Abstract
As part of this project, this section of the wiki is devoted to the analysis of fluorinated molecules. Our main objectives were to verify the addition of the ring to TFA by Spl lipase and, consequently, to confirm that the enzymatic reaction had indeed taken place. This verification was carried out by HPLC-UV analysis.
Another objective of the study was to monitor the degradation of PFOA (perfluorooctanoic acid) by fluorine NMR (¹ F NMR). This technique is particularly suitable because PFOA is highly fluorinated, and each fluorine atom, depending on its chemical environment, has a characteristic chemical shift. ¹ F NMR thus not only detects the disappearance of the parent compound and the appearance of degradation products (such as shorter-chain perfluorocarboxylic acids or inorganic fluoride), but also provides quantitative data. It is therefore a relevant tool for monitoring the degradation kinetics of PFOA.
However, in the time available, we were unable to perform NMR analysis of PFOA. We only performed HPLC-UV analysis, without having time to complete all the planned manipulations. In particular, not all calibration curves could be established, and we were unable to accurately determine all retention times.
Introduction
The aim of the analytical section was twofold: first, to quantify the cyclized TFA produced by the enzymatic cyclization of TFA using SpL lipase, and second, to quantify the cyclized TFA during its degradation by dehalogenases.
But how did we get here?
Since PFAS quantification is complex, we were supported and guided by various researchers and engineers specializing in the field.
Initially, PFAS quantification is known to be challenging and is typically performed using High-Performance Liquid Chromatography coupled with tandem mass spectrometry (HPLC-MS/MS). This method is the most accurate but also the most expensive and time-consuming to implement. Researcher Carlos Alfonso helped us understand the principles of mass spectrometry-based quantification and advised us to use a targeted approach by obtaining internal standards of the molecules to be analyzed.
We also consulted three researchers from the Institute of Analytical Sciences (Jérôme Lemoine, Xavier Saupin, and Karine Faure), who introduced us to the different types of mass spectrometers they work with. They explained the difficulties of setting up an analytical method using mass spectrometry within such a limited timeframe. Indeed, the preliminary steps required for HPLC-MS/MS analysis are lengthy to establish and optimize. Sample preparation steps prior to analysis, as well as the determination of analytical instrument parameters, are necessary:
- Extraction of fluorinated molecules from enzyme-containing samples (extraction protocol)
- Purification and concentration of these molecules in a suitable buffer (purification and concentration protocol)
- Separation by HPLC (parameters to be determined)
- Analysis by mass spectrometer (detection parameters to be determined)
These steps are even more complex with PFAS because these molecules are toxic, hydrophobic, ubiquitous, and quickly contaminate the equipment used.
To help us develop extraction and purification protocols, in addition to literature research, we turned to our sponsor, the Carso analysis laboratory, and to ecotoxicologist Maxime Louzon. PFAS extraction and purification are generally performed using solvent extraction (liquid-liquid extraction) complemented by solid-phase extraction on WAX cartridges. Special filters can also be used. However, since each fluorinated molecule has different physicochemical properties, protocols can vary from one molecule to another. As cyclized TFA is a molecule synthesized by our team, it has not been studied in the literature, so we had to draw inspiration from protocols for similar fluorinated molecules such as TFA or other short-chain PFAS. Unfortunately, due to lack of time, none of the protocols could be tested, but we hope to test them in the future!
For the analytical part and molecule quantification, it was during a meeting with Karine Faure that we realized mass spectrometry would not be necessary for analyzing the enzymatic reaction of cyclization on TFA. Indeed, both the 4-chlorobenzylamine (reactant) and the cyclized TFA (product) contain a ring detectable by UV! A simpler and less expensive method was therefore used for analysis: HPLC-UV. The method development protocols for HPLC-UV were created with the help of Karine Faure and optimized during laboratory experiments with the support of Stéphane Chambert.
Enzymatic Reaction Setup
Figure 1: Reaction scheme for adding the cycle to TFA with Lipase SpL
To carry out the reaction of adding the cycle to TFA, we focused heavily on the protocol from the article by Zeng. S, 2018:
Methods and Procedures:
The above reaction is carried out at 30°C with 10 mM TFA and 20 mM 4-chlorobenzylamine in 5 mL of n-hexane. 500 µL of a 20 mg/mL aqueous solution of purified SpL lipase is added to the mixture and stirred at 250 rpm. The reaction should be complete within 2 hours, but as a precaution we can choose to wait up to 24 hours. To permanently stop the reaction after 24 hours, 5 mL of methyl tert-butyl ether (MTBE) is added to the reaction mixture. Next, anhydrous sodium sulfate is added to remove water. This step is crucial to stop the enzymatic activity of SpL, as water is generally necessary for catalysis by lipases.
As part of the study of the enzymatic reaction of adding a ring to TFA (see reaction page), several steps were defined to verify that the transformation had indeed taken place, with the help of Karine Faure, a researcher at the Institute of Analytical Sciences (ISA) in Lyon, and Stéphane Chambert, a lecturer at INSA Lyon.
Purpose: The objective was to develop an analytical method to identify and quantify 4-chlorobenzylamine (reagent) and cyclized TFA (product) from the above reaction. To do this, it was necessary to:
- Determine retention times: measure the retention times of 4-chlorobenzylamine (reagent) and N-[(4-chlorophenyl)methyl]-2,2,2-trifluoroacetamide (product), which we will refer to as cycled TFA under the HPLC protocol conditions, as this information was not available in the literature.
- Create calibration curves: construct calibration curves to quantify the amount of product formed and the reagent remaining after the enzymatic reaction.
The initial trifluoroacetic acid (TFA) is not detectable by HPLC-UV, as it does not have a UV absorbing chromophore, making direct measurement of this reagent impossible using this method.
Due to delivery delays, the commercial standard for cycled TFA was not available at the time of our analyses. In order to move forward, we had to synthesize the molecule ourselves in parallel with our HPLC manipulations.
This approach proved to be relevant to the project because it allowed us to:
- Have a reliable standard for identifying and quantifying the product formed by the enzymatic reaction
- Experimentally verify the feasibility of cyclized TFA synthesis
- Integrate chemical synthesis and analytical analysis, thereby enriching the overall methodology of the project and understanding of the mechanism under study
Accordingly, our work will be structured into three main parts:
- Determination of retention times for target molecules
- Synthesis of cycled TFA
- Construction of the calibration curve
I. Determination of Retention Times for Target Molecules
This section presents the results of HPLC analyses performed to determine the retention times (rt) of the various target molecules studied at the INSA Lyon laboratory. These data are essential for identifying compounds by liquid chromatography, verifying the reproducibility of analyses, and adjusting methods according to the chemical nature of the substances.
I.A. The General Conditions of the Analyses
All analyses were carried out using reverse-phase HPLC (Agilent 1260 Infinity) equipped with a Hypersil™ ODS-2 C18 column (250 mm × 4.6 mm, 5 µm; Thermo Scientific).
The following conditions apply to all experiments, unless otherwise stated.- Type of samples analyzed: organic compounds dissolved in a water/methanol mixture (toluene, 4-chlorobenzylamine, cycled TFA)
- Detection: UV, mainly at 230 nm, the optimal wavelength for observing the aromatic compounds studied (also detection at 210 and 250 nm)
- Injection volume: 5 µL or 10 µL, depending on the signal intensity
- Initial pressure observed: between 108 and 117 bar depending on the solvent composition
- Solvents:
- Solvent A: Methanol with 20 mM ammonium acetate
- Solvent B: Water with 20 mM ammonium acetate
- Preparation of solvents with 20 mM ammonium acetate: 0.7760–0.7822 g of ammonium acetate
I.B. The Separation Method (Gradient)
We had to find the most optimal separation method (gradient) for our experiments. The initial gradient lasted 30 minutes. This gradient was used in most of the preliminary analyses, particularly for the tests with toluene and 4-chlorobenzylamine.
0 min: 99% B (water), 1% A (MeOH)
20 min: 1% B, 99% A
24 min: 1% B, 99% A
26 min: 99% B, 1% A
30 min: 99% B, 1% A
The objective was to cover a wide range of polarity in order to separate both highly polar and less polar compounds.
After several gradient tests, an optimized 20-minute gradient was subsequently implemented to improve the efficiency of the analytical protocol and reduce analysis time, while maintaining good resolution.
0 min: 80% B, 20% A
10 min: 10% B, 90% A
15 min: 10% B, 90% A
18 min: 80% B, 20% A
20 min: 80% B, 20% A
It was used in particular for tests on cycled TFA and 4-chlorobenzylamine after dilution, as well as for constructing the final calibration curve.
I.C. The Retention Times
I.C.1. Toluene
- Preparation: dilution to 10% and 1% (v/v) in methanol
- Gradient: 30 min
- Solvents: methanol/water (without additives)
- rt ≈ 15.0 min
The toluene is a low-polarity compound, expected at the end of the gradient, towards the methanol-rich proportions. It is used to verify the proper functioning of the column and the reproducibility of the gradient. It should be noted that for the very first tests, we did not add ammonium acetate to our solvents. Two tests showed excellent reproducibility:
rt = 15.020 min (10%)
rt = 14.970 min (1%)
In a subsequent test with solvents containing 20 mM ammonium acetate (solvents A and B), the rt of toluene increased slightly: tr = 15.225 min.
I.C.2. 4-Chlorobenzylamine
- Preparation: 10 mM solution in a 50:50 water/methanol mixture; injected pure or diluted 5x depending on the test
- Solvents: methanol/water with 20 mM ammonium acetate
- 30-minute gradient: several peaks observed around 9 minutes.
- 20-minute gradient (optimized):
- tr = 8.026 min (5x diluted solution)
- tr = 7.601 min (in a mixture with cycled TFA)
Note: the UV spectrum shows a main peak around 8 min. However, repeated analyses have revealed the appearance of two high-concentration peaks, probably related to: 1. a molecular or ionic association of the compound 2. a concentration effect on the column 3. an impurity
This phenomenon disappears after dilution, confirming that it is a concentration-dependent effect.
I.C.3. Cycled TFA
- Preparation: 10 mM solution in water/methanol, then diluted 5x.
- Solvents: Methanol/water with 20 mM ammonium acetate.
- Optimized gradient: 20 min.
- Observed tr:
- rt = 9.721 min (5 μL injection)
- rt = 9.170 min (10 μL injection)
- rt = 9.2 min (in a mixture with amine)
The peak is well defined at 230 nm, with sufficient intensity from 10 μL injected. Retention is consistent with the intermediate nature of this molecule (neither too polar nor too hydrophobic).
I.D. Co-injection and Separation of Mixtures
A test was performed by injecting an equimolar mixture of 4-chlorobenzylamine and cyclized TFA. The peaks are correctly separated:
rt(4-chlorobenzylamine) = 7.601 min
rt(cyclized TFA) = 9.2 min
This confirms the separation capacity of the gradient optimized at 20 minutes, even for compounds with relatively similar structures.
I.E. Observations
- The use of 20 mM ammonium acetate provides better resolution for basic or ionizable compounds
- The initial gradient at 30 min is too long for the analytical requirements of the project, and the 20-minute gradient has proven to be sufficiently effective for the analysis of target compounds
II. Synthesis of Cycled TFA
The molecule N-[(4-chlorophenyl)methyl]-2,2,2-trifluoroacetamide (cycled TFA) was synthesized to provide a standard for verifying the enzymatic reaction. The protocol was based on the method of N-acylation of amines by esters, using acetic acid as a catalyst (Sanz et al., 2010). The method used is based on the publication: Daniel D. Sanz, Sharleya and Jonathan M. J. Williams, "Acetic acid as a catalyst for the N-acylation of amines using esters as the acyl source," Chem. Commun., 2010, DOI https://doi.org/10.1039/C6CC09023K
This synthesis provided a pure standard of cyclized TFA, which is essential for identification and quantification by HPLC-UV. No catalyst was used in our synthesis, as the 4-chlorobenzylamine/ethyl trifluoroacetate pair is sufficiently reactive to form the amide when stirred at room temperature overnight. Fluorinated esters such as ethyl trifluoroacetate are more electrophilic than conventional esters, which facilitates nucleophilic attack by the amine.
Figure 2: Reaction scheme for synthesis of cycled TFA
Reagents
- A: 4-chlorobenzylamine, 4 mmol
- B: ethyl trifluoroacetate, 4 mmol
- Solvent: toluene, 2 mL
The protocol is structured into four parts:
- Preparation of the reaction medium
- Control by TLC (Thin Layer Chromatography)
- Evaporation and purification of the sample
- Synthesis yield
- HPLC spectrum after purification
II.A. Preparation of the Reaction Medium
- Using a P200 micropipette, withdraw 488 μL of ethyl trifluoroacetate (B).
- Using a P200 micropipette, withdraw 488 μL of 4-chlorobenzylamine (A).
- Add 2 mL of toluene to the reaction mixture using a P1000 micropipette.
- Allow the mixture to react with stirring (under a fume hood) overnight.
The next day, the reaction was stopped and a TLC check was performed.
II.B. Control by TLC (Thin Layer Chromatography)
A small sample of the reaction medium was taken for TLC. The plates were developed under UV light.
Cycled TFA and the reagent 4-chlorobenzylamine have aromatic groups (phenyl ring) that absorb UV light. When the plate is exposed to a UV lamp, these compounds become visible as dark or bright spots, allowing the product to be located and its purity verified. A small amount of the sample was recovered for HPLC analysis prior to complete purification on the column.
Figure 3: TLC analysis showing reaction products
Two tasks were observed for what corresponds to a single product. A check was performed to determine whether the second task could have originated from the reaction solvent (toluene).
Figure 4: TLC comparison with toluene
On a new TLC, two deposits were made: one deposit containing only toluene (the reaction solvent), which migrates directly with the solvent front, and one deposit of the product from the reaction medium prior to purification, which has two distinct stains.
Comparison of the two deposits shows that the second spot observed in the raw sample is not due to the toluene solvent, which invalidates the initial hypothesis. The second spot observed on the TLC cannot correspond to the reagent, as the latter migrates very little on silica and would remain close to the starting point, unlike the spot observed further away. This hypothesis was therefore discarded.
Figure 5: HPLC spectrum of reaction mixture
The spectrum of the reaction mixture prior to purification still shows the presence of the 4-chlorobenzylamine reagent. The cyclized TFA product appears at a retention time of 15.117 min under the HPLC conditions presented above, with a duration of 30 min. This spectrum serves as a reference prior to purification, for comparison with the spectrum obtained after passage through the column.
II.C. Evaporation and Purification of the Sample
- The solvent was evaporated in a rotary evaporator, yielding a pale yellow solid.
- A chromatography column was prepared: 1 cm of sand, silica with dichloromethane, 1 cm of sand on top.
- The sample was dissolved in a small amount of dichloromethane and placed on the prepared column.
Observation: The yellow compound (responsible for the pale yellow color of the crude solid) remains on the surface of the column and migrates little, suggesting that it is probably the 4-chlorobenzylamine reagent. The yellow spot on the surface of the column could be the reagent, but this is not certain: it could also be a by-product or impurity that migrates little.
- The fractions were collected in test tubes.
- Each fraction was analyzed on a silica plate under UV light to identify those containing the product.
Figure 6
Figure 6: Fraction analysis on TLC plate
After collecting the fractions in test tubes, a grid was drawn on a silica plate to locate the fractions containing the product. Each fraction was spotted on the grid and revealed under UV light. Fractions 4 to 8 contain the cycled TFA product. Fractions 26 to 28 show a second product, with faint spots.
Fractions 4 to 8 containing the product were migrated to TLC to verify the presence of the product. These fractions were transferred to a flask and the solvent was evaporated in a rotary evaporator, yielding a white solid. After complete recovery and removal of the product from the flask walls, the final purified quantity was 614 mg of white powder.
Fractions 26 were also migrated, confirming the presence of the second product. Fractions 25 to 27 were evaporated for NMR to identify this second compound. Unfortunately, we did not have time to identify this second compound.
II.D. Synthesis Yield
For the synthesis of N-[(4-chlorophenyl)methyl]-2,2,2-trifluoroacetamide (cycled TFA), we used 4 mmol of 4-chlorobenzylamine and 4 mmol of ethyl trifluoroacetate. The molar mass of the product is 237.60 g/mol, and the amount obtained after purification was 0.7209 g. n = 0.7209 / 237.60 = 0.00303 mol
The maximum amount of product that could be obtained is 0.004 mol (amount of limiting reagent). The yield is therefore: (0.00303 / 0.004) × 100 = 75.9%
The reaction yield is 75.9%. Thus, the reaction yield is 75.9%, which shows that the synthesis was relatively efficient. NMR allowed us to confirm the successful synthesis of cyclized TFA.
Figure 7: NMR confirmation of cycled TFA
II.E. HPLC Spectrum After Purification
Figure 8: HPLC spectrum after purification
After purification of the cycled TFA, an HPLC spectrum was obtained under the same conditions as those used for the mixture prior to purification. The spectrum shows the absence of the 4-chlorobenzylamine reagent, and the product appears with a similar retention time of 14.911 minutes. This observation confirms that column purification has isolated the pure product.
III. Construction of the Calibration Curve
The purpose of this step is to produce a reliable calibration curve for cycled TFA so that we can quantify this compound in our samples using HPLC. To ensure the accuracy and reproducibility of the measurements, we use a calibration method with an internal standard (toluene).
III.A. 10 mM Stock Solution of Cycled TFA
- Weight measured: 0.0243 g of synthesized cyclized TFA
- Solvent used: water/methanol solution (50:50 v/v) containing 1% toluene - final volume: 20 mL
In the following protocol, the mixing order is important to ensure solubility:
- 2.5 mL of pure methanol
- Addition of cycled TFA (0.0243 g)
- Addition of toluene (50 μL) to obtain 1% (v/v)
- 2.5 mL of ultrapure water
If water is added too early, the cycled TFA does not dissolve properly: a whitish cloudiness appears, which disappears when methanol is added.
III.B. Preparation of Standard Solutions (1 mL)
The dilutions were made from the 10 mM stock solution, with the addition of a water/methanol mixture (50:50) still containing 1% toluene. Each final solution has a volume of 1 mL, adapted to the volume of an HPLC vial.
| Concentration of the standard in mM | Volume A of stock solution to be withdrawn in mL | Solvent volume (water/methanol, 50/50) in mL |
|---|---|---|
| 0 | 0 | 1 |
| 2.5 | 0.125 | 0.875 |
| 5 | 0.25 | 0.75 |
| 6 | 0.3 | 0.7 |
| 7 | 0.35 | 0.65 |
| 8 | 0.4 | 0.6 |
| 9 | 0.45 | 0.55 |
| 10 | 0.5 | 0.5 |
Before injecting the entire calibration range, a check is performed to ensure that the toluene and cycled TFA are sufficiently separated to be used together (e.g., section I.D)
- Injection: 10 μL
- Detection at 210 nm (because toluene does not absorb at 230 nm)
Results Tables - Cycled TFA (with Internal Standard Toluene)
| Concentration of the standard in mM | Rt cycled TFA (minute) | Surface area cycled TFA | Rt Toluene (minute) | Surface area Toluene |
|---|---|---|---|---|
| 0 | 0 | 0 | 10.012 | 25829.3 |
| 2.5 | 9.103 | 12001.5 | 9.996 | 23408.8 |
| 5 | 9.192 | 18697.9 | 9.999 | 23923.6 |
| 6 | 9.186 | 21032.8 | 10.006 | 23589.1 |
| 7 | 9.190 | 22328.6 | 10.015 | 22589.6 |
| 8 | 9.186 | 22551.3 | 10.032 | 21983.1 |
| 9 | 9.189 | 23721.6 | 10.029 | 21472.9 |
| 10 | 9.192 | 25667.8 | 10.039 | 22007.2 |
Figure 9: Calibration curve - Linear relationship between area ratio (TFA/toluene) and concentration of cycled TFA
The calibration curve obtained shows a linear relationship between the area ratio (TFA/toluene) and the concentration of cycled TFA, with a correlation coefficient of 0.978.
Conclusion
All of the work carried out as part of this project aimed to verify the feasibility and effectiveness of the enzymatic reaction of adding a halogenated ring (4-chlorobenzylamine) to TFA, catalyzed by Spl lipase, as well as to develop reliable analytical tools for monitoring this transformation.
Although not all of the planned analyses could be completed—in particular, ¹F NMR for monitoring PFOA degradation—the results obtained by HPLC-UV validated several key steps in the experimental protocol.
The development of chromatographic conditions allowed a clear separation between the reagent (4-chlorobenzylamine, tr ≈ 7.6 min) and the product formed (cycled TFA, tr ≈ 9.2 min) thanks to an optimized 20-minute gradient in reverse phase. This protocol demonstrated its robustness and reproducibility, providing a solid basis for future quantitative analyses.
The chemical synthesis of cyclized TFA, carried out in parallel to compensate for the absence of a commercial standard, was successful with a satisfactory yield of 75.9%. This step yielded a purified and characterized product, which was essential for establishing calibration curves and identifying reaction products.
The calibration curve for cycled TFA, produced using toluene as an internal standard, showed good linearity (R² = 0.978), validating the HPLC-UV quantification method.
In conclusion, the work carried out made it possible to:
- Develop and validate an HPLC-UV method suitable for analyzing the system studied;
- Successfully synthesize and purify the target product, cyclized TFA;
- Lay the analytical foundations necessary for future confirmation of the enzymatic reaction catalyzed by Spl lipase.
The next steps will consist of repeating the complete enzymatic reaction and confirming the formation of the product by direct comparison with the synthesized standard.
