFIP, these compounds lack buffering ability at pH values lower than 7. Therefore, experiments with HFTB and HFIP were
not conducted at lower pH values. Ammonium hydroxide was used
in most cases to modify the buffer pH, except for ammonium ac-

2. Material and methods
Eluent
additives:

ammonium
acetate,
25%
ammonium
hydroxide,
acetic
acid,
1,1,1,4,4,4-hexafluoro-2,3bis(trifluoromethyl)butane-2,3-diol (perfluoropinacol, PP, Table 1),
1,1,1,3,3,3-hexafluoro–tert–butyl alcohol (HFTB) were LC-MS grade
and obtained from Sigma-Aldrich (Missouri, USA), 1,1,1,3,3,3hexafluoroisopropanol (HFIP) was obtained from Fluka (Buchs,
Switzerland). LC-MS grade MeOH used in the mobile phases was
obtained from Sigma Aldrich (Missouri, USA), water was purified
in-house using a Millipore Advantage A10 system (18.2 M cm
at 25 °C and a total organic carbon (TOC) value 2–3 ppb) from
Millipore (Bedford, USA).
Analytes: 4-nitrobenzoic acid, 2,3,4,5,6-pentafluorophenol,
2,4-dichlorophenol, phenol, p-cresol, hydroquinone, isopropylphenol, 3-fluorophenol, 2 -hyrdoxyacetophenone, diisopropylamine,
cyclohexylamine, pyrrolidine, piperidine, 2,6-dimethylpyridine,
2


K. Lossmann, R. Hecht, J. Saame et al.

Journal of Chromatography A 1666 (2022) 462850

etate at pH = 5.0, where acetic acid was taken as the pH modifier
instead.
The pH values measured in the aqueous buffer before adding
the organic phase are denoted as w
w pH [23] further in the text - if

making a distinction was necessary from the pH of the whole mobile phase (where the organic component has been added). This
is because the addition of the organic phase changes the solvated
proton activity and thus also the final mobile phase’s pH. For expressing the pH of the whole mobile phase, we have employed
H2 O
the absolute pH scale (further denoted as pHabs
) values [24]. The
w
aqueous buffer w pH values were measured with an Elmetron EPP1 combination electrode connected to the Evikon pH metre E6115,
which was calibrated daily with standard aqueous buffers at pH
H2 O
values 4.01, 7.00 and 10.00. The mobile phase pHabs
measurements were conducted using the modified version of the 2020 differential potentiometric method described by Heering et al. [5].
Measurements were done with a Metrohm 713 pH metre (with a
Pt wire) and a Keysight B2987A Electrometer (without a Pt wire).
No pre-soaking was done. Data was collected for 60 min at 10 s
intervals and points from 30 min to 60 min were used for the
analysis. Measurements were done only at one polarity. The w
w pH
H2 O
and pHabs
values of the used mobile phases are listed in Table S2
in the Supplementary data. For some of these mobile phases the
H2 O
pHabs
values were reported by us already previously [12], using a
simplified measurement method. The values reported in this work
have been obtained using the more accurate differential potentiometric method and they differ from the earlier values by up to
0.5 pH units, demonstrating the low accuracy of the previous simH2 O
plified method. The pHabs
values from this work should be preferred.

Table 2
Model analytes, their pKa and pKaH values (all values come from the iBonD
database, unless indicated otherwise) and structures. pKaH value corresponds to the
pKa of the protonated base.
Analyte

pKa

Structure

Strong acids
4-nitrobenzoic acid

3.4

2,3,4,5,6-pentafluorophenol

5.5

2,4-dichlorophenol

7.9

N–hydroxy-6-bromobenzotriazole

3.97 [16]

N–hydroxy-6-trifluoromethyl
benzotriazole

3.60 [13]

N–hydroxy-5–chloro-4-methyl
benzotriazole

4.09 [16]

Weak acids
phenol

9.99

p-cresol

10.28

hydroquinone

11.40;
11.65 [17]

2-isopropylphenol

10.53

3-fluorophenol

9.24

3. Results and discussion
3.1. Column stability

2 -hydroxyacetophenone

The recommended w
w pH range for the used PFPcolumn was 2–8.
A broader pH range was tested in this work due to the chemical
properties of fluoroalcohols and to gain a better understanding of
their effect. Using various aqueous phases with a w
w pH above the
recommended range (the w
w pH varied from 8.5 to 9.5) caused the
stationary phase to degrade rapidly, as was to be expected. The influence of the column condition and mobile phase pH on the column was monitored using the SST solution (Figure S1). All analytes in the SST mixture eluted at progressively shorter times over
the use of the column at high w
w pH values with the most extreme
change being toluene whose retention time changed from 3.5 min
to 2.1 min. Increased tailing was also observed with degraded stationary phases. Measured retention factors were confirmed with
unused PFP column after the SST solution had demonstrated the
loss of retention. The column was in use for 65 h at pH = 8.5 and
23 h at pH = 9.0 before system pressure reached the set limit and
H2 O
the column became blocked. In addition, pHabs
measurements re-

10.28 [18]

Strong bases
diisopropylamine

11.05 [19]

cyclohexylamine

10.49 [19]

pyrrolidine

11.27

piperidine

11.22

Weak bases
2,6-dimethylpyridine

6.72

H O

2-methoxypyridine

3.28

2-methylpyridine

5.96

1-naphthylamine

3.92 [20]

aniline

4.62

2
vealed that the pHabs
values for the eluents were higher than the
w pH of the aqueous buffer component of the mobile phase (Table
w
H2 O
S2 in Supplementary data). The pHabs
values of the mobile phases
w
with a w pH in the range of 8.5 to 9.5 ranged from 8.17 to 10.17.
Therefore, no measurements were performed with HFIP at pH 9.5,
and with HFTB only acidic analytes were tested. Additional chromatograms can be seen in Supplementary data.

3.2. Trends observed
The main interactions that can influence retention when using
fluoroalcohols as eluent additives with a PFP column are the following: (a) hydrophobic retention typical to RP, (b) π -π interac3


K. Lossmann, R. Hecht, J. Saame et al.

Journal of Chromatography A 1666 (2022) 462850

Fig. 1. Retention factors of weak bases using different eluent additives at different pH values. Error bars represent standard deviations.

tions between the analytes containing an aromatic ring and aromatic rings on the stationary phase, (c) charge-charge or hydrogen
bond interactions between the analytes and partially deprotonated
fluoroalcohols adsorbed on the stationary phase and (d) a hydrogen bond and/or charge-charge interaction between the free silanol
groups or their deprotonated forms on the stationary phase and
the analytes. The first three are expected to markedly affect retention time, while the main effect of the silanol groups is in the peak
asymmetry, and their effect on retention time is probably smaller.
The average pKa values of silica’s silanol groups in water have
been estimated to be at around 7 [25]. Thus, at first sight, in most
of our mobile phases silanols are predominantly ionised. However, there are three factors that additionally influence ionisation
of SiOH groups: (1) adding organic solvent to aqueous buffer increases the pHabs , (2) adding organic solvent to the aqueous buffer
increases the pKa of silanol groups and (3) the protonated form
of silanol can be additionally stabilised by hydrogen bond (HB)
formation with a base or protonated form of a base can be stabilised by charge-charge interaction and HB formation with a deprotonated silanol. These three factors put together mean that HB
interaction between silanol groups and bases are possible also if
the pH value of the aqueous buffer is above 7. To a large extent,
the same considerations apply also to the fluoroalcohols.

erate, but the free silanol groups on the stationary phase are now
to a large extent ionised and have limited ability to form hydrogen
bonds with analytes.
With PP up to w
w pH 7.0 the analytes were partially ionised, as
with ammonium acetate. As they became more neutral, the retention of analytes increased. In addition, the partially ionised PP
on the stationary phase could have ion-ion interactions with the
ionised analytes. Therefore, the retention factors in this pH region
were higher when using PP rather than when using ammonium acetate. At higher pH values, the retention mechanisms were largely
the same as with ammonium acetate. PP was strongly ionised in
the mobile phase and did not attach to the stationary phase to the
same extent. Using HFTB and HFIP enabled the analytes similar retention mechanisms, thus creating similar retention trends as with

PP.
3.4. Strong bases (pKaH 9.6–11.3)
Strong bases with pKaH 9.6–11.3 were to a large extent in a
cationic (protonated) form throughout the researched pH range.
They demonstrated an increase in the retention factors as the pH
increased (Fig. 2) both when PP and ammonium acetate were used.
This is due to a gradual decrease in the extent of protonation of
the analytes – they become less polar when the mobile phase pH
increases. Throughout the used w
w pH range there is a possibility of
interacting with the free silanol groups: either HB between neutral base and SiOH or charge-charge interaction between the protonated base and SiO– . None of the selected analytes in the strong
bases group had an aromatic ring and thus π -π interactions did
not occur.
With ammonium acetate as the eluent additive, the analytes’
ionisation decreased with the increase in the pH, causing stronger
retention, especially at the w
w pH above 7.0 (Table S4). In the case
of PP, the above considerations hold as well. Additionally, partially
ionised PP can interact with ionised analytes (ion-pairing effect),
which in turn causes a stronger retention of analytes than with
ammonium acetate. Otherwise, the retention mechanisms are similar to those of ammonium acetate. When HFIP or HFTB were used
as mobile phase components, the retention mechanism was expected to be similar to that of PP. However, the analytes’ retention factors were significantly higher with HFIP or HFTB than either with ammonium acetate or PP. This implies that the ion pair
effect might be more prominent with HFIP and HFTB than with PP
indicating strong influence of additives.

3.3. Weak bases (pKaH 3.3–7.0)
The weak bases with pKaH 3.3–7.0 are only partially ionised in
the used pH range and are mostly in a neutral form in the upper part of the used pH range. When using ammonium acetate as
an eluent, the retention factors increased in the w
w pH range 5.0–

7.0 and with w
w pH 8.5 began to decrease (Fig. 1). In the case of PP,
such a trend was observed only in the case of 1-naphthylamine.
The retention factors of other analytes decreased with increasing
the pH of the eluent. The full data set can be seen in Table S5 and
chromatograms in Figure S3.
With ammonium acetate as the eluent additive, up to w
w pH 7.0
the analytes were partially ionised. An increase in the eluent pH
influences analytes to become more neutral and increases retention, as is common with RP LC. In addition, π -π interactions and
hydrogen bonding between the neutral analytes and free silanol
groups on the stationary phase are possible. Starting from w
w pH 8.5,
the analytes are essentially in their neutral forms and the proportion of the neutral form no longer increases with an increase in
the pH. The hydrophobic retention and π -π interactions still op4


K. Lossmann, R. Hecht, J. Saame et al.

Journal of Chromatography A 1666 (2022) 462850

Fig. 2. Retention factors of strong bases using different eluent additives at different pH values. Error bars represent standard deviations.

Fig. 3. Retention factors of weak acids using different eluent additives at different pH values. Error bars represent standard deviations.

3.5. Weak acids (pKa 9.3–11.4)

With PP as the eluent additive, there is a more visible increase
in the retention factors with the increase in the w
w pH up to 7.0

(Fig. 3). A further increase in the mobile phase pH leads to a decrease in the retention factors (Table S3). Using HFTB and HFIP, the
analytes have retention trends similar to PP. Most likely, the initial
increase in retention is due to an increase in the proportion of the
ionised form of PP, which may bind the (still almost fully neutral)
analytes via hydrogen bonds. As in the case of ammonium acetate,
the main reason for the decrease in retention above pH 7.0 is that
the analytes are increasingly in an anionic form.
With all mobile phases the interactions between anionic analytes and free silanol groups as well as π -π interactions are weakened due to the high hydrophilicity of the anions and increasing
deprotonation of the silanol groups. Anionic analytes are repelled
from deprotonated silanol groups.

Weakly acidic analytes with pKa 9.3–11.4 are mostly in a neutral
form throughout the used pH range. Thus, they were expected to
have similar retention factors at every mobile phase pH researched.
However, a more nuanced pattern of retention factor changes was
observed with the changing of the pH, as seen in Fig. 3. The full
data set can be seen in Table S3 and chromatograms can be seen
in Figure S5.
Using ammonium acetate as the eluent additive, the analytes
are mostly in a neutral form up to w
w pH 8.5 and their retention is
essentially constant. The π -π interactions are possible, depending
on the analyte’s structure. At higher pH values, the proportion of
the ionised (anionic) form of the analytes begins to increase and
thus the retention decreases.

5


K. Lossmann, R. Hecht, J. Saame et al.

Journal of Chromatography A 1666 (2022) 462850

Fig. 4. Retention factors of strong acids using different eluent additives at different pH values. Due to 4-nitrobenzoic acid’s very weak retention, this analyte is not
shown in the graph. A – 2,3,4,5,6-pentafluorophenol, B – N–hydroxy-6-bromobenzotriazole, C – N–hydroxy-6-trifluoromethylbenzotriazole, D – N–hydroxy-5–chloro-4methylbenzotriazole. Error bars represent standard deviations.

Fig. 5. Comparison of the logarithms to the base 10 of the retention factors of pyrrolidine (strong base), 2,6-dimethylpyridine (weak base) and phenol (weak acid) with
different stationary and mobile phases.

3.6. Strong acids (pKa 3.4–7.9)

All analytes except 2,4-dichlorophenol, which had the highest
pKa value of the strong acids and thus was deprotonated to a
smaller degree than the others, eluted close to dead time using
HFIP and HFTB as eluent additives.

A decrease in the retention factors was observed with an increase in the pH for the analytes mostly in an anionic form (analytes with pKa < 7.9), as can be seen in Fig. 4. This can be explained with the shift of the acid’s dissociation equilibrium from
the neutral acid HA to the anionic (deprotonated) A– form, which
results in an increase in the analyte’s polarity and decrease in
its retention. The full data set can be seen in Table S2 and chromatograms can be seen in Figure S4.
When using an ammonium acetate buffer with the w
w pH values
up to 7.0, there is a possibility for π -π interactions between the
analytes and the stationary phase. Although at w
w pH values up to
7.0 an appreciable share of silanol groups are in neutral form, hydrogen bonds between silanols and anions are unlikely to form due
to the high hydrophilicity of the anions. In addition, the deprotonated silanols will repel the ionic analytes. At higher w
w pH values
(8.5–9.5), the analytes are essentially fully ionised and retention is
weak.

With PP as the eluent additive, a similar trend was observed
as with ammonium acetate (see the paragraph above). In the w
w pH
range 5.0–7.0, the analytes’ retention was weaker than with ammonium acetate because partially anionic PP can be present on
the stationary phase and repel anionic analytes. At higher w
w pH values (8.5–9.5), the analytes are almost fully ionised and retention is
weak.

4. Comparison of the PFP stationary phase with the C18 and
biphenyl stationary phases
When fluoroalcohols were used as eluent additives with the
C18 and biphenyl columns, the retention patterns displayed both
similar and different trends in comparison to the PFP column. We
examined the situation on the basis of three exemplary analytes:
pyrrolidine (strong base), 2,6-dimethylpyridine (weak base) and
phenol (weak acid), see Fig. 5. We left out strong acids as their
retention was weak overall and clear comparisons could not be
made.
Fig. 5 reveals that very large differences in the retention factors were achievable for the same analyte using the same mobile
phase composition with the three different RP stationary phases.
The difference in the order of tens of times is not rare with basic
analytes (the largest is around 60 times). Short retention times of
basic analytes on C18 column could be improved by changing the
stationary phase. The strong bases included in this work and in the
works by Veigure et al. [12,22] did not contain aromatic rings and
therefore could not have π -π interactions, which makes the differences in the retention mechanisms between biphenyl, PFP and C18
6


K. Lossmann, R. Hecht, J. Saame et al.

Journal of Chromatography A 1666 (2022) 462850

stationary phases smaller. With phenol, the difference was around
3 times.
These differences led to the observed highly differentiated extents of separation. As an example, with HFTB at w
w pH 8.5 as
seen on Fig. 5, the selectivity factor (the ratio of retention factors) between the strongest and weakest retained analyte (2,6dimethylpyridine and phenol) was around 4 in the case of C18,
while with the PFP column it was over 50 (pyrrolidine being the
strongest retained and phenol being the weakest retained analyte).
As a generalisation, Fig. 5 clearly shows that on average the best
differentiation of retention with these analytes was obtained with
the PFP phase.
When using the same stationary phase and the same w
w pH,
while changing the nature of the buffer system, the difference in
retention can also be striking (see Figure S2). Although the PFP on
average ensured the highest selectivity factors between the compounds, it was the remaining two stationary phases that displayed
the most interesting retention order behaviour. In the case of the
biphenyl phase, just at w
w pH 8.5 the four different buffer systems
led to three (!) different retention orders and the selectivity factor
between the strongest and weakest retained analyte ranged from
2.2 to 6.5. While PP has shown promising influence in our earlier
study on C18 column, on PFP column HFIP and HFTB showed much
stronger influence.
In Fig. 5, the analyte that stands out is pyrrolidine – representative of strong bases. The figure shows that PFP column with fluorinated eluent additives is the best stationary and mobile phase
combination for pyrrolidine and similar analytes. Using common
mobile and stationary phases may not provide good enough retention or separation for strongly basic analytes. Even when using a
common eluent additive, the difference in retention factors is remarkable between different stationary phases: using ammonium

acetate pyrrolidine has retention factors well below 1 on C18 column, while the retention factor is well over 10 on the PFP stationary phase, showing an over 60 times increase in retention. Similar trends can be seen for the fluorinated eluent additives as well.
While ammonium acetate is possibly the most common eluent additive in LC-MS, it generally leads to lower retention factors compared to results obtained with fluoroalcohols as eluent additives.
Thus, when using ammonium acetate, it would be necessary to use
higher pH to be able to separate strong bases. Higher pH values,
however, are detrimental for any silica-based column and thus, not
advisable. The greatest difference obtained in retention factors between ammonium acetate and fluoroalcohols was more than 13
times (for pyrrolidine), remaining in the range of 3–10 times on
average for other model analytes.
In conclusion, very different selectivity is possible with the
same organic phase content and w
w pH when using a different type
of buffer and a different RP phase, however PFP with combined influence of fluoroalcohols has demonstrated a great potential of use
with pyrrolidine-like analytes.

charge-charge interactions between protonated analytes and deprotonated silanols) and charge-charge interactions between the
analytes and partially deprotonated fluoroalcohols which have attached to the stationary phase. The effect of different fluoroalcohols was also varied, but HFIP and HFTB had a more noticeable
effect on the retention of analytes than PP.
The comparison of the used mobile and stationary phases
showed that different selectivity was possible by changing the mobile buffer or stationary phase. Using the PFP column gave on average the best selectivity compared to the C18 or biphenyl columns.
Fluoroalcohols significantly improved the weak retention of basic
analytes. For weakly retained basic analytes this offers the possibilities of improvement of retention by either changing the stationary phase to PFP, changing the eluent additive to fluoroalcohols or
both, without increasing mobile phase pH to highly basic.
Authors understand that fluoroalcohols might not outcompete
tried and trusted eluent additives like ammonium acetate –as the
routinely used buffer (as few as they are for LC-MS methods)
components usually outcompete fluoroalcohols in price and performance combined competition, as well as for most aspects when
acidic analytes are to be researched and analysed. However, authors find this research very valuable nonetheless, for the cases,
when the results obtained with the routine eluent additives are
sub-optimal or not fit for the purpose at all – especially for basic analytes (as is the most common case when analysing active
pharmaceutical ingredients). Both broadening the selection of LCMS compatible eluent additives and giving explanation for most

probable interactions and thus elution patterns to expect, the authors hope to save time and effort for any bioanalytical chemist
analysing medical products.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.
CRediT authorship contribution statement
Krit Lossmann: Investigation, Formal analysis, Writing – original draft, Visualization. Ruta Hecht: Investigation, Writing – original draft, Visualization, Supervision. Jaan Saame: Investigation,
Formal analysis. Agnes Heering: Investigation, Formal analysis. Ivo
Leito: Conceptualization, Resources, Writing – review & editing,
Supervision. Karin Kipper: Conceptualization, Methodology, Writing – review & editing, Supervision.
Acknowledgments
The current research was supported by Restek Corporation
through the Academic Support Program (RASP). This research was
funded from the EMPIR programme (project 17FUN09 “UnipHied”,
www.uniphied.eu) co-financed by the Participating States and from
the European Union’s Horizon 2020 research and innovation programme, by the Estonian Research Council grants (PRG690) and by
the EU through the European Regional Development Fund under
the project TK141 “Advanced materials and high-technology devices for energy recuperation systems” (2014–2020.4.01.15–0011).
This work was carried out using the instrumentation at the Estonian center of Analytical Chemistry (www.akki.ee).

5. Conclusions
When using fluoroalcohols as eluent additives, the retention
factors of strong acids were generally lower, and the retention factors of strong bases were generally higher than when using ammonium acetate as the eluent additive. The retention factors obtained for weak bases and weak acids were more comparable in
the case of both ammonium acetate and fluoroalcohols. Retention
mechanisms with the PFP column using fluoroalcohols as the eluent additives appeared to be due to a combination of four interactions. These interactions are RP hydrophobic retention mechanism, π -π interactions between aromatic analytes and the aromatic ring on the stationary phase, hydrogen bond interactions between the analytes and free silanols on the stationary phase (or

Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi:10.1016/j.chroma.2022.462850.
7

K. Lossmann, R. Hecht, J. Saame et al.

Journal of Chromatography A 1666 (2022) 462850

References

[12] R. Veigure, K. Lossmann, M. Hecht, E. Parman, R. Born, I. Leito, K. Herodes,
K. Kipper, Retention of acidic and basic analytes in reversed phase column using fluorinated and novel eluent additives for liquid chromatographytandem mass spectrometry, J. Chromatogr. A. 1613 (2020) 460667, doi:10.1016/
j.chroma.2019.460667.
[13] E. Parman, L. Toom, S. Selberg, I. Leito, Determination of pKa values of fluorocompounds in water using 19F NMR, J. Phys. Org. Chem. (2018) e3940.
[14] R. Filler, R.M. Schure, Highly acidic perhalogenated alcohols. A new synthesis
of perfluoro-tert-butyl alcohol, J. Org. Chem. 32 (1967) 1217–1219, doi:10.1021/
jo01279a081.
[15] W.J. Middleton, R.V. Lindsey, Hydrogen bonding in Fluoro alcohols, J. Am.
Chem. Soc. 86 (1964) 4948–4952, doi:10.1021/ja01076a041.
[16] I. Koppel, J. Koppel, I. Leito, V. Pihl, L. Grehn, U. Ragnarsson, The acidity of
substituted 1-Hydroxybenzotriazoles in water and Dimethyl sulfoxide, J. Chem
Res. (1993) 446–447.
[17] C. Li, M.Z. Hoffman, One-electron redox potentials of phenols in aqueous solution, J. Phys. Chem. B. 103 (1999) 6653–6656, doi:10.1021/jp983819w.
[18] J. Stradins, B. Hasanli, Anodic voltammetry of phenol and benzenethiol derivatives.: part 1. Influence of pH on electro-oxidation potentials of substituted
phenols and evaluation of pKa from anodic voltammetry data, J. Electroanal.
Chem. 353 (1993) 57–69, doi:10.1016/0022- 0728(93)80286- Q.
[19] S. Zhang, A reliable and efficient first principles-based method for predicting
pKa values. 4. organic bases, J. Comput. Chem. 33 (2012) 2469–2482, doi:10.
1002/jcc.23068.
[20] N.F. Hall, M.R. Sprinkle, Relations between the structure and strength of certain
organic bases in aqueous solution, J. Am. Chem. Soc. 54 (1932) 3469–3485.
[21] Restek, Stationary Phase: fluoroPhenyl, (2018). />GNBR2368B-UNV.pdf (accessed 11 December 2021).

[22] R. Hecht (formerly Veigure), Novel Eluent Additives for LC-MS Based Bioanalytical Methods, University of Tartu, 2020.
[23] M. Rosés, Determination of the pH of binary mobile phases for reversed-phase
liquid chromatography, J. Chromatogr. A. 1037 (2004) 283–298, doi:10.1016/j.
chroma.2003.12.063.
[24] A. Suu, L. Jalukse, J. Liigand, A. Kruve, D. Himmel, I. Krossing, M. Rosés, I. Leito,
Unified pH values of liquid chromatography mobile phases, Anal. Chem. 87
(2015) 2623–2630, doi:10.1021/ac504692m.
[25] J. Nawrocki, The silanol group and its role in liquid chromatography, J. Chromatogr. A. 779 (1997) 29–71, doi:10.1016/S0 021-9673(97)0 0479-2.

[1] M.R. Euerby, A.P. McKeown, P. Petersson, Chromatographic classification and
comparison of commercially available perfluorinated stationary phases for
reversed-phase liquid chromatography using principal component analysis, J.
Sep. Sci. 26 (2003) 295–306, doi:10.1002/jssc.200390035.
[2] D.S. Bell, A.D. Jones, Solute attributes and molecular interactions contributing
to “U-shape” retention on a fluorinated high-performance liquid chromatography stationary phase, J. Chromatogr. A. 1073 (2005) 99–109, doi:10.1016/j.
chroma.2004.08.163.
[3] J... Wells, Pharmaceutical Preformulation: the physicochemical properties of
drug substances, 1998.
[4] J.W. Dolan, Back to basics: the role of pH in retention and selectivity, LC GC N.
Am. 35 (2017) 22–28.
[5] A. Heering, D. Stoica, F. Camões, B. Anes, D. Nagy, Z. Nagyné Szilágyi, R. Quendera, L. Ribeiro, F. Bastkowski, R. Born, J. Nerut, J. Saame, S. Lainela, L. Liv,
E. Uysal, M. Roziková, M. Vicˇ arová, A. Snedden, L. Deleebeeck, V. Radtke,
I. Krossing, I. Leito, Symmetric potentiometric cells for the measurement of
unified pH values, Symmetry (Basel). 12 (2020), doi:10.3390/sym12071150.
[6] A. Tan, J.C. Fanaras, Use of high-pH (basic/alkaline) mobile phases for LC–MS
or LC–MS/MS bioanalysis, Biomed. Chromatogr. 33 (2019) e4409, doi:10.1002/
bmc.4409.
[7] R. Veigure, R. Aro, T. Metsvaht, J.F. Standing, I. Lutsar, K. Herodes, K. Kipper, A highly sensitive method for the simultaneous UHPLC–MS/MS analysis of clonidine, morphine, midazolam and their metabolites in blood plasma
using HFIP as the eluent additive, J. Chromatogr. B. 1052 (2017) 150–157,
doi:10.1016/j.jchromb.2017.03.007.

[8] K. Kipper, K. Herodes, I. Leito, Fluoroalcohols as novel buffer components for
basic buffer solutions for liquid chromatography electrospray ionization mass
spectrometry: retention mechanisms, J. Chromatogr. A. 1218 (2011) 8175–8180.
[9] B. Basiri, H. van Hattum, W.D. van Dongen, M.M. Murph, M.G. Bartlett, The
role of fluorinated alcohols as mobile phase modifiers for LC-MS analysis of
oligonucleotides, J. Am. Soc. Mass Spectrom. 28 (2017) 190–199.
[10] L. Gong, R. Liu, Y. Ruan, Z. Liu, The role of fluoroalcohols as counter anion for ion-pair reversed-phase liquid chromatography/high resolution electrospray ionization mass spectrometry analysis of oligonucleotides, Rapid Commun. Mass Spectrom. 33 (2019).
[11] K. Kipper, K. Herodes, I. Leito, L. Nei, Two fluoroalcohols as components of
basic buffers for liquid chromatography electrospray ionization mass spectrometric determination of antibiotic residues, Analyst 136 (2011) 4587–4594.

8