Ethyl Lactate as a Pharmaceutical-Grade Excipient and Development of a Sensitive Peroxide Assay

Pharmaceutical Technology, Pharmaceutical Technology-05-02-2009, Volume 33, Issue 5

The authors investigate whether the addition of an antioxidant could be used to stabilize the solvent ethyl lactate by preventing the formation of peroxides

The use of organic solvents for coating solid oral dosage forms has diminished over the past decade because many harmful effects are associated with exposure to these solvents. Acute effects of solvent inhalation in humans and animals include narcosis, anesthesia, central-nervous-system depression, and respiratory problems (1). As a result, the industry has shifted from organic-based systems to aqueous-based technologies. Tablet coating is a prime example of this change.

Advantages and properties of ethyl lactate

Ethylcellulose and acrylic polymers, which traditionally used either acetone or isopropyl alcohol (IPA), have been replaced by popular latex and pseudolatex systems such as the Eudragit (Evonik, Darmstadt, Germany) and Aquacoat (FMC BioPolymer, Philadelphia) product lines of coating dispersions. The advantages of these systems include a reduced health risk for the operator, a reduction in environmental contamination, low cost, and low risk of combustion or explosion. A disadvantage of aqueous coating systems is that they generally require more energy because of the low vapor pressure of water. Also, an increased quantity of plasticizer is required in an aqueous-based system to obtain a coating finish that is comparable with a coating finish in an organic-based system. Assuming appropriate equipment is used, organic-based systems are limited because of toxicity or environmental issues. A novel approach for pharmaceutical applications would be a user-friendly organic solvent that is environmentally friendly and poses little health risk. An example is ethyl lactate of pharmaceutical grade (ELPG) (2).

Crude ethyl lactate is prepared by the esterification of lactic acid and ethanol (3). Lactic acid and ethanol are from renewable sources and may be obtained by fermenting corn starches, an environmentally acceptable process (4). Ethyl lactate can be manufactured from a continuous process (3). Enhanced yields of ethyl lactate may be prepared by using an acid-catalyzed preparation step, and zeolites may be used to remove water from the reaction vessel, a byproduct of the esterification process (5). The pharmaceutical grade is prepared by purifying raw ethyl lactate (2).

Ethyl 2-hydroxypropionate (ethyl lactate) is used in many applications, including inks, flavorings, coatings, silicone oil and grease removal, and in the semiconductor industry (6). It is a replacement solvent for N-methylpyrrolidone, toluene, acetone, and xylene (7). As a "green" solubilizing agent, ethyl lactate has many advantages compared with other organic-based solubilizing agents. It may be readily purified because it is prepared from natural and renewable sources. It is completely biodegradable to carbon dioxide and water. It is easy and inexpensive to recycle (7). It is approved by the US Food and Drug Administration as an adjuvant solubilizing ingredient for flavors and pharmaceutical dyes. (8). It is a nonozone-depleting chemical, poses no hazard as an air pollutant, and is noncarcinogenic and noncorrosive (7). The authors propose that this ultrapure pharmaceutical grade of ethyl lactate is suitable in various pharmaceutical applications.

Ethyl lactate is susceptible to degradation by hydrolysis and oxidation, and as such is packaged under an inert blanket (e.g., nitrogen or argon) (9). It is highly recommended that it be stored under inert conditions to prevent oxidation. The main routes of degradation are summarized in Figures 1(a) and 1(b). Under dry conditions, degradation because of oxidation is most prevalent. Peroxides are formed during or after the oxidation process and may be monitored as a direct indication of the extent of oxidation of ethyl lactate that has occurred. Previous efforts used antioxidant materials to stabilize a similar organic solvent, ethyl oleate, a nonaqueous solvent used for injections (10, 11). The focus of the research in this article deals with the stabilization of ELPG by the inclusion of various generally recognized as safe (GRAS) excipients with certain antioxidant properties. The authors also present an improved method for determining peroxide levels within ethyl lactate solutions.

Figure 1: The basic steps of degradation of ethyl lactate of pharmaceutical grade: (a) the mechanism of hydrolysis to form acid and alcohol and (b) oxidation forming an unstable pyruvate intermediate and water. (ALL FIGURES ARE COURTESY OF THE AUTHORS)

Materials and methods

The following materials were purchased: peroxide test strips, 0–100 ppm (Merckoquant, EM Science, Gibbstown, NJ); ascorbic acid (AA), butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), citric acid (CA), methionine (METH), sodium ascorbate (SA), sucrose (SU), and tocopherol (vitamin E [VE]), 2-propanol, high performance liquid chromatography grade and 3% w/w hydrogen peroxide (Spectrum Chemicals, Gardena, CA); and ELPG (Riba-Fairfield, Decatur, IL).

ELPG stability study. Partially oxidized ELPG (30 ppm peroxide) was dispensed (15 mL) into 20 glass scintillation vials (VWR International, West Chester, PA). The following antioxidants were selected from the GRAS list: AA, BHA, BHT, CA, METH, SA, SU, and VE. Each antioxidant (3% w/v) was weighed individually into a separate scintillation vial. This concentration was above the saturation solubility for AA, CA, METH, SA, and SU (i.e., residual undissolved powder was observed in each of these containers at 25 °C). Additionally, two control groups were prepared: a control group with an ambient-air head (C1) and one control group containing a nitrogen head (N2). A replicate study of the remaining 10 scintillation vials was also prepared for storage at 40 °C. During the six-month stability test, each vial was opened daily for the first seven days, and peroxide levels were determined using peroxide test strips. N2 vials were sealed with a septum, and nitrogen was flushed through to obtain a fresh nitrogen head daily. After seven days, the samples were checked at one, two, four, and six months to determine peroxide levels. Additionally, samples were observed and photographed during the same sampling periods, and a scoring system of 1–5 was applied for yellow discoloration (i.e., 5 being the most discolored and 1 being a colorless solution). Ethyl pyruvate, the degradation product from the oxidation process, is known to have a yellow color.

Solubility of AA in ELPG. An ultraviolet–visible (UV–vis) spectrophotometric determination of AA concentration in ELPG was used in a method developed from Memon et al. (12). A solution was prepared by dissolving 1,10-phenanthroline monohydrate (1.97 g) into 1.0 N HCl (10 mL). Iron (III) ammonium sulfate (1.67 g) was added to the solution before dilution with deionized water to make a 1-L stock solution. This iron (III)-tris-1,10-phenanthroline complex (I3PC) has no absorption peak in the visible spectrum (at 510 nm) until it is reduced to the iron (II)-tris-1,10-phenanthroline complex (I2PC).

To determine the maximum wavelength of absorption, I2PC was prepared by adding a 5-mg/mL ascorbic-acid solution (5 mL) to a 5% v/v solution of the I3PC stock solution (5 mL) prepared as previously described. A scan of the resultant solution revealed a maximum absorption wavelength (λ max) of 510 nm. Absorption of this solution was checked at 0.5, 1, 5, 10, 15, 20, 30, and 60 min to determine the stability of the I2PC formed in solution.

Aliquots of 5% v/v I3PC stock solution (3.5 mL) were added to separate aliquots of AA standard solutions (6.5 mL) prepared at 1, 2, 3, 4, 5, 7, and 10 mg/L. Each of the seven samples was allowed to react for precisely 10 min, prior to analysis. A calibration curve was constructed at 510 nm. AA (1.5 g) was weighed into 10 separate brown-glass scintillation vials (because AA is photosensitive), and ELPG (5 mL) was added. The vials were sealed and shaken at 25 °C (five vials) and at 37 °C (five vials) for 48 h. An aliquot (1 mL) of each vial was dispensed into 1.5-mL centrifuge tubes, and the ELPG was evaporated to dryness in a vacuum desiccator for 24 h. Each dried sample was reconstituted into 1 L of deionized water, and the same procedure for determination of the calibration curve as previously described was used to ascertain each unknown concentration.

Solubility of BHT in ELPG. To determine an absorption wavelength, a 10 mg/mL solution of BHT in a mixture of IPA and deionized water (1:1) was prepared, and a UV–vis scan revealed a λ max of 280 nm. Following this step, a stock solution of BHT (100 mg) in a 100-mL mixture of IPA and deionized water (1:1) was prepared. A calibration curve was constructed at concentrations of 100, 90, 80, 70, 60, 50, 40, 30, 20, and 10 mg/L. BHT (9 g) was weighed into 10 separate brown-glass scintillation vials, and ELPG was added (5 mL). The vials were sealed and shaken at 25 °C (five vials) and 37 °C (five vials) for 48 h. An aliquot (1 mL) of each vial was dispensed into a 1.5-mL centrifuge tube, and the ELPG was evaporated to dryness in a vacuum desiccator for 24 h. An 18-gauge needle was inserted into the top of each sample during the drying period with a stainless-steel wire to prevent the formation of a BHT crust layer. A crust layer of BHT, which would prevent further evaporation of the ELPG, could form at the interface of the solvent and air. Each dried sample was washed with 3 mL of IPA into a scintillation vial before adding 3 mL of deionized water. An aliquot (1 mL) from each reconstituted vial was diluted into 100 mL of the mixture of IPA and deionized water (1:1). The unknown concentration was determined by analysis at 280 nm. It was necessary to reconstitute in the IPA because interference at 280 nm was observed with pure ELPG.

Table I. Construction of a reverse-absorption calibration curve for hydrogen peroxide (H202) and iron (II)-tris-1,10-phenanthroline (I2PC).

Accurate determination of peroxide concentration in ELPG. A method was developed using a reverse UV–vis absorbance technique based on the oxidation of a I2PC to I3PC (as opposed to the reduction of I3PC to I2PC as previously described). A 2% w/v stock solution of I2PC was prepared. Iron (II) ammonium sulfate (1.3635 g) and 1,10-phenanthroline (0.9717 g) was weighed into a 1-L volumetric flask, and 1.0 N hydrochloric acid (10 mL) was added. The solution was made with deionized water to form a 2% w/v blood-red complex in solution. A 1% w/v hydrogen peroxide solution was prepared by adding an aliquot of 3% w/v hydrogen peroxide (1 mL) to deionized water (2 mL). The freshly prepared 1% w/v hydrogen peroxide solution was diluted with deionized water (1:10) to give a stock solution of hydrogen peroxide (1000 mg/L). The stock solution was diluted to obtain the standard concentrations of peroxide as indicated (see Table I), and a calibration curve was constructed. Five aliquots of ELPG (each 200 μL) were added to 2% I2PC (1.5 mL) and deionized water (2.3 mL) to determine the unknown concentration of the peroxides in the sample.

Figure 2: Results of a stability study of ethyl lactate of pharmaceutical grade (ELPG) performed at 25° C. Each antioxidant was added at 3% w/v into ELPG. The antioxidants studied were acorbic acid (AA), butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), citric acid (CA), methionine (METH), sodium ascorbate (SA), sucrose (SU), and tocopherol (vitamin E [VE]). C1 and N2 represent the control groups. (ALL FIGURES ARE COURTESY OF THE AUTHORS)

Results and discussion

ELPG stability study. At 25 °C, it is apparent that all the antioxidant additives reduced the concentration of peroxides in the ELPG, and several showed reduced levels of peroxide for some period of time (see Figure 2). The best performing additives were AA, BHT, SA, and VE because the peroxide concentrations did not exceed 30 mg/L for at least six months. BHA, CA, and SU displayed an initial reduced concentration of peroxides for up to one month, but then there was a rapid increase in levels, so these additives were considered to be ineffective at the concentrations used. At 40 °C, there is a similar display of effectiveness (see Figure 3). The profiles for BHA, CA, and SU, however, showed a rapid increase in peroxide concentrations after three days. This observation indicated that these additives were only effective for a limited period and so could be eliminated as potential stabilizing antioxidants. The control groups with no additives showed similar degradation at 25 and 40°C, and 100 mg/L peroxide levels were observed after seven days. A rapid rise and fall of peroxide levels were observed within the control group containing the purged nitrogen blanket head. This phenomenon was observed within several groups during the course of the study, but it is not an accurate indication of the actual peroxide levels within the container. The sharp rise and fall in peroxide quantification was merely an artifact of the test-strip determination method, as this method relies on the operator's assessment of a colorimetric change observed on the test strip when exposed to peroxide levels. This type of assessment is generally a guide and may be subject to individual interpretation. For the ELPG sample that contained METH, the results indicated a lag time in the reduction of peroxide concentration that lasted for up to three to four days at 25 °C. This lag time was observed to be more rapid at 40 °C and was probably because of the solubility of METH within the ELPG. (The METH was more soluble at this elevated temperature and therefore had increased antioxidant capacity within the ELPG.)

Figure 3: Results of the stability study of ethyl lactate of pharmaceutical grade (ELPG) performed at 40 °C. Each antioxidant was added at 3% w/v into ELPG. The antioxidants studied were ascorbic acid (AA), butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), citric acid (CA), methionine (METH), sodium ascorbate (SA), sucrose (SU), and tocopherol (vitamin E [VE]). C1 and N2 represent the control groups. (ALL FIGURES ARE COURTESY OF THE AUTHORS)

Observing the vials upon storage revealed another issue with the long-term stability of the samples. Although the peroxide levels had been deemed low, in some cases, degradation occurred during the stability study. The numbers assigned to a yellowish colorimetric change for samples stored at 25 °C are specified in Table II. No discernible color change was visible in the ELPG containing AA, BHT, CA, or SU in addition to the control samples C1 and N2. The samples containing BHA and VE showed only a moderate change to a pale straw yellow color. The samples of METH and SA stored at 25 °C showed a strong change to a deep yellow color that occurred more rapidly in the sample containing SA. The same trend in yellowing was observed for the samples when stored at 40 °C (see Table III). The onset of discoloration, however, was much more rapid and turned to a darker yellowish-brown color at this temperature.

Table II. Observed colorimetric change of stability vials upon storage at 25 °C.*

Taking into account the peroxide concentration data obtained from the stability studies at the two temperatures, it was clear that the most effective antioxidants were AA, BHT, METH, and SA. By observing the vials for any colorimetric change, METH and SA could be further eliminated because they displayed some yellowing during the tests. Elimination of those two antioxidant additives prompted further investigation of AA and BHT as potential stabilizers for ELPG.

Table III. Observed colorimetric change of stability vials upon storage at 40 °C.

Solubility of AA in ELPG. As previously described, the proposed antioxidants were added to ELPG to obtain a 3% w/v mixture. In some instances, the antioxidant was completely dissolved in the ELPG. In the case of AA, however, excess powder was observed at the bottom of the sealed scintillation vial. This situation is highly undesirable from a commercial point of view because the presence of excess powder within a solvent vehicle has the potential to interfere with formulation design and can severely limit the diversity of manufacturing processes.

After reaction of a 5% v/v solution of the I3PC stock solution with an AA solution (5 mg/mL), the resultant I2PC absorption at 510 nm reached a maximum value after 10 min and was stable for up to 20 min. As a result, the authors decided that 10 min of developing time was needed to allow for full conversion of I3PC to I2PC and to maintain maximum absorption.

After determining the calibration curve for the various concentrations of AA, the solubility of AA in ELPG was determined. This was only 33 ± 5 mg/L at 25 °C and 42 mg/L at 37 °C. Because this stability group was observed to have excess AA powder within the ELPG throughout the entire course of the stability study, it is safe to assume that the ELPG maintained a saturated solution of AA at approximately 33 mg/L. Furthermore, the solubilized AA was the active component that was able to act as an antioxidant within the ELPG. Because the amount of AA was such a low concentration, it was hypothesized that solubilized AA was constantly being removed from the solution after reacting with the oxygen radicals, thus allowing more of the excess AA to be dissolved and to maintain a constant saturated solution of AA. This hypothesis could be further investigated by monitoring the stability of a 33 mg/L solution of AA in ELPG.

Solubility of BHT in ELPG. In the case of BHT, no excess powder was observed with a 3% w/v solution in ELPG, so it was clear that a nonsaturated solution had been formed within the solvent. The absence of excess powder within a solvent vehicle is desirable from a commercial point of view to prevent potential cross-contamination during processing. Additionally, the absence of powder in the ELPG may negate the need for a "settling" period (where the antioxidant powder is able to settle out of the ELPG after shipping and before use.

It may be assumed that the only component of BHT that is active in limiting the degree of oxidation that can occur within the ELPG is that which is in solution. It was relevant, therefore, to determine the solubility of BHT in the ELPG (not previously reported in the literature). At 25 °C, the solubility of BHT in ELPG was 0.311 ± 0.004 g/mL, and at 37 °C, the solubility was 0.378 ± 0.002 g/mL. These results reveal that BHT demonstrated an extremely high degree of solubility within ELPG. BHT can be solubilized to a maximum level of approximately 30% w/v, which is considerably higher than the initial 3% w/v used throughout the stability study. Considering the extended stability allowed by using a 3% w/v solution of BHT within ELPG, it is reasonable to assume that higher concentrations of BHT would allow for even longer stability by reducing the potential of solvent oxidation. This aspect of the research warrants further investigation to determine the optimal levels of BHT required for stabilization of ELPG over specific time periods.

Accurate determination of peroxide concentration in ELPG. A calibration curve obtained for the standard peroxide concentration using the reverse-absorbance method is shown in Figure 4. The curve displays good linearity (R2 value = 0.9994) with an absorbance intercept of 4.2 ×10-3 and a slope of 1 ×10-4, in accordance with the Beer–Lambert law. The initial concentration of peroxides within the partially oxidized sample of ELPG was 42 ± 8 mg/L (± standard deviation). This peroxide concentration is comparable to a value of 30 mg/L obtained from the peroxide test-strip analysis. Although the standard deviation of the analysis is relatively high, it may be considered an improvement in detection compared with the peroxide test-strip method. The biggest problems encountered while using the test strips were operator-dependent variability and variation in the uniformity of the colorimetric change associated with the developing time of the test strips. The second problem has the potential to cause great variation in the data as observed in the stability study data (see Figures 2 and 3). It is an artifact that appears with the peroxide concentrations in the control samples when the concentration peaks at 100 mg/L and suddenly drops to 30 mg/L. This type of observed observation is the biggest variability that can occur because the color change on the test strip is rapid within this concentration range. It was also highly likely that this fluctuation in concentration also occurred at the low range of analysis, but was less obvious from the data range under investigation during these studies.

Figure 4: Calibration curve used for the accurate determination of peroxide in ethyl lactate of pharmaceutical grade using a reverse–absorption method.

Conclusion

Pharmaceutical-grade, purified ethyl lactate showed extended stability when AA or BHT were used as antioxidants. ELPG is an attractive solubilizing agent because of its low cost, low toxicity, environmentally sound nature, and its derivation from renewable resource. Stabilized ethyl lactate has potential in various pharmaceutical applications such as tablet coating, granulation, taste-masking, and inhalation propellants. Additionally, it may improve the solubility of many active pharmaceutical agents during processing.

Jason T. McConville* is an assistant professor and Thiago C. Carvalho is a graduate student and PhD candidate at the College of Pharmacy, Division of Pharmaceutics, University of Texas at Austin, Austin, TX 78712–0231, tel. 512.471.0942, jtmcconville@mail.utexas.edu. Shawn A. Kucera is a senior applications scientist at Evonik–Degussa Corporation, and Elizabeth Garza is an MD student at the Pritzker School of Medicine, University of Chicago.

*To whom all correspondence should be addressed.

Submitted: June 10, 2008. Accepted: Aug. 12, 2008.

What would you do differently? Submit your comments about this paper in the space below.

References

1. "Organic Solvent Neurotoxicity" in Current Intelligence Bulletin (National Institute for Occupational Safety and Health, Atlanta, GA), No. 48, 1987.

2. J.T. McConville et al., "The Use Of Antioxidants to Stabilize Ethyl Lactate," presented at the Annual Meeting and Exposition of the American Association of Pharmaceutical Scientists, Baltimore, MD, Nov. 2004.

3. N. Asthana et al., "A Continuous Reactive Separation Process for Ethyl Lactate Formation," Org. Process Res. Dev. 9 (5), 599–607 (2005).

4. R.A. Sheldon, "Green Solvents for Sustainable Organic Synthesis: State of the Art," Green Chemistry 7 (5), 267–278 (2005).

5. J.J. Jafar, P.M. Budd, and R. Hughes, "Enhancement of Esterification Reaction Yield Using Zeolite—A Vapor Permeation Membrane," J. Membr. Sci. 199 (1–2), 117–123 (2002).

6. "Ethyl Lactate—Overview" in Solvent Alternatives Guide, www.p2pays.org/ref/19/18161/alt.cfm-id=yl&cat=ov.htm, accessed Apr. 15, 2009.

7. M. Henneberry, "Green Solvents: Agrochemicals in Place of Petrochemicals," Paints & Coatings Industry, www.pcimag.com/Articles/Feature_Article/42581cc0096a7010VgnVCM100000f932a8c0, accessed Apr. 15, 2009.

8. "Food and Drug Administration Department of Health and Human Services," in the Code of Federal Regulations, Title 21, Food and Drugs (Government Printing Office, Washington, DC), 2005.

9. Material Safety Data Sheet for Ethyl-(S)-Lactate, Purac America, (Lincolnshire, IL), 2005.

10. N.M. Nikolaeva and M.K. Gluzman, "Conditions for Stabilizing Ethyl Oleate During Storage," Farmatsiya 26 (1), 25–28 (1977).

11. J.M.K. Gluzman and N.M. Nikolaeva, "Method of Obtaining Ethyl Oleate, A Solvent for Parenteral Administration," Farm. Zh. 27 (6), 57–61 (1972).

12. M.A. Memon et al., "Spectrophotometric Determination of Ascorbic Acid in Pharmaceuticals by Flow Injection Analysis Using Brown Mono 1,10–Phenanthroline-Iron (III) Complex as an Oxidant," Med. J . Islamic Acad. Sci. 13 (2), 69–74 (2000).