Alternative HPLC-DAD Direct-Phase Approach to Measurement of Enantiopurity of Lactic Acid Derivatives

Abstract

Lactic acid (LA) is a natural organic acid that can be used in a wide variety of industries. Recently, the production of lactic acid by fermentation protocols has gained relevance. However, these biotechnological processes often encounter challenges in the production of enantiopure D- or L-lactic acid. Thus, the measurement and control of the enantiopurity of lactic acid and its derivatives is a crucial step in these productions, especially when monomers have to be used to synthesize polymers, such as polylactic acid (PLA). In the present work, we propose a measurement of the enantiopurity of lactic acid mixtures with HPLC-DAD by using direct-phase conditions, which are mild, and a large set of different employable chiral columns. To this end, we report the synthesis of two new LA derivatives (2-nitrobenzyl 2-hydroxypropanoate and 2,4-dinitrobenzyl 2-hydroxypropanoate) and benzyl 2-hydroxypropanoate, with a selective and non-racemizing chemical functionalization. Then, three commercially available HPLC direct-phase chiral columns are tested to achieve a method for measuring the enantiomeric purity of lactic acid derivatives. The 2-nitrobenzyl lactate was chosen as the best lactic acid derivative and the Chiralpak IA column was chosen as the most effective chiral column to achieve a new and efficient protocol (Rs = 3.57 and α = 1.13) for the measurement of the enantiopurity of lactic acid.

1. Introduction

Lactic acid (LA) is a versatile and natural organic acid that can be produced chemically and, especially in recent times, biotechnologically [1]. It is commonly used as an acidifier, flavor enhancer and preservative in the food, pharmaceutical, leather and textile industries as well as to produce fine chemicals [2]. Large amounts of lactic acid are required to produce polylactic acid (PLA), a bio-based thermoplastic polyester used for higher technological applications [3,4,5]. However, enantiopure, or highly enantioenriched, LA monomers are required to yield an isotactic PLA with a high degree of crystallinity and better physical and rheological properties [6]. For example, the PLA stereo-complexes D-PLA and L-PLA have a high thermostability and a high melting point, about 50 °C higher than that of the respective DL-polymers [7]; thus, they are more appropriate for higher technological applications.

Recently, biotechnological protocols starting with biomass and agricultural waste, as alternative feedstock, have been widely exploited for LA production. However, these protocols are often limited to the production of mixtures of D- and L-lactic acid [8,9], instead of enantiopure materials. At this stage, the determination of the optical purity of lactic acid produced by biotechnological processes has become a crucial step to validate the production process of LA through fermentation and to check the quality of the monomer leading to polylactic acid. The development of chromatographic methods to determine the enantiopurity of lactic acid and lactic acid derivatives has been studied for about 30 years [10]. High-performance liquid chromatography (HPLC) is the most widely used chromatographic analytical technique for this purpose. However, several techniques of chromatographic analysis have also been explored. One of these is based on chiral derivatization combined with gas chromatography and mass spectrometry (GC–MS). L-Menthol and acetyl chloride were used to derivatize LA enantiomers that were subsequently separated on a DB-5 MS capillary column. This method was also successfully applied to the analysis of mouse plasma samples [11].

In 1999, two HPLC methods for the determination of lactic acid and its enantiomers in calf serum were described for the first time [12]. In 2007, Norton et al. [13] developed a novel method for the separation and simultaneous determination of urinary D- and L-lactic acid enantiomers based on high-performance liquid chromatography–tandem mass spectrometry (HPLC/MS/MS). The chiral separation was optimized utilizing the Chirobiotic TAG column, chosen because the polar ionic mobile phase (17% H₂O, 83% EtOH/0.12% HOAc/0.30% TEA delivered isocratically at 0.2 mL/min) was well suited for the separation of small polar molecules, such as lactic acid. Some years later, the same method was exploited by Henry et al. [14]. Urinary D- and L-lactic acid enantiomers were separated at 4 °C on an Astec Chirobiotic™ R chiral HPLC column. The mobile phase consisted of 15% (v/v) 33.3 mmol/L ammonium acetate and 85% (v/v) acetonitrile. Elution was performed at a flow rate of 0.7 mL/min. Xu et al. [15] reported a protocol based upon the use of a reverse-phase chiral column with a “Pirkle”-type stationary phase, using 1 mM copper (II) sulfate in water as the mobile phase. The addition of copper sulfate to the mobile phase is needed to generate a complex that can be detected spectrophotometrically by a diode array detector (DAD), avoiding a preliminary chemical derivatization step. The HPLC-DAD approach was also used to test various types of macrocyclic glycopeptide-based chiral stationary phases for chiral HPLC separation of lactic acid in reversed-phase mode. Chromatographic separation was performed using different analytical columns; for example, ristocetin was used with Chirobiotic R, vancomycin was used with Chirobiotic V, teicoplanin was used with Chirobiotic T, etc. Optical isomers of lactic acid were efficiently separated using chiral stationary phases based on teicoplanin (R_S = 1.9) and ristocetin (R_S = 1.7) modes at a column temperature of 25 °C [16]. All the HPLC techniques described above are based upon the use of a reverse-phase chiral column.

As this review shows, HPLC procedures require highly sensitive detectors, such as MS or DAD, especially when biological samples and low concentrations of analytes are involved. Moreover, lactic acid needs to be properly functionalized by introducing an appropriate chromophore to be detected by UV-Vis detectors (DAD) (Figure 1). At this stage, we reasoned that the introduction of a suitable chromophore could also change the polarity of lactic acid, thus exploiting the possibility of a direct-phase HPLC approach. This type of determination has the advantage that direct-phase chiral columns are cheaper than reversed-phase ones and a wide variety of different stationary phases are available, thus paving the way for a broad screening of potentially useful stationary phases. In this paper, considering our experience in this field [17,18,19,20], we have investigated a different HPLC-DAD approach based upon the use of a direct-phase chiral column for the measurement of the enantiopurity of lactic acid derivatives [21]. Our work hinges first on the functionalization step that changes the polarity and introduces a chromophore on the lactic acid. Then, a screening of different HPLC columns and conditions follows to individuate the better resolution of the enantiomers of lactic acid. Finally, a check of enantiospecificity of the functionalization is required to prevent the preliminary step from altering the enantiopurity of the analyte.

Different methods have been reported in the literature to obtain lactic acid esters, often requiring strong bases, high temperatures and anhydrous solvents. Gellmann et al. [22] synthesized the benzyl ester of lactic acid with a selective procedure that leaves the hydroxyl portion of LA unchanged. We modified their procedure by using a base such as potassium carbonate or triethylamine to generate the lactate and a water-miscible aprotic polar solvent such as acetonitrile or only water to which we added the benzyl halide. The products were analyzed by HPLC-DAD with three commercially available direct-phase chiral columns: Chiralcel OD-H, Chiralpak IA and (R,R)-Whelk O2. In this screening, different separation conditions (stationary phase, mobile flow, eluent polarity) were investigated to determine the best method for separating the lactic acid enantiomers.

2. Materials and Methods

2.1. Synthesis

All reagents and solvents purchased were of the highest commercial quality and used without further purification.

DL-Lactic Acid (85% in water) (CAS RN50-21-5) and L-lactic acid (85% in water) (CAS RN79-33-4) were purchased from TCI Europe (Paris, France). NEt₃ (CAS RN121-44-8), K₂CO₃ (CAS RN584-08-7), NaOCH₃ (CAS RN124-41-4), DBU (1,8-diazabiciclo[5.4.0]undec-7-ene) (CAS 229-713-7), NaOH (CAS 1310-73-2) and CH₃CN (CAS 75-05-8) were purchased from Merck (Rahway, NJ, USA) and were used for the synthesis. Preparative column chromatography was carried out using Macherey–Nagel silica gel (Macherey–Nagel, Düren, Germany) (60, particle size 0.063–0.2 mm). Macherey–Nagel aluminum sheets with silica gel 60 F254 were used for TLC analyses. New compounds were characterized by ¹H-NMR, ¹³C-NMR, HR–MS analysis and FTIR.

2.2. Instruments

¹H-NMR and ¹³C-NMR spectra were recorded on an Agilent 500 spectrometer (Agilent Technologies, Santa Clara, CA, USA) at 500 and at 125 MHz, respectively. High-resolution mass spectra were recorded with an Agilent HPLC QTOF spectrometer via direct infusion of the samples. Liquid chromatographic studies were performed on an Agilent 1100 analytical HPLC system consisting of a Manual Injector (25 µL), an Agilent HPLC Quaternary pump and Agilent DAD. The sequence programming and collection of chromatographic data were performed using software from Agilent. FT-IR spectra were measured on a spectrophotometer Bruker Invenio-S FT-IR (Bruker, Billerica, MA, USA) using dry KBr pellets with the scan range 4000–400 cm⁻¹.

2.3. Chiral Chromatography

We chose the analytical (25 × 0.46 cm) chiral HPLC columns Chiralcel OD-H, Chiralpak IA and (R,R)-Whelk-O2 as being particularly representative among the many commercially available columns. In detail, the Chiralcel OD-H has a cellulose tris(3,5-dimethylphenylcarbamate) coated on a 5 µm silica gel packing. The Chiralpak IA is an amylose tris (3,5-dimethylphenylcarbamate) immobilized on a 5 µm silica gel stationary face. The (R,R)-Whelk-O2 is a chiral stationary phase based on 1-(3,5-dinitrobenzamido)-1,2,3,4-tetrahydrophenanthrene as the chiral selector covalently bonded to trifunctionalized 5 µm silica nanoparticles.

The HPLC separations were performed at room temperature with an optimal flow rate of 0.5 mL/min. Mixtures of n-hexane/2-propanol were used for normal-phase HPLC. The compounds were detected with DAD between 190 and 300 nm. Lactate 1b was best detected at 210 nm; lactates 2b and 3b were best detected at 254 nm.

3. Results and Discussion

After preliminary tests, we chose to transform the lactic acid to be analyzed into the corresponding benzylic lactates. The benzylic fraction was chosen because it is more reactive in nucleophilic substitution reactions even towards non-optimal nucleophiles such as lactate. This choice allowed us to use mild reaction conditions avoiding racemization of the product. Furthermore, the effect of introducing one or two nitro groups on the benzylic fraction was evaluated, first to improve the HPLC separation conditions and then to enhance the UV absorbance of the analyte.

3.1. Screening of Reaction Conditions

We tested different bases to carry out the selective formation of benzylic lactates, without involving the hydroxyl fraction. The racemic lactate reacted with three benzyl halides (benzyl bromide 1a, 2-nitrobenzyl chloride 2a and 2,4-dinitrobenzyl chloride 3a) to obtain a pair of enantiomers of the benzyl lactate esters.

Below, we report the synthesis of two new LA derivatives (2-nitrobenzyl 2-hydroxypropanoate and 2,4-dinitrobenzyl 2-hydroxypropanoate) and benzyl 2-hydroxypropanoate (1b-3b) (Figure 2 and

Table 1. Optimization of selective synthesis of lactate esters.

Entry	Benzyl Halide	Product	Base a	Solvent	Temperature	t (h)	Yield (%)
1	1a	(±)-1b	DBU c	Toluene	90 °C	12	50
2	2a	(±)-2b	K2CO3	CH3CN	60 °C	12	67
3	2a	(±)-2b	K2CO3	EtOH	Rt	12	Nr b
4	2a	(±)-2b	KOH	EtOH	Rt	12	Nr b
5	2a	(±)-2b	NaOCH3	Acetone	Rt	12	Nr b
6	2a	(±)-2b	NaOH	Acetone	50 °C	2	Nr b
7	1a	(±)-1b	NEt3	CH3CN	60 °C	4	72
8	2a	(±)-2b	NEt3	CH3CN	60 °C	4	89
9	3a	(±)-3b	NEt3	CH3CN	60 °C	4	56
10	2a	(−)-2b	NEt3	CH3CN	60 °C	4	82
11	2a	(±)-2b	NEt3	no solvent	60 °C	1	50
12	2a	(±)-2b	NEt3	H2O	60 °C	1	38

^a Molar ratio: lactic acid/Base/Benzyl halide 1:1:1. ^b No reaction. ^c DBU (1,8-diazabiciclo[5.4.0]undec-7-ene).

).

In the first instance, racemic lactic acid was reacted with DBU (1,8-diazabiciclo[5.4.0]undec-7-ene) as the organic base in toluene (Table 1, entry 1) [22]. After 12 h of reaction at 90 °C using benzyl bromide 1a as the benzyl halide, we obtained a 50% yield of the isolated product (±)-1b. Later, we tested some inorganic bases (potassium carbonate, potassium hydroxide, sodium hydroxide, sodium methylate), in polar protic and aprotic solvents, at room temperature or at reflux. Except for potassium carbonate in the polar aprotic solvent at reflux (Table 1, entry 2), the remaining inorganic bases did not yield useful results (Table 1, entries 3–6). Then, we tested triethylamine in acetonitrile under reflux (Table 1, entry 7). After 4 h, the reactions yielded lactates 1b, 2b and 3b in good yields (56–89%, entries 7–10, Table 1). At this stage, we considered these reaction conditions to be satisfactory, and then we tested them with a sample of enantiopure lactic acid (Table 1, entry 10). L-lactic acid and 2a were reacted in the best reaction conditions to yield (−)-2b (82%, entry 10, Table 1). The obtained lactate, (−)-2b, was found to be enantiopure (HPLC, Figure 3b,d), thus giving us the confidence that the optimized reaction conditions do not cause racemization of these compounds. Lactates 1b, 2b and 3b were isolated and purified to provide a complete characterization. We checked the possibility of extending the validity of the method even to mixtures with a higher water content. The reaction was repeated at 60 °C in solvent-free conditions by simply mixing L-lactic acid (85% in water), triethylamine and 2a. The desired product (±)-2b was obtained after 1 h in 50% yield (Table 1, entry 11). The reaction was, finally, repeated using water as a solvent to simulate the biological environment or a dilute solution of the analyte. In this case, after a 1 h reaction time, product 2b was obtained in 38% yield (Table 1, entry 12). These crude reaction mixtures were analyzed with HPLC without any purification to set up a method for the fast indirect determination of the enantiopurity of lactic acid.

3.2. Chiral HPLC-DAD Analysis of Lactates

The racemic products ((±)-1b, 2b and 3b) were synthesized in previous steps and injected using different chiral columns and conditions (Supplementary Materials, Figure S14). The best separation conditions for racemic lactates (1b–3b) are reported in

Table 2. Chiral HPLC-DAD analysis of lactate derivatives.

Entries	Compound	Column	α 1	Rs 2	Elution Conditions (n-Hexane:i-Propanol)	λobs
1	1b	Chiralcel OD-H	1.05	1.77	98:2	210 nm
2		Chiralpak IA	1.04	1.98	98:2	210 nm
3		(R,R)-Whelk O2	- 3	- 3	95:5	210 nm
4	2b	Chiralcel OD-H	1.08	3.24	90:10	230, 254 nm
5		Chiralpak IA	1.13	3.57	80:20	230, 254 nm
6		(R,R)-Whelk O2	1.04	1.17	90:10	230, 254 nm
7	3b	Chiralcel OD-H	1.09	2.03	70:30	230, 254 nm
8		Chiralpak IA	1.08	2.12	70:30	230, 254 nm
9		(R,R)-Whelk O2	- 3	- 3	80:20	230, 254 nm

¹ Separation factor, as a ratio between capacity factors (Table S1). ² Resolution is calculated using the separation of two peaks in terms of their average peak width at the base (Table S1). ³ No separation.

. The peaks chosen as representative of enantiomers are those with equal area and equal absorption profile. The D and L peaks were confirmed by injecting a standard of racemic solution of 1b, 2b and 3b and observing the resulting chromatogram (Table 2, Figure 3 and Supplementary Materials, Figure S14). For 2b, we also injected pure (−)-2b (Figure 3). From the above, it was noted that the first peak, with a retention time of 19.77 min (Area%: 45.8028), was for enantiomer L and the second peak with a retention time of 21.56 min (Area %:40.2158) was for enantiomer D of compound 2b.

First, lactate ester 1b was analyzed. As reported (Table 2, entries 1–3), the separation factors and resolution factors are low, and in one case (Table 2, entry 3) no separation was observed, even after increasing the hexane/i-propanol ratio. On the other hand, lactate 2b was more successful (Table 2, entries 4–6). Interesting separations were achieved with the first two columns, whereas (R,R)-Whelk O2 showed lower performance (α = 1.04; Rs = 1.17). Between Chiralcel OD-H and Chiralpak IA, the latter gave the best separation factor (1.13 vs. 1.08). Finally, lactate 3b also gave satisfactory separations with the first two columns (Table 2, entries 7–8), whereas (R,R)-Whelk O2 gave no separation.

By comparing all these results, it is found that the best separation was achieved with the Chiralpak IA column on lactate 2b (Figure 3). From the values of α and Rs obtained, in particular with the Chiralpak IA column, and by comparing them with the values of α and Rs from similar separation and detection protocols already present in the literature (Rs 1.9) [16], it appears evident that our procedure is an efficient and alternative method of analysis of lactic acid. Thus, 2-nitrobenzyl lactate 2b can be considered a successful derivative for the measurement of the optical purity of mixtures of lactic acid. Figure 3 for 2b also confirms that the derivatization reaction of lactic acid does not cause any modification of the chirality of the mixture. In fact, when using the pure enantiomer L of lactic acid as a substrate, the integrity of the chiral compound is maintained.

4. Experimental Part

Synthesis of Lactate Derivatives

The lactate derivatives (±)-1b, (±)-2b, (−)-2b and (±)-3b were synthesized in a 50 mL round-bottomed flask by adding 0.595 mmol of lactic acid (equal to 63 mg of solution of 85% lactic acid in water) to 0.595 mmol triethylamine in 2 mL of acetonitrile or water. The mixture was reacted for 30 min at 60 °C. After this time, the benzyl halide (0.595 mmol) was added and the mixture was reacted for 3.5 h at 60 °C. After this time, the solvent was evaporated under vacuum. The crude mixture was extracted three times with ethyl acetate and purified by silica gel column chromatography using ethyl acetate/hexane (4:6) as the eluents to give the following products:

Benzyl 2-hydroxypropanoate -(±)-1b. The product is a colorless oil (77 mg, 72% yield).

¹H-NMR (CDCl₃, 500 MHz) δ 7.37–7.33 (m, 5 H), 5.20 (s, 2 H), 4.31 (q, J = 5.0 Hz, 1 H), 2.89 (broad s, 1 H), 1.43 (d, J = 5.0 Hz, 3 H).

¹³C-NMR (CDCl₃, 125 MHz) δ 175.4, 135.2, 128.6, 128.5, 128.2, 67.2, 66.8, 20.4.

HRMS (ESI-TOF), m/z: calcd for [C₁₀H₁₂O₃ + Na]⁺ 203.0679, found [M + Na]⁺ 203.0670.

FT-IR (KBr): ν 3449.61 (m), 1739.77 (m), 1455.05 (m), 747.58 (m), 698.96 (m) cm⁻¹.

2-nitrobenzyl 2-hydroxypropanoate-(±)-2b. The product is a yellow oil (120 mg, 89% yield).

¹H-NMR (CDCl₃, 500 MHz) δ 8.15–8.10 (m, 1 H), 7.71–7.63 (m, 1 H), 7.58–7.51 (m, 2 H), 5.64 (d, J = 15 Hz, 1 H), 5.57 (d, J = 15 Hz, 1 H), 4.39 (q, J = 7.0 Hz, 1 H), 1.48 (d, J = 7.0 Hz, 3 H).

¹³C-NMR (CDCl₃, 125 MHz) δ 175.1, 133.9, 133.8, 131.2, 129.2, 129.1, 125.2, 66.8, 64.0, 20.4.

HRMS (ESI-TOF), m/z: calcd for [C₁₀H₁₁NO₅ + Na]⁺ 248.0529, found [M + Na]⁺ 248.0520.

FT-IR (KBr): ν 3455.83 (s), 2987.95 (m), 1746.44 (s), 1613.06 (m), 1528.51(s), 1347.31(s) cm⁻¹.

2-nitrobenzyl 2-hydroxypropanoate-(−)-2b. The product is a yellow oil (109 mg, 82% yield). [α]_D = −9.06 (c = 0.75, CH₃CN).

2,4-dinitrobenzyl 2-hydroxypropanoate-(±)-3b. The product is a colorless oil (90 mg, 56% yield).

¹H-NMR (CDCl₃, 500 MHz) δ 8.90 (d, J = 2.4 Hz, 1 H), 8.52 (dd, J = 8.5, J = 2.4 Hz, 1 H), 7.83 (d, J = 8.5 Hz, 1 H), 5.70 (s, 1 H), 5.69 (s, 1 H) 4.40 (q, J = 6.9 Hz, 1 H), 1.51 (d, J = 6.9 Hz, 3 H).

¹³C-NMR (CDCl₃, 125 MHz) δ 174.7, 147.6, 147.4, 138.1, 130.0, 127.9, 120.7, 66.8, 63.2, 20.4.

HRMS (ESI-TOF), m/z: calcd for [C₁₀H₁₀N₂O₇ + Na]⁺ 293.0380, found [M + Na]⁺ 293.0386.

FT-IR (KBr): ν 3440.97 (s), 2986.90 (m), 1747.23 (s), 1606.41 (s), 1536.78 (s), 1347.66 (s) cm⁻¹.

5. Conclusions

In summary, we reported the synthesis of three lactate esters, after a screening of reaction conditions, to yield a selective synthetic methodology that does not cause racemization of the stereogenic center of the lactic acid. We also verified that our reaction conditions can be carried out in solvent-free conditions or in aqueous mixtures. This simple and general methodology was applied to a new and alternative analytical protocol for the measurement of the enantiomeric purity of lactic acid, even in a non-anhydrous environment. The newly synthesized lactate esters 1b, 2b and 3b were analyzed by direct-phase chiral HPLC-DAD with stationary phases differing both in the chiral selector and in the methodology of immobilization. The Chiralcel OD-H and Chiralpak IA columns show similar behaviors in retention times, enantioselectivity and resolution. On the other hand, the Whelk-O2 column is low-performing in this type of separation. The resolution of the Chiralpak IA column is greater than that of the Chiralcel OD-H column. The resolution increases passing from ester 1b to ester 2b, in which the addition of one nitro group in the ortho-position is highly beneficial. On the other hand, the resolution decreases passing from ester 2b to ester 3b, in which the addition of a second nitro group is detrimental; in fact, retention times increased, but the overall resolution decreased. Ester 2b is indeed the best choice for the functionalization of lactic acid (α 1.13 and Rs 3.57).

In conclusion, we define a new protocol for the measurement of the enantiomeric purity of lactic acid by direct-phase chiral HPLC-DAD. This method involves a preliminary functionalization of the starting material by means of a reaction that is potentially feasible on biological samples and analytes at low concentrations in the 2-nitrobenzyl ester. Subsequently, the measurement of the enantiomeric purity of the same ester is performed using the Chiralpak IA column.