Introduction

A respiratory illness outbreak in Wuhan, China, led to the discovery of a new infectious disease named Coronavirus Disease-2019 (COVID-19), caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Since then, COVID-19 has advanced into an unprecedented global pandemic that sparked grave public health emergencies. Preventive measures, therefore, remained inevitable to control the spread of infection. Accordingly, public health agencies across the globe recommended hand hygiene as one of the measures to reduce COVID-19 transmissions1.

Hand hygiene has been proven to stop outbreaks, reduce transmission of antimicrobial-resistant organisms, and reduce overall infection rates in healthcare facilities2. Hence, in 2002, the United States (US) Centers for Disease Control and Prevention revised their hand hygiene guideline in healthcare settings by recommending that healthcare personnel use alcohol-based hand rubs alongside traditional handwashing with soap and water for patient care3. The World Health Organisation (WHO) subsequently published a guideline on hand hygiene, which also promotes the use of alcohol-based hand sanitiser (ABHS) among healthcare workers4. ABHSs are recommended for non-visible soiled hands by applying an adequate amount of sanitisers to the palm of one hand, successively rubbing both hands, whilst covering all surfaces of the hands and fingers until they are dry5.

In a study published in the Emerging Infectious Diseases, Swiss and German researchers found that the original and modified ABHS formulations recommended by the WHO effectively kill the SARS-CoV-2 with a virus reduction factor of ≥ 3.8 and ≥ 5.9, respectively. The study also found that ethanol and isopropyl alcohol (IPA) were efficient in inactivating the virus in 30 s at a concentration of > 30% v/v6. The WHO recommends two formulas for ABHS, which differ only in their alcohol constituent, and is at present widely accepted throughout the world. These formulations are Formulation 1 with 80% v/v of ethanol and Formulation 2 with 75% v/v of IPA, in addition to other ingredients such as glycerol, hydrogen peroxide, and sterile/distilled water. No other active or inactive ingredients should be added, which may impact the quality and potency of the product7. Generally, ABHS containing 60 to 95% v/v of alcohol kills microorganisms most effectively, provided that at least 2.4 mL of hand sanitisers were applied for 25 to 30 s1.

The combination of healthcare experts’ recommendations for ABHS and consumers’ fear of contracting the novel influenza A (H1N1) swine flu have contributed to the earliest significant spike in hand sanitiser sales in 20098. During the current COVID-19 pandemic, again, ABHS proved to be a crucial tool to reduce the number of SARS-CoV-2 transmissions via skin-to-skin contact. At the early stage of the COVID-19 outbreak, a sudden surge in demand for hand sanitisation products has led to shortages in their supply. Hand sanitiser sales increased by 470% in the first week of March 2020 compared to the same week the previous year, according to Nielsen market research in the US8. As a result, many regulatory bodies have relaxed their regulations requirements in response to the significant shortages of hand sanitisers8,9.

Recognising the opportunity and at the same time keeping up with consumers’ demands, cosmetic and skincare brands have also diversified their product lines to include ABHS10. While these companies often have the technology and ingredients to manufacture ABHS, there are concerns that these new manufacturers are not aware of their regulatory obligations. Although health regulatory authorities have guided the industry to produce ABHS during a public health emergency, some manufacturers may have resorted to in-house formulations that are not validated and licenced for use11.

The US Food and Drug Administration (FDA) has recently uncovered grave safety concerns with some hand sanitisers during their quality testing in July 2020. These include low levels of active ingredients, including the detection of hazardous amounts of impurities such as methanol. In addition, some of these products were even labelled with false, misleading, or unproven claims. Consequently, more than 150 hand sanitisers have been recalled and are no longer advised for use12. Based on these findings, health regulatory authorities were alerted about substandard or unsafe ABHS products that may become toxic to human health and the environment when used.

In Malaysia, ABHSs are classified as cosmetics or generic products depending on their intended use. Since the COVID-19 pandemic, various manufacturers and companies have filed notification applications with the National Pharmaceutical Regulatory Agency (NPRA) to import, manufacture, or market ABHS as cosmetics. Generally, cosmetic products must be notified, adhere to the requirements set in the Guidelines for Control of Cosmetic Products in Malaysia, and be manufactured by a premise that meets the requirements of Good Manufacturing Practice for cosmetic products or equivalent13. In addition, the NPRA has set a specific requirement for ABHS, stating that the active ingredient must contain at least 60% of alcohol and the manufacturer can only make hygiene and antibacterial claims14. Notified cosmetic products in Malaysia are monitored from time to time, with regulatory penalties issued upon products that did not comply with the requirements set by the NPRA, including cancellation of notification and removal from the market.

Alcoholmeter, hydrometer, or other chemical analyses with equivalent or greater accuracy are among the most common methods used for determining alcohol concentration15,16. During the COVID-19 public health emergency, the National Institute of Standards and Technology (NIST) has supported the industry by developing and evaluating analytical methods for their applicability in identifying and quantifying ethanol and other impurities in 72 hand sanitisers17. These methods include gas chromatography with flame ionisation detection (GC-FID), liquid chromatography with ultraviolet absorbance detection (LC-UV), quantitative nuclear magnetic resonance (q-NMR) spectroscopy, and attenuated total reflectance Fourier-transform infrared (ATR-FTIR) spectroscopy.

The NPRA, as the regulatory body in charge of cosmetic product compliance and quality control in Malaysia, herein reported the development of a gas chromatography-mass spectrometry (GC–MS)-based analytical method to simultaneously identify and quantify ethanol, IPA, and methanol in ABHSs. The method was carefully optimised and validated, ensuring robust and reliable performance for determining alcohols and related impurity in liquid and gel ABHS products. The applicability of this method was demonstrated using 69 samples obtained from the Malaysian market under the NPRA’s post-registration market surveillance programme.

Methods

Materials

All reagents and solvents were of LC grade and purchased from Merck KGaA (Darmstadt, Germany). The purity of ethanol, IPA, methanol, 2-butanol, and acetonitrile were ≥ 99.90%, ≥ 99.90%, ≥ 99.97%, ≥ 99.00%, and ≥ 99.90%, respectively, as stated on the solvents’ specification.

Instrumentation and analytical method

The analysis was carried out using a GC system coupled to a quadrupole MS (GC–MS TQ8040, Shimadzu Corporation, Japan). The compounds were separated using a BP-624 GC capillary column (Part number: 054840, Trajan, Australia) with dimensions of 30 m (length) × 0.25 mm (internal diameter) × 1.4 µm (film thickness). Helium was used as the carrier gas at a flow rate of 1 mL/min. The split injection mode was selected, using a split ratio of 1:100 with an injection volume of 0.2 µL. The GC oven temperature was programmed using an initial temperature of 60 °C with a hold time of 3.00 min, then increased to 90 °C at 30 °C/min, and subsequently to 230 °C at 55 °C/min. The final temperature was held for 3.00 min with a total run time of 9.55 min. The injection and interface temperatures were set at 230 °C.

The MS was operated at 70 eV in Electron Ionisation (EI) mode, while the ion source temperature was set at 200 °C. Target analytes were identified by collecting the full-scan mass spectra at an m/z range of 25 to 200. The identity of these analytes in a sample was confirmed using the NIST 14 mass spectral library. Selected Ion Monitoring (SIM) mode was chosen as the data acquisition method for quantitation. One target ion was selected as a quantifier ion for each analyte, while two reference ions were chosen for identification purposes. Table 1 shows the quantifier and qualifier ions for each target analyte and internal standard (IS). All data were processed using Lab Solution Software (GC–MS Solution version 2.50, Shimadzu Corporation, Japan).

Table 1 The quantifier and qualifier ions for target analytes, including the internal standard.
Full size table

Standard preparation

Standard stock solutions of labelled alcohols (ethanol and IPA), methanol, and 2-butanol (IS) were prepared at concentrations of 5.0% v/v, 0.5% v/v, and 10.0% v/v, respectively, by dissolving their neat standards in acetonitrile. For ethanol and IPA, the calibration solutions were prepared at concentrations of 0.45, 0.55, 0.65, 0.75, 0.85, and 0.95% v/v by diluting the standard stock solution with an appropriate amount of acetonitrile. Calibration solutions for methanol were prepared by aliquoting the standard stock solution in acetonitrile to obtain final concentrations of 0.0115, 0.0175, 0.0275, 0.0375, 0.0450, and 0.0550% v/v. Each ethanol, IPA, and methanol calibration solution was then added with an exact amount of 2-butanol as an IS to yield a final concentration of 0.55% v/v.

Sample preparation

The developed method was applied to 69 ABHS products obtained from the Malaysian market in liquid and gel forms; collected under the NPRA’s post-registration market surveillance programme. These products were selected based on their over-claim advertisements, public complaints, manufacturer’s track record and random market sampling. These samples were initially homogenised before analysis. For liquid-type hand sanitisers, 0.2 mL of a sample was accurately pipetted into a 20 mL volumetric flask. For gel-type hand sanitisers, their specific gravity at 25 °C (weight/mL) was initially determined using a density meter (Mettler Toledo, Greifensee, Switzerland). Based on the specific gravity, an amount equivalent to 0.2 mL of a gel sample was weighed in a 20 mL volumetric flask. Next, an exact amount of IS stock solution was added to obtain a final concentration of 0.55% v/v. The sample was further diluted to volume with acetonitrile and subsequently vortexed for 2 min until well mixed. The sample solution was finally filtered using a 0.45 µm nylon filter into a GC vial for analysis.

Method validation

The method was validated according to the International Council for Harmonisation (ICH) Harmonized Tripartite Guideline: Validation of Analytical Procedures: Text and Methodology Q2(R1)18. The validation parameters for determining the two labelled alcohols (ethanol and IPA) in liquid and gel ABHSs include specificity, linearity and range, accuracy, and precisions. Methanol, in contrast, is regarded as an impurity in hand sanitiser products, and thus, two additional validation parameters, i.e., limit of detection (LOD) and limit of quantitation (LOQ), were performed. Blank matrices of liquid and gel ABHSs, free from any analyte of interests, were used for method optimisation and validation.

The specificity was evaluated by establishing the ability of the method to discriminate each target analyte in the presence of various interferences originating from other analytes, IS, diluent, and sample matrices. The linearity was established using a six-point calibration curve, analysed in triplicate. The calibration curve was constructed by plotting the peak area ratio of each target analyte versus the IS against their corresponding concentration in % v/v. The linearity range of each target analyte was selected according to their respective specification or limit. The accuracy was established using a recovery study of samples spiked at three quality control levels (low, medium, high) covering the range of the analytical method. The precisions were determined by evaluating the repeatability and intermediate precision. The repeatability was ascertained by analysing six different liquid and gel samples spiked at a single concentration (0.75% v/v ethanol and IPA, and 0.0375% v/v methanol). The intermediate precision was determined using a similar procedure to that of repeatability, albeit the analysis was performed by a second analyst and on a different day. The LOD and LOQ of methanol were initially calculated based on the signal-to-noise (S/N) ratio approach. Based on these estimates, the LOD was then determined experimentally at the lowest concentration of methanol that can be reliably detected, with at least a 3:1 S/N ratio. At the same time, the LOQ was set at a concentration where methanol can be quantified with acceptable accuracy and precisions, giving an S/N ratio of no less than 10.

Results

Method development and optimisation

The simultaneous separation of multiple analytes can be a challenging task. Therefore, careful consideration must be given to each target analyte to achieve an excellent chromatographic separation for reliable identification and quantitation. A non-polar (5%-phenyl)-methylpolysiloxane, HP-5 ms (30 m × 0.25 mm × 0.25 µm) column was initially assessed during the early stage of method development, with an initial column temperature of 40 °C. Although a relatively good separation (resolution factor of 1.933 to 4.993) was observed between each target analyte, the peak shape was poor, with tailing factors in the range of 0.96 to 2.359. Moreover, the target analytes’ capacity factors (K’) were observed between 0.059 and 0.320, indicating poorly retained alcohols in the HP-5 ms column. Therefore, another column, i.e., an intermediate polar (6%-cyanopropylphenyl)-dimethylpolysiloxane, BP-624 (30 m × 0.25 mm × 1.4 µm), was evaluated in a review of the first column. Contrarily to the HP-5 ms column, the BP-624 column's thicker stationary phase film of 1.4 µm necessitated a higher initial column temperature of 60 °C. The BP-624 column significantly improved the peak shape of all target analytes. The peaks of all three alcohols were sharp and symmetrical with tailing factors between 1.032 and 1.206. Furthermore, target analytes retention in the column improved significantly, with K’ ranging from 0.219 to 1.107. The BP-624 column was chosen as it provides an optimal chromatographic separation.

The injection volume was evaluated at 0.2, 0.3, and 0.5 µL. The detector response for ethanol and IPA was observed to be saturated, notably at concentrations higher than 0.75% v/v using injection volumes of 0.3 and 0.5 µL. Conversely, the detector saturation for ethanol and IPA was not visible at the injection volume of 0.2 µL. The detector response for methanol, however, was not significantly reduced, with a good response recorded at the lowest injection volume. Based on these findings, an optimum injection volume of 0.2 µL was selected for the final method.

Method validation

Specificity

All target analytes were chromatographically resolved from one another; identified by their specific retention time, as shown in Fig. 1A, and selectively resolved using the MS SIM detection with unique quantifier and qualifier ions, presented in Table 1. No interferences were found at the peaks of methanol, ethanol, and IPA, as evidenced in the GC–MS total ion chromatograms of an acetonitrile blank solution, an IS of 2-butanol, and a sample solution spiked with IS. Furthermore, the retention times of each standard in the unspiked and spiked samples correspond to their respective peaks in the chromatograms of the mixed standard solution, as shown in Figs. 1B and 1C, respectively. In addition, the reference ions ratio of an unspiked sample and a spiked sample was within the maximum-permitted tolerance of ± 10%, set according to the reference ions ratio of a mixed standard solution.

Figure 1
figure 1

A representative GC-MS total ion chromatograms (TIC) with the addition of 2-butanol as an internal standard. TIC of mixed standard solution (A), unspiked sample solution (B) and spiked sample solution (C), demonstrating the specificity of the method.

Full size image

Linearity and range

A linear regression model was applied to the standard curves, and the linearity was determined based on the coefficient of determination (r2). The linearity model was further verified using the lack-of-fit test to confirm that the selected regression analysis is accurate at a 95% confidence interval. The r2 obtained using six standard concentration levels for ethanol, IPA, and methanol were 0.9994, 0.9995, and 0.9996, respectively. Ethanol and IPA standards were prepared using a range of 4.5 to 9.5% v/v, which is equivalent to 45 to 95% v/v in real samples. Methanol, in contrast, was prepared within a range of 0.01128 to 0.055% v/v, corresponding to 1.128 to 5.5% v/v in real samples. The linear regression model demonstrated excellent linearity between the peak area ratio of the target analytes with respect to 2-butanol (IS) versus their concentration. In addition, all y-intercept values passed zero, established via the student t-test.

Accuracy

Results obtained from the spike recovery studies are presented in Table 2. For liquid hand sanitiser samples, the mean recoveries of ethanol, IPA, and methanol were 100.31%, 101.09%, and 99.98%, respectively. While for gel hand sanitiser samples, the mean recoveries were 99.26% for ethanol, 98.99% for IPA, and 99.79% for methanol.

Table 2 The accuracy of ethanol, isopropyl alcohol, and methanol at low, medium, and high-quality control levels for liquid and gel hand sanitisers (established via spike recovery studies).
Full size table

Repeatability

The repeatability data are shown in Table 3. The relative standard deviation (RSD) percentages of ethanol, IPA, and methanol spiked in liquid hand sanitiser samples were 1.50%, 0.82%, and 0.83%, respectively. The RSD percentages for gel hand sanitiser samples were 0.57% for ethanol, 0.47% for IPA, and 0.57% for methanol. Overall, each target analyte demonstrated acceptable repeatability with RSD percentages of < 5%.

Table 3 The repeatability of ethanol, isopropyl alcohol, and methanol for liquid and gel hand sanitisers.
Full size table

Intermediate precision

The intermediate precision results are shown in Table 4. The RSD percentages for ethanol, IPA, and methanol spiked in liquid hand sanitiser samples were calculated at 1.11%, 3.04%, and 0.81%, respectively. The gel hand sanitiser samples yielded RSD percentages of 0.95% for ethanol, 0.91% for IPA, and 1.56% for methanol. All target analytes demonstrated acceptable intermediate precisions with RSD percentages of < 5%.

Table 4 The intermediate precision of ethanol, isopropyl alcohol, and methanol for liquid and gel hand sanitisers.
Full size table

Limit of detection and limit of quantitation

The S/N for methanol via the predetermined LOD and LOQ levels were 517.97 and 1447.44, respectively. Upon confirmation, the RSD percentages of the peak area ratio for LOD (n = 3) was 1.003%, while LOQ (n = 10) was 2.053%. Accordingly, the LOD for methanol was established experimentally at 0.004% v/v (equivalent to 0.4% v/v in a sample); whereas the LOQ was verified at 0.0115% v/v or 1.15% v/v of methanol in a sample.

Discussion

GC, in general, provides information on the retention time, verifying the identity of alcohols and, simultaneously, quantifying their concentration, making it useful for the quality control of ABHS products. In contrast to the recommended method in the US Pharmacopeia (USP) General Chapter 611, Alcohol Determination19, a GC coupled to an MS detector was utilised in this study instead of an FID.

An FID is recommended for alcohol determination because it is easy to use, generating a consistent and adequate response within the specified concentration range, even after dilutions during the sample preparation process. Hence, most analytical instrument companies have published GC-FID application notes as a tool for hand sanitiser analysis20,21,22. Nonetheless, an MS is more sensitive and capable of detecting trace amounts of impurities23. Thus, a GC–MS-based method was selected to evaluate ABHS products formulated with ethanol or IPA as the labelled active ingredient.

The method described in this study was developed following the updated US FDA guideline and their subsequent published method12, which requires the active ingredient of either ethanol or IPA to be tested for the presence of methanol (as an impurity) if it is obtained from a third-party source. This reference method, however, was established to screen potentially harmful impurities with interim limits, which differs from the guideline published by the NPRA. For instance, under the US FDA temporary policies, ethanol used to manufacture hand sanitiser products containing less than 630 ppm of methanol is not considered contaminated; and is not subjected to adulteration charges under the Food, Drug, and Cosmetic Act. Conversely, the NPRA focused on detecting methanol as an impurity because it is listed in the NPRA Cosmetic Guideline ‘Annex III-Part 1-List of substances which cosmetic products must not contain except subject to restrictions and conditions laid down’24. This annexe outlines that methanol content in cosmetic products should not exceed 5%, calculated as a percentage of either ethanol or IPA. If any manufacturer fails to follow this limit, the ABHS notification will be revoked, and their product will be withdrawn from the market.

Numerous global reports have revealed that ABHSs may contain undeclared methanol25,26. Therefore, NPRA has been actively analysing ABHSs for compliance with the regulations and legislation. Methanol is toxic when inhaled, taken orally, or applied to the skin; and should never be incorporated into any hand hygiene product. Methanol poisoning may lead to central nervous system depression and severe metabolic acidosis, causing blindness, confusion, and eventual death27. Generally, methanol poisoning is rare; typically occurs due to accidental ingestion or suicidal intent. Unfortunately, during the COVID-19 pandemic, there was a proliferation of fatal methanol poisoning resulting from the inappropriate use of hand sanitisers illegally containing methanol28,29.

Between August 2020 and November 2021, 69 distinct ABHSs were analysed by the NPRA, with the results presented in Table 5. A discrepancy between the label claim and the analysis result regarding the type of alcohol was evident in one sample. Sample SL18 was labelled with 80.0% ethanol but was instead detected to contain 66.4% IPA. According to the US FDA, IPA should be in the range of 70.0 to 91.3% v/v in any hand sanitiser product9. Ethanol, on the contrary, should be between 60.0 and 95.0% v/v9. This finding leads to another grave concern regarding the potency of alcohol in the ABHSs. Of the 69 samples tested, 14 were less potent as they contained insufficient amounts of the active ingredient. For instance, five samples were quantified below 60% for ethanol, while nine were quantified below 70% for IPA. Hand sanitisers with a lower alcohol percentage may not work as well for many germs and may merely inhibit the growth of germs rather than killing them altogether.

Table 5 The identification and quantitation of labelled alcohols (ethanol and isopropyl alcohol) and their impurity (methanol) in liquid and gel hand sanitiser samples.
Full size table

Alarmingly, four samples, viz. samples SL3, SL5, SG14, and SG17, may pose a significant threat to consumers as they contained a high amount of methanol ranging from 5.3 to 19.4% with respect to the active alcohol percentage. Therefore, consumers who have been exposed to ABHSs containing methanol should discontinue using them immediately and seek medical attention if they experience any concerning symptoms. NPRA, at the same time, will cancel the products’ notification and remove these products from the market to protect consumers’ health and well-being.

Conclusion

A GC–MS-based method was developed, optimised, and validated for the quality, safety, and efficacy assessment of ABHS products. The method can be utilised to identify and quantify ethanol or IPA as the active ingredient in ABHS, with simultaneous determination of methanol, an impurity that should never be present in any hand sanitiser products. The method's applicability was demonstrated using 69 ABHS samples obtained in liquid and gel forms from the Malaysian market. The strategies discussed herein would be beneficial to protect the public against the potential harm of substandard or unsafe ABHS products. Therefore, it is critical for a regulatory body like the NPRA to ensure that ABHSs contain the required percentage of alcohol and are free of any potentially hazardous impurities like methanol.