Fluorinated carboxylic acids have been in use as ion-pairing reagents for over three decades. It has been observed that ion-pairing reagents not only increase the retention of oppositely charged analytes on reversed-phase HPLC columns but also decrease the retention of similarly charged analytes.
Journal of Chromatography A 1631 (2020) 461575 Contents lists available at ScienceDirect Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma Fluorinated carboxylic acids as “ion repelling agents” in reversed-phase chromatography Bassam Lajin∗, Walter Goessler Institute of Chemistry—Analytical Chemistry for Health and Environment (ACHE), University of Graz, Austria a r t i c l e i n f o Article history: Received 22 July 2020 Revised 17 September 2020 Accepted 18 September 2020 Available online 21 September 2020 Keywords: Ion-pair chromatography Alkyl sulfonate Fluorinated acetic acids Heptafluorobutyric acid ICPMS a b s t r a c t Fluorinated carboxylic acids have been in use as ion-pairing reagents for over three decades It has been observed that ion-pairing reagents not only increase the retention of oppositely charged analytes on reversed-phase HPLC columns but also decrease the retention of similarly charged analytes; these latter effects, however, have not been thoroughly investigated for the fluorinated carboxylic acids, and the application of these reagents has been rather restricted to their ion-pairing capacity to separate basic analytes In the present study, we report a systematic investigation about the effects of three fluorinated carboxylic acids (trifluoroacetic acid (TFA), pentafluoropropionic acid (PFPA), and heptafluorobutyric acid (HFBA)) on the retention and selectivity of the separation of halogenated carboxylic acids and sulfonic acids by reversed-phase chromatography with an inductively coupled plasma mass spectrometry detector (ICPMS) Several eluents were tested and compared at different concentrations (0–100 mM) and pH values, including sulfate, nitrate, phosphate, oxalate, TFA, PFPA, and HFBA The fluorinated carboxylic acids resulted in a consistent decrease in the retention factors (up to ca 9-fold with HFBA) in a concentration dependent manner, which plateaued at around 50 mM Significant improvement of the peak symmetry of the chromatographed acids was also observed We highlight the advantages of incorporating the fluorinated carboxylic acids in modifying the selectivity and retention of organic acids in reversed phase chromatography in general, and particularly when employing chromatographic detectors with limited compatibility with organic mobile phases such as the ICPMS © 2020 The Authors Published by Elsevier B.V This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Introduction The fluorinated carboxylic acid derivatives have been in use as ion-pairing reagents since the early 1980s when they were first introduced to modify the retention for the reversed-phase chromatography of peptides [1–3] The most commonly used and widely known of these compounds is trifluoroacetic acid, which is frequently employed to improve the retention and peak shape of basic analytes on silica-based reversed-phase chromatography through its ion-pairing capacity The fluorinated carboxylic acids are relatively strong acids with pKa values 1.0 Even though the use of negatively charged ion-pairing reagents is generally known to decrease the sensitivity for the detection of basic analytes by ESI-MS through the formation of the ion-pair, one advantage of the fluorinated carboxylic acids as ion-pairing additives over the classically employed alkyl sulfonates is their high volatility which confers better compatibility with LC-ESI-MS [5,6] ∗ Corresponding author E-mail address: Bassam.lajin@uni-graz.at (B Lajin) A further consequence of the volatility of the fluorinated carboxylic acids is their removability which can be an advantage in preparative separations [7] Most importantly, the equilibration times for these ion-pairing reagents is very fast [8,9] Fluorinated carboxylic acids such as PFPA and HFBA have also been employed as ionpairing reagents for the reversed-phase chromatography of small basic molecules [8–11], albeit less frequently than have TFA and the alkyl sulfonates As with other types of ion-pairing reagents, the use of the fluorinated carboxylic acids has been dominated by improving the retention and separation of analytes that can harbor an opposite (i.e positive) charge through their ion-pairing capacity, which has been thoroughly investigated [12,13] The general effects of negatively charged ion pairing reagents (e.g alkyl sulfonates) on the retention of similarly charged (i.e acidic) compounds have been observed [14–16] However, the effects of the fluorinated carboxylic acids on the retention of acidic compounds has not been systematically investigated In the present work, we show the strong and concentrationdependent influence of the fluorinated carboxylic acids on the retention of negatively charged compounds in reversed-phase chro- https://doi.org/10.1016/j.chroma.2020.461575 0021-9673/© 2020 The Authors Published by Elsevier B.V This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) B Lajin and W Goessler Journal of Chromatography A 1631 (2020) 461575 matography with the incentive of highlighting their rather underappreciated potential as powerful tools for modifying the selectivity and retention for the reversed-phase separation of organic acids Furthermore, we expand on the practical advantages of the fluorinated acids particularly when employing a carbon-prone chromatographic detector such as the inductively coupled plasma mass spectrometry (ICPMS), where mobile phase selection is relatively limited and the detection can be greatly compromised by high content of organic solvents in the mobile phase Phenomenex Luna C5 (2.0 mm I.D x 150 mm long, μm particle size) using mobile phases containing 0–10% methanol and 0– 100 mM of one of three investigated fluorinated carboxylic acids (TFA (trifluoroacetic acid), PFPA (pentafluoropropionic acid), and HFBA (heptafluorobutyric acid) We controlled the pH by employing a buffer containing phosphate, which interacts minimally with the hydrophobic stationary phase, at a constant concentration of 30 mM, including the addition of varying concentrations of one of the different eluents (TFA, PFPA, HFBA, oxalate, sulfate, and nitrate) with concentrations in the range of 0–100 mM The pH of the final mobile phase was adjusted to 2.3 (± 0.1) with ammonia The non-fluorinated eluents were also compared with the fluorinated carboxylic acids at various but equivalent concentrations and pH values Column temperature was controlled at 40 °C for all experiments The detailed chromatographic conditions including mobile phase compositions are mentioned in detail in each figure caption Materials and methods 2.1 Test compounds Acidic compounds of various hydrophobicity and pKa values, including halogenated acetic acid (HAA) derivatives and sulfonic acids were chromatographed as model compounds The studied HAAs and their dissociation acid constants [17,18] were: chloro acetic acid (CAA, pKa = 2.87), dichloroacetic acid (DCAA, pKa = 1.35), trichloroacetic acid (TCAA, pKa = 0.66), bromoacetic acid (BAA, pKa = 2.90), dibromoacetic acid (DBAA: pKa = 1.39), and tribromoacetic acid (TBAA: pKa = 0.72) The sulfonic acids (pKa < 0) were: 1-propanesulfonic acid, 1-pentanesulfonic acid, benzenesulfonic acid, and 1-octane sulfonic acid Additionally, neutral compounds were employed for reference, namely chloroethanol, chloroacetamide, and thiourea The compounds were purchased from Sigma-Aldrich (Steinheim, Germany) and prepared in pure water (18.2M cm) at a concentration in the range of 5.0–30 mg L−1 Results and discussion 3.1 The general effects of the fluorinated acids on similarly charged analytes The haloacetic acids were selected as model compounds for separation as their pKa values span a wide range (0.5–3.0 [17,18]), which allows an in-depth investigation of the effect of the extent of the partial negative charge on the analyte Furthermore, the number of halogen atoms on the haloacetic acid can modify the hydrophobicity and chromatographic retention on C18 in opposing ways In other words, an increase in the number of the halogen atoms increases hydrophobicity but also decreases the pKa and increases the acid strength through the inductive effects of the halogen atom, which in turn increases ionization and decreases hydrophobicity Therefore, at certain pH values, the overall hydrophobicity, and hence the retention on the hydrophobic C18 stationary phase, of some of the haloacetic acids would be comparable, while the partial negative charge would be quite different This creates an exemplary situation where the ionic interaction mechanisms can be useful Such a situation can be seen in Fig where the separation between chloroacetic acid, CAA (peak 1) and dichloroacetic acid, DCAA (peak 2) within the pH range of 2.5–3.0 was significantly improved when using 50 mM pentafluoropropionic acid instead of 50 mM sulfuric acid as the eluent (pH adjusted with ammonia) The reversal in the peak order for CAA (pKa = 2.87) and DCAA (pKa = 1.35) at pH 3.0 observed when comparing Fig 1A and Fig 1B and the associated improvement in the separation can be clearly explained by stronger adsorption of PFPA relative to sulfate on the C18 stationary phase and a resulting ion-ion repulsion mechanism Such a repulsion effect is much bigger for the more negatively charged DCAA at that pH (partial charge ca −1.0) than for CAA (partial charge −0.6) Another observation is the decrease in retention for the trichloroacetic acid (TCAA) (pKa = 0.66, partial charge ca −1.0 at pH > 1.5) Overall, a fast separation of the three HAAs was achievable within TFA > PFPA > HFBA In additional experiments, small amounts of methanol were added to the mobile phase to investigate whether the coexistence of an organic mobile phase component can significantly decrease the adsorption of the hydrophobic fluorinated carboxylic acids on the C18 stationary phase and therefore counteract the observed ion-repulsion effects observed This would be expected to manifest itself as flattening in the curves observed under the 100% aqueous mobile phase conditions shown in Fig However, we did not clearly observe such trends up to 10% methanol (Fig S1) Higher methanol concentrations were not tested due to the incompatibility of high organic mobile phase content with the ICPMS under standard conditions Apart from the ion-repulsion and the slight decrease in the hydrophobicity of the C18 stationary phase, the employment of the fluorinated carboxylic acids as mobile phase additives in reversedphase chromatography might also modify the selectivity of the stationary phase by introducing the polarized carbon-halogen bonds and therewith additional types of interaction Such interactions might have also played a small additional role in modifying the retention of the separation of the chlorinated and brominated test 3.2 Concentration-dependent effects We investigated the effects of eluent concentration on the retention of TCAA and TBAA as negatively charged model compounds and 2-chloroacetamide and 2-chloroethanol as reference neutral compounds As can be observed in Fig 4, the retention on the C18 stationary phase decreased more steeply with increasing hydrophobicity of the eluent for the charged compounds The retention for the reference neutral compounds showed only a slight decrease (up to 15% decrease with HFBA, Fig 4C & D) This slight charge-independent decrease could be explained by a decrease in the general hydrophobicity of the C18 stationary phase due to coating with the polar ionized carboxylic acid groups, which can decrease the access of the analyte to the C18 stationary phase B Lajin and W Goessler Journal of Chromatography A 1631 (2020) 461575 Fig A comparison between different eluents for the separation of six haloacetic acids The chromatograms show the separation of a mixture of standards containing 5.0, 10, and 15 mg Cl L−1 of chloroacetic acid (CAA), dichloroacetic acid (DCAA), and trichloroacetic acid (TCAA), respectively, and 10, 20, and 30 mg Br L−1 of bromoacetic acid (BAA), dibromoacetic acid (DBAA), and tribromoacetic acid (TBAA), respectively Stationary phase: YMC Triart-C18; column temperature: 40 °C; mobile phase flow rate: 0.8 mL min−1 ; injection volume: 5.0 μL; mobile phase: 50 mM of sulfuric acid (A), oxalic acid (B), nitric acid (C), trifluoroacetic acid (D), or pentafluoropropionic acid (E), all adjusted to pH 1.5 with ammonia Signal offsets were applied to the Cl and Br signals to facilitate visualization The pKa values [17, 18] and the corresponding calculated partial negative charge on the analytes at the mobile phase pH 1.5 are as follows: CAA: pKa = 2.87, −0.04; DCAA: pKa = 1.35, −0.59; TCAA: pKa = 0.66, −0.87; BAA: pKa = 2.90, −0.04; DBAA: pKa = 1.39, −0.56; TBAA: pKa = 0.72, −0.86 The void time was ca 0.8 (estimated based on the retention time of chloride with ammonium sulfate as the eluent) Note the dramatic change in selectivity and the decrease in retention time with the increased hydrophobicity of the negatively charged eluent (sulfate < oxalate < nitrate < trifluoroacetic acid < pentafluoropropionic acid) in a manner dependent on the partial negative charge on the analytes, which can be explained by adsorption of the more hydrophobic eluents on the C18 stationary phase and negative ion repulsion This effect is remarkably manifested in the change in peak order under pentafluoropropionic acid (E) B Lajin and W Goessler Journal of Chromatography A 1631 (2020) 461575 Fig The differential effects of the fluorinated carboxylic acid on the retention of negatively charged versus neutral compounds The overlaid chromatograms show the separation of 10 mg Cl L−1 of 2-chloroacetamide (Cl-AA), 10 mg Cl L−1 trichloroacetic acid (TCAA), and 10 mg Br L−1 of tribromoacetic acid (TBAA) Stationary phase: YMC Triart-C18; column temperature: 40 °C; mobile phase flow rate: 0.8 mL min−1 ; injection volume: 5.0 μL; mobile phase: 30 mM of phosphoric acid in addition to 10 mM of oxalic acid (A), nitric acid (B), trifluoroacetic acid (TFA) (C), pentafluoropropionic acid (D), or heptafluorobutyric acid (E), all adjusted to pH 2.3 Phosphate interacts minimally with the C18 stationary phase and was employed as a buffer (pKa1 = 2.1) at a constant concentration The pH was selected as to yield approximately a full negative charge on trichloroacetic acid (pKa = 0.66) and tribromoacetic acid (pKa = 0.72) 2-chloroacetamide was used to serve as a neutral reference compound Note the relative stability in the retention of the neutral compound (peak 1) and the marked decrease in the retention of the negatively charged compounds in a manner dependent on the hydrophobicity of the varied eluent Heptafluorobutyric acid resulted in a change in peak order compounds and should be taken into consideration when interpreting the results of the present study Kamiusuki et al reported an increase in the retention of fluorinated compounds in direct proportion to their content of fluorine atoms on a fluorinated stationary phase relative to a non-fluorinated ODS stationary phase [20] Our data, however, indicate that ion-repulsion is by far the dominant mechanism explaining the observed trends Nevertheless, we tested other non-halogenated acids, namely sulfonic acids of different chain length along with thiourea as a neutral compound, and observed similar trends (Fig S2) B Lajin and W Goessler Journal of Chromatography A 1631 (2020) 461575 Fig The variation in the retention factor for negatively charged and neutral compounds as a function of the eluent concentration Stationary phase: YMC Triart-C18; column temperature: 40 °C; mobile phase flow rate: 0.8 mL min−1 ; injection volume: 5.0 μL; mobile phase: 30 mM of phosphoric acid in addition to variable concentrations of oxalic acid, nitric acid, trifluoroacetic acid (TFA), pentafluoropropionic acid (PFPA), or heptafluorobutyric acid (HFBA), all adjusted to pH 2.3 The pH was selected as to yield approximately a full negative charge on trichloroacetic acid (pKa = 0.66) (A), and tribromoacetic acid (pKa = 0.72) (B) As neutral reference compounds we used 2-chloroethanol (C), and 2-chloroacetamide (D) Note that the retention factors for the neutral compounds showed only a small decrease (10–20%) correlating with the hydrophobicity of the studied eluent, which was in sharp contrast with the dramatic decrease in the retention of the negatively charged compound (A & B) The retention time repeatability was investigated for the different eluents and the% RSD was found to be