Development and Validation of an HPLC Method for Simultaneous Assay of MCI and MI in Shampoos Containing Plant Extracts

A simple, easy-to-implement HPLC method was developed and validated for simultaneous determination of two isothiazolinone preservatives, methylchloroisothiazolinone (MCI) and methylisothiazolinone (MI), in hair care shampoo containing plant extracts. In this method, shampoo samples were first dissolved in isopropyl myristate and then MCI and MI were extracted from isopropyl myristate layer by a mixture of methanol and 0.02 M phosphate buffer solution pH 3.0 (30: 70, v/v) and analyzed on an analytical biphenyl column maintained at 25°C with a mixture of methanol and water (10: 90, v/v) in isocratic elution mode as mobile phase. Total flow rate of mobile phase was maintained at 1.0 mL per minute. The UV detection was performed at 274 nm. Injection volume was 50 μl. The method was fully validated in terms of specificity, linearity, precision, accuracy, and robustness according to requirements of AOAC International and was proved as reliable and suitable for the intended application.


Introduction
e use of preservative is necessary for many types of cosmetics and toiletries because certain components in these products, such as plant extracts, can be favorable for the development of microorganism. Isothiazolinone-type biocides are effective preservatives with antimicrobial activity against a broad spectrum of fungi and bacteria [1]. eir effectiveness at low concentrations made these biocides a common choice for preservatives in cosmetic products [2]. e most widely used biocides of this group are methylchloroisothiazolinone (MCI) and methylisothiazolinone (MI), in which MI can be used alone or MCI and MI can be used together in 3 :1 combined mixture. However, MCI and MI have been known to cause contact dermatitis [2][3][4][5]. erefore, the use of the mixture MCI-MI was restricted in the European Union at 15 ppm (0.015 mg/g) [6] or recommended at 7.5 ppm for leave-on products and 15 ppm for rinse-off products [7]. e use of MI independently without the presence of MCI was claimed as safe at not more than 100 ppm in European Union [6]. However, the use of isotriazolinone biocides in cosmetic products has became more and more restricted in the last few years. In 2017, ASEAN countries, including Vietnam, only accepted MCI-MI (3 :1) mixture and MI for use in rinse-off products at levels not more than 15 ppm [8]. In Vietnam, the consumption of cosmetics has been rapidly increased in recent years, with commercial products coming from many sources, including highly dubious ones. erefore, it is necessary to have reliable and easy-to-implement analytical methods for controlling the actual levels of bioactive components with potential risk for human health in cosmetics in order to assure that regulation requirements are respected. In case of components with low concentration in cosmetics like MCI and MI, the method must be able to minimize the interference from sample matrix, which can be very complex, and to detect specifically the analytes with suitable limits of detection and quantification.
e interference from product matrices can be very challenging, such as those of shampoos containing plant extracts. ese products can contain the extracts of many plants, such as Panax ginseng, Ginkgo biloba, Fallopia multiflora, Gleditsia australis, Eleusine indica, Oroxylum indicum, Ageratum conyzoides, Morus alba, Agastache rugosa, Cymbopogon citratus, and Ocimum gratissimum, just to cite the most frequently declared ones on the shampoo labels (see Table 1 for some examples). ese plant extracts can be used separately or combined, and sometimes a shampoo can contain extracts of 7 different plants in its composition. To extract MCI and MI from cosmetics and toiletries, different approaches have been employed, including solid-phase extraction [9,10], matrix solid-phase dispersion [11], or direct dissolving from sample matrix with different solvents [12][13][14][15]. However, within the limit of our bibliographic research, we did not find any sample preparation process dealing with the matrix of shampoo containing plant extracts. e analysis of MCI and MI has been executed with gas chromatography coupled with mass spectrometric detection [16], but the majority of published works regarding the quantitative determination of MCI and MI in different types of samples have been carried out by using reverse-phase liquid chromatography on C18 column [10][11][12][13][14][15] or C30 column [9] with UV-Vis detection [9,12] or mass spectrometric detection [10,11,[13][14][15].
In this study, an HPLC method using phenyl column combined with sample treatment using liquid-liquid extraction was developed and validated for simultaneous assay of MCI and MI in shampoos containing plant extracts.

Instrumentation.
Shimadzu LC-20AT HPLC system (Shimadzu, Kyoto, Japan) was used for method development and validation.

Chemicals and Reagents.
Reference substances of MI (purity: 98.2%) and MCI (purity: 96.0%) were purchased from Sigma Aldrich Singapore (Singapore). Methanol HPLC grade and other PA grade chemicals (potassium dihydrogenphosphate, orthophosphoric acid, and isopropyl myristate) were purchased from Merck Vietnam (Ho Chi Minh City, Vietnam). e 0.02 M phosphate buffer solution pH 3.0 was prepared by dissolving 2.72 g of potassium dihydrogenphosphate in 900 mL of water, adjusting the pH to 3.0 ± 0.1 by orthophosphoric acid if necessary and diluting with water to make 1000 mL.
For method development, 3 different shampoos containing plant extracts were used to optimize the conditions of sample preparation and chromatographic separation (see Table 1).
Blank sample for method validation containing the same components as those of ai Duong 7 shampoo without MI and MCI and was kindly provided as a gift by Sao ai Duong company (Vietnam).

Analytical Method.
To obtain the suitable method, the conditions for sample preparation and chromatographic analysis were optimized in method development and optimization process, which is discussed in detail in 3.1. From the outcomes of this process, the final conditions of analytical method were fixed as in 2.3.1 and 2.3.2.

Chromatographic Conditions.
Mobile phase was mixture of methanol and water (10: 90, v: v). e flow rate of mobile phase was maintained at 1.0 mL/min. e analysis was carried out on an Shimadzu LC-20AT series HPLC system equipped with a photo diode array detector set at 274 nm for recording chromatograms. e chromatographic separation was conducted on a Apollo Phenyl column (250 × 4.6 mm, 5 μm) maintained at 25°C. e injection volume was 50 μl.

Sample Preparation
(1) Standard Solutions. Stock standard solutions of MCI (1.0 mg/mL) and MI (1.0 mg/mL) were prepared by dissolving an accurately weighed quantity of corresponding reference standard in methanol. Working mix standard solutions were prepared by accurately diluting stock standard solutions to intended concentration using a mixture of methanol-0.02 M phosphate buffer solution pH 3.0 (30: 70, v: v). Standard solutions were filtered through 0.45 μm membrane filter before being used for chromatographic analysis.
(2) Sample Solution. About 1.0 g of shampoo samples, accurately weighed into a 50-mL separation funnel, was dispersed in 10 mL of isopropyl myristate and extracted 2 times, each time by shaking for 10 minutes with 8 mL mixture of methanol-0.02 M phosphate buffer solution pH 3.0 (30: 70, v: v). e lower layer was then collected into a 20-mL volumetric flask and the mixture of methanol-0.02 M phosphate buffer solution pH 3.0 (30: 70, v: v) was added into the flask to make 20 mL.
is solution was transferred into a centrifuge tube and left at 10°C in 30 minutes before centrifuging at 5000 rpm for 10 minutes at 10°C. A portion of supernatant was immediately filtered through 0.45 μm filter membrane for chromatographic analysis.

Method Validation.
e conditions of final analytical method as being described in 2.3 were used for the method validation. e assessment of validation results was based on performance requirements of AOAC International [17] and those proposed in other published papers [18][19][20][21][22][23]. e method was validated in terms of specificity, linearity, sensitivity, accuracy, precision, robustness, and the stability of test solutions.

Specificity.
Specificity is the ability of the analytical method to distinguish between the analyte(s) and the other components in the sample matrix [18,19]. In case of this HPLC method, it is assured by complete separation of MI and MCI from each other and from other peaks originated from sample matrix. Specificity evaluation was carried out by injecting separately 50 μl of standard, sample, spiked sample, and blank into the chromatographic system. ree injections from each concentration were analyzed under the same conditions. Linear regression analysis was used to evaluate the linearity of the calibration curve by using least square linear regression method, and the significance of linear regression was confirmed by one-way ANOVA test if P (Sig) < 0.05 [18].

Sensitivity.
e limit of detection (LOD) and limit of quantitation (LOQ) of MCI and MI were determined by analyzing different solutions of MCI and MI and measuring the signal-to-noise ratio for each analyte. e limit of detection (LOD) is the concentration giving a signal-to-noise ratio not less than 3 : 1, and the limit of quantitation (LOQ) is the concentration giving a signal-to-noise ratio not less than 10 : 1 with RSD of less than 10% with triplicate analysis [20,21].

Accuracy.
e accuracy of the method was determined by recovery studies for MCI and MI from placebo matrix [17][18][19]. Exact amounts of reference substances of MCI and MI were mixed with blank matrix in such a way that the spiked samples, after preparation process, yielded solutions containing each analyte at three concentration levels within the linear range: at lowest concentration, at middle concentration, and at 80% of highest concentration of the calibration curve, i.e., about 0.084, 0.169, and 0.203 μg/ mL with MI and about 0.255, 0.510, and 0.612 μg/mL with MCI. At each concentration level, nine samples were prepared and analyzed. e percentage recovery of added MCI and MI and the RSD were calculated for each replicate samples.

Precision.
e proposed method was validated in terms of system precision and method precision according to current guidelines and published papers [17,18,22]. e system precision was determined by six measurements of mixed standard solution containing about 0.169 μg/ mL of MI and 0.510 μg/mL of MCI on the same day [18,22]. e method precision includes repeatability and intermediate precision [17,18,22]. ey were determined by estimating the dispersion of assay results obtained with recovery studies in 2.4.4 at each spiked level the same day and on two different days, respectively.

Robustness.
e robustness of the method was verified by investigating the effects caused by deliberate minor changes in experimental conditions to analyze the results [18,22,23]. In this study, following changes were applied: (i) Flow rate: ±0.1 mL/min (ii) Percentage of methanol in mobile phase: ±1% At each condition, a mixed standard solution containing about 0.169 μg/mL of MI and 0.510 μg/mL of MCI and three sample solutions of a shampoo product containing MI and MCI as preservatives were prepared and injected into chromatography system. e robustness of method was evaluated from the RSD of peak area for each analyte after three consecutive injections of standard solution and the RSD of the content of MCI and MI determined from sample solutions.

Stability of Analytical Solution.
e stability of analytical solutions was determined by analyzing the standard

Method Development and
Optimization. e matrix of shampoo with plant extracts was very complex due to the presence of many components coming from the various plants added into the product besides the normal composition of an usual shampoo. Direct dissolving of MI and MCI from this matrix has been carried out using methanol and mixtures of methanol-water, methanol-0.02 M phosphate buffer solution pH 3.0 at different percentages, and the obtained liquids were analyzed on C18 and phenyl stationary phases. With all above-mentioned solvent and solvent mixtures, analysis results revealed significant codissolving of other components from sample matrix. Due to low concentration of MI and MCI in shampoo and similar chromatographic behavior of many codissolved components, it was impossible to separate completely MI and MCI with C18 column. With phenyl column, the elution program permitting complete separation of MI and MCI from other matrix peaks was too long (more than 60 minutes per injection). However, HPLC analysis results showed that codissolving effect was less significant by using mixture methanol-0.02 M phosphate buffer solution pH 3.0 (30: 70, v: v) than by using methanol, methanol-water, and other ratio of methanol-0.02 M phosphate buffer solution pH 3.0 (preliminary results not shown). To reduce further the codissolving effect, the shampoo was first dissolved in isopropyl myristate before liquid-liquid extraction with mixture methanol-0.02 M phosphate buffer solution pH 3.0 (30: 70, v: v). By this process, MI and MCI were extracted effectively into the methanol-buffer phase due to their high polarity and the acidic pH of the methanol-buffer phase. e liquid obtained after liquid-liquid extraction step using conditions in 2.4.2 contained less coextracted components than the liquid obtained from direct dissolving (see Figure 1(b)). e cold centrifuge step after liquid-liquid extraction permitted further cleansing of the extracted liquid by reducing the solubility of certain coextracted components and eliminating them by precipitation, while the solubility of MI and MCI was not affected. is amelioration in sample preparation step permitted simplifying chromatographic conditions and faster analysis on phenyl column (only 18 minutes per injection, see Figure 2 for typical chromatograms and see 2.3 for chromatographic conditions). On C18 column, it was also possible to separate completely MI and MCI, but analysis time was longer than on phenyl column (more than 30 minutes per injection, (see Figure 1(a))). erefore, liquid-liquid extraction combined with chromatographic analysis on biphenyl column has been selected for the final method. e UV-Vis absorption spectra of MI and MCI showed a maximal at wavelength about 274 nm; therefore, this wavelength was selected to record chromatograms (see Figures 2(g) and 2(h)).  Figures 2(a)-2(d). In selected chromatographic conditions, MI and MCI each were eluted in one completely resolved peak. e peak of MI was eluted before the peak of MCI. It can be observed from the peak purity analysis (Figures 2(e) and 2(f )) that there are no coeluting peaks at the retention times of MCI and MI to interfere with peaks of analytes. is result indicated that the peak of the analyte was pure and this confirmed the specificity of the method.

Linearity and Range.
Analytical method linearity is defined as the ability of the method to obtain test results that are directly proportional to the analyte concentration, within a specific range. e mean peak area obtained from the chromatograms was plotted against corresponding concentrations to obtain the calibration graph. e results of linearity study (Figure 3) gave linear relationship over the concentration range of 0.084-0.253 μg/mL for MI and of 0.255-0.765 μg/mL for MCI. From the regression analysis, the linear equation was obtained: y � 241158x − 592.6 for MI and y � 138508x + 32.83 for MCI, and the coefficient of determination R-square was 0.998 for both analytes, indicating a linear relationship between the concentration of analyte and area under the peak. ANOVA analysis for both analytes (as shown in (Tables 2 and 3)). also proved that the regression model statistically significantly predicts the outcome variable (P < 0.05) [18].

Limit of Detection (LOD) and Limit of Quantification (LOQ).
e limit of detection (LOD) is the lowest amount of analyte in a sample that can be detected, but not necessarily quantitated, while the limit of quantification (LOQ) is the lowest amount of analyte in a sample that can be quantitatively determined with suitable precision [18,20,21]. For MI, the concentration of injected solution at LOD and LOQ was 0.030 μg/mL and 0.084 μg/mL, equivalent to the content of MI in shampoo of 0.60 ppm and 1.68 ppm, respectively. For MCI, the concentration of injected solution at LOD and LOQ was 0.090 mg/mL and 0.255 mg/mL, equivalent to the content of MCI in shampoo of 1.80 ppm and 5.10 ppm, respectively. According to the current legal requirements in Vietnam [8], with the limit of mixture MCI-MI (3: 1, w: w) being set at 15 ppm, the acceptable content of MI and MCI in rinse-off cosmetics must be not more than 3.75 ppm and 11.25 ppm, respectively. With LOQ of both MI and MCI lower than the current legal acceptable limits, this method has suitable sensitivity for control of MI and MCI in shampoo for regulatory purposes.

Accuracy.
e accuracy of an analytical method expresses the closeness of results obtained by that method to the true value. In this study, the recovery rate fell within the range from 86.5% to 101.8% for the two analytes. At each level of concentration where recovery study was done, the RSD values varied from 2.5% to 5.2%, as shown in Table 4. ese results were within the accepted limit for recovery each concentration level according to AOAC International [17].

Precision.
e precision of the method is defined as "the closeness of agreement between a series of measurements  obtained from multiple sampling of the same homogeneous sample under the prescribed conditions," [20] and it is normally expressed as the relative standard deviation. In terms of system precision, the RSD of retention time, peak area, and the performance of chromatographic system, represented by the number of theoretical plate and tailing factor, were all less than 2.0%, and the number of theoretical plate was higher than 1000 for all analyte peaks, as shown in Table 5. In terms of method precision, the RSD of assay results for MCI and MI in the evaluation of repeatability and   intermediate precision were all within the accepted limit for precision each concentration level according to AOAC International [17], as shown in Table 4. erefore, the results of both system and method precision showed that the method is precise within the acceptable limits for intended application.

Robustness.
e analytical method robustness was tested by evaluating the influence of minor modifications in HPLC conditions on system suitability parameters of the proposed method, as mentioned in Section 2.4.6.
In the robustness test, a minor change of method conditions, such as the composition and flow rate of the mobile phase, caused variations in analytical results within acceptable limit, i.e., RSD less than 2.0% (Table 6). ese results demonstrated that the method was robust in case of minor variations in experimental conditions. In all modifications, good separation was achieved between peak of MI and peak of MCI, as well as between peaks of these substances and other peaks on chromatograms, and the RSD values of peak area were obtained from repeated injections of standard solution. e RSD values of MI content and MCI content determined from ai Duong 7 shampoo (Table 1) containing approximately 3 ppm of MI and 9 ppm of MCI were all less than 2.0% and lower than limit proposed by AOAC International for this range of concentration [17]. Furthermore, the minor changes applied in this robustness test produced no significant difference in content of MCI   and MI found in sample, as one-way ANOVA analysis found F < F inscrit for both analytes (as shown in Table 7).

Solution
Stability. e percentage of recovery was within the range of 98.0% to 102.0% and RSD was not more than 2.0%, indicating a good stability of the sample and standard solutions for 24 hr at 25°C in autosampler, as shown in Table 8. ese results proved that both analytes were stable in sample and standard solutions prepared as described in 2.4, and the preparation procedure for sample and standard solution was suitable for intended application of the method.

Conclusion
In this study, a simple, accurate, precise, and robust HPLC method has been developed for simultaneous assay of MCI and MI in shampoos containing plant extracts. In the extent of our literature research, this was the first method using chromatographic separation on phenyl column and liquidliquid extraction for analysis of MI, MCI in cosmetics. e method was validated according to the requirements of AOAC International on analytical method performance, proved to be suitable for the intended application, and able to provide accurate and precise quantitative results under minor variation of chromatographic conditions.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare no conflicts of interest in publication of this study.