1. Introduction
Wax coating is applied to fruits and vegetables before shipping them over long distances. This reduces the quality loss of fruits and vegetables between harvest and consumption by retaining moisture and preventing weight loss. The coating is also applied to fruits at set periods after flowering to prevent damage to fruit during ripening. Fruit wax coating protects against sun damage in the summer by preserving the color of fruit. However, the wax coating is difficult to remove by typical washing of fruits and vegetables; thus, wax coatings can be absorbed in the body without degradation [
1]. Active compounds such as ascorbic acid, malic acid, calcium chloride, calcium lactate, citric acid, and glutathione are often incorporated in the formulations of edible wax coating matrix (e.g., apple puree/pectin alginate, whey protein concentrates, alginate/apple puree, whey protein concentrates, and beeswax, etc.) of apple fruits to improve shelf life [
2]. Morpholine is a colorless secondary amine ether that has been used in the edible coating matrix [
3]. Morpholine has been used as a salt of fatty acid form (e.g., stearate salt) as an emulsifier to wax coating [
4,
5,
6]. Morpholine has been added to the wax coating applied thinly and evenly to the fruit and vegetable surface [
7]. However, there are safety concerns regarding the use of morpholine. Animals exposed to morpholine showed liver and kidney damage [
8]. During digestion in the gut, morpholine can undergo nitrosation with excess nitrites from the diet to form
N-nitroso-morpholine (NMOR), a genotoxic carcinogen in rodents [
9]. A previous study showed that when morpholine was added to human saliva, NMOR was formed [
10]. NMOR may be formed in the human gut when morpholine-treated fruits and vegetables are consumed, especially when they are consumed with wax-coated peel. The safety of morpholine has been extensively examined in many countries. Therefore, it has not been approved for use in the European Union [
11]. Health Canada set morpholine acceptable daily intake as 0.48 mg kg
−1 of body weight (bw) day
−1 based on no observed adverse effect level as 96 mg kg
−1 of bw day
−1 in a chronic oral toxicity study [
9]. Morpholine is approved for use in the United States and was included on the Environmental Protection Agency (EPA) master list and as an EPA registered pesticide in 1996 [
8]. Canada and Australia have permitted the use of morpholine as fruit coating additive [
9]. Furthermore, morpholine use for coating fruit was permitted in Chile and South Africa [
3].
Recently, morpholine-free waxes have been introduced to avoid health concerns; however, the residues are present in package lines because morpholine is also used for various purposes such as fungicidal coating of paper. In China, Korea, and Japan, morpholine salts of fatty acids have been approved for coating the surface of fruit and vegetable commodities at Good Manufacturing Practice maximum levels [
12,
13,
14].
In Korea, apple production increased from 474,712 tons to 545,349 tons from 2014 to 2017 [
15]. Citrus production was 597,294 tons in 2017 [
15]. Morpholine has been used as a coating component of apple and citrus [
16]. Therefore, risk assessment of morpholine is necessary. The development of a method of analyzing for morpholine in apple and citrus matrix should also be essential before conducting a risk assessment of morpholine.
Methods for analyzing morpholine fatty acid salts are complex because it is linked to fatty acids such as stearate. The official standard analysis method of morpholine salts in fatty acids in Japan detects fatty acids rather than morpholine [
13]. However, fatty acids are also present in fruit peel, so even if fatty acids are detected, they may be from the intrinsic fruit peel, not from morpholine fatty acid salts. Recent studies have used morpholine as a target analyte [
17,
18]. To analyze morpholine in fruit and vegetable matrix, various sample preparation methods were used. For example, the cleanup method (i.e., dispersive micro solid phase extraction) was included in a previous study [
17,
19,
20,
21]. In the previous study, dispersive micro solid phase extraction was conducted using PCX (polymer cation exchange) powder that can adsorb alkaline chemical compounds [
17]. Furthermore, a dual solid phase extraction cartridge system was utilized for morpholine analysis in pineapple [
22]. However, the system was time-consuming and costly [
22]. In another previous study, no cleanup method was included [
18].
In previous studies, gas chromatography coupled with a thermal conductivity detector and liquid chromatography coupled with a thermal energy analyzer were used to analyze morpholine in various matrices [
23,
24]. The thermal energy analyzer detects nitrosamines on the basis of chemiluminescence produced by the decay of the NO
2 group when it is electronically excited [
25,
26]. However, the thermal conductivity detector and thermal energy analyzer have higher limits of detection than other detectors such as the flame ionization detector or mass spectrometry [
27]. In a recent study of morpholine analysis by ultrahigh-performance liquid chromatography–high-resolution mass spectrometry (UHPLC-HRMS), the limit of detection of morpholine was reported as 2 µg/kg [
17]. The method used whole fruits rather than separately analyzing the peel and pulp of fruit samples [
17,
18]. A previous study developed a gas chromatography-mass spectrometry (GC-MS) method for morpholine analysis in apple juice and whole apples [
28] but was not used to examine the fruit peel itself. Because of the high lipid contents in fruit peel, lipids may lower extraction efficiency. Particularly, citrus fruit peels have a lipid content of 4.4% [
29]. Moreover, previously reported methods for analyzing morpholine often did not use internal standards [
18]; food samples often have a matrix effect, and thus more accurate analysis can be achieved when internal standards are used, particularly isotope-labeled standards which have very similar chemical structures to the target analyte.
There is little information on morpholine content in fruit commodities. Recently, Chen et al. reported morpholine residues of 80.5–598.7 μg/kg for citrus and 43.4–328.2 μg/kg for apples, respectively [
17]. However, in this method, the peel was not separated from the pulp. Thus, the origin of the morpholine residues was unclear.
To address this issue, we developed a method for analyzing morpholine in the peel and pulp of fruits using GC-MS for the first time. As an internal standard, d8-morpholine was used. Additionally, the method was optimized by involving a lipid removal step and a different pH during the derivatization step. The validation results (i.e., method detection level (MDL), method quantification level (MQL), linearity, accuracy, precision, cross lab validation, measurement uncertainty, etc.) are presented. Additionally, morpholine monitoring was performed on 30 apples and citrus samples purchased from local markets in three countries, Korea, China, and the U.S., after separating the peel and pulp of the fruits. Vegetable samples (cucumber, squash, and paprika) were also analyzed.
2. Materials and Methods
2.1. Chemicals and Reagents
Morpholine standard (99.9% purity) and d8-morpholine (98.0% purity) were purchased from Sigma-Aldrich (St. Louis, MO, USA) and C/D/N Isotope, Inc. (Pointe-Claire, QC, Canada), respectively. Sodium nitrite, n-hexane, dichloromethane, HPLC-grade methanol, and fatty acid methyl ester standards were purchased from Sigma-Aldrich. Hydrochloric acid was purchased from Junsei Chemical Co. (Tokyo, Japan). Ultrapure water was prepared using a Milli-Q water system (Millipore, Billerica, MA, USA).
2.2. Sample Collection
Seventeen apple samples, 7 orange samples, 2 mandarin samples, 1 cucumber, 1 squash, and 1 paprika sample were purchased from local markets from Anseong in South Korea, Beijing in China, and Boston in the U.S. The oranges purchased from Korea were produced in the U.S.A. or Australia and imported to Korea. Other fruits and vegetables purchased from Korea were produced in Korea.
More than 50 fruits or vegetables per sample were purchased to prepare a composite sample. After purchase, the (unwashed) apple and citrus fruits were peeled and separated into peel and pulp. After peeling, each fruit was sliced into six pieces and two pieces were used for making a composite sample for pulp. For cucumber, squash, and paprika, whole vegetables were used for analysis. They were diced and one third of the diced pieces were used to make a composite vegetable sample. The composite samples were lyophilized and stored at −80 °C until analysis within 1 month. The freeze-dried samples were ground before the analysis.
2.3. Optimization of Sample Preparation Method for Morpholine Analysis in Fruit Peel and Pulp
A previously described sample preparation method for morpholine was modified by adding a lipid removal step and changing the pH during the derivatization step [
28]. Sequential extraction was performed. The first step employed the lipid removal method and the second step involved a derivatization step to
N-nitroso-morpholine. As described in the Introduction, Cao et al. prepared samples of apple juice rather than of fruits [
28]. When the method of Cao et al. was used for fruit samples, particularly fruit peels, the final extract solution was unclear, possibly because of the presence of lipids. Therefore, after spiking morpholine standard, lipids were removed from the fruit samples as follows. First, 18 mL of nano-pure water was added to 2.0 g of freeze-dried fruit (apple peel, apple pulp, citrus peel, or citrus pulp) powder in a 15 mL tube. Then, morpholine and d8-morpholine were added. The final isotope labelled internal standard concentration was 100 μg/kg fruit dry weight (DW). The tube was vortexed for 15 min, ultrasonicated for 15 min, and centrifuged at 19,587×
g for 10 min. Next, 17 mL of n-hexane was added to 5 mL of the supernatant, followed by vortex mixing for 15 min. After centrifugation at 9598×
g for 10 min, the n-hexane layer was removed, and the non-organic layer was collected. The lipid removal step was repeated twice more using the sample residue.
Morpholine was spiked and analyzed by GC-MS after defatting and derivatization of the fruit samples. By adding the defatting step to the previous method, the final extract solution was clear. For derivatization, 200 μL of 0.05 M HCl and 200 μL of saturated sodium nitrite were added to 2.0 mL of the defatted sample extract and the mixture was vortexed for 30 s. HCl was added to improve recovery and the pH was optimized by adding different amounts of HCl. Herein, different pH levels (pH 1.5, 3.0, and 6.5) were tested and recovery was compared. To optimize the derivatization method, accuracy and precision were determined at pH 1.5, 3.0, and 6.5. Recovery of pH 6.5 was 124.3%, whereas, the recoveries of morpholine stearate at pH 1.5 and 3.0 were 102.2% and 106.5%, respectively. The precision (relative standard deviation, RSD%) of morpholine stearate measurements at pH 1.5, 3.0, and 6.5 were 0.8, 11.6%, and 7.0%, respectively. The morpholine recovery% and RSD% were best at pH 1.5. Thus, pH 1.5 should be used to analyze morpholine.
Then, the extract was heated at 40 °C for 5 min and cooled. Finally, 0.5 mL of dichloromethane was added, and the mixture was vortexed for 1 min and left to stand for 10 min. An aliquot of the organic layer was collected and placed in an amber vial after filtration through a 0.22-μm filter. All samples were stored at −20 °C before analysis, which was performed in 1–2 days.
2.4. Method Validation (Method Detection Limit, Method Quantification Limit, Linearity, Accuracy, Precision, Cross-Lab Validation, and Measurement Uncertainty)
For method validation, apple and oranges were purchased from a local market in Anseong, Korea. Apple peel, apple pulp, orange peel, and orange pulp were lyophilized and stored at −80 °C until analysis. This method was validated for MDL, MQL, linearity, accuracy, precision, cross-lab validation, and measurement uncertainty. The MDL and MQL were based on International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) guidelines and calculated as LOD (limit of detection) = 3.3σ/S and LOQ (limit of quantification) = 10σ/S, where σ is the standard deviation and S is the slope of the standard curve [
30]. To evaluate intra-day accuracy and precision, known amounts of morpholine were added to the four matrices at final concentrations of 10–400 μg/kg and measurements were repeated 5 times. For inter-day accuracy and precision, known amounts of morpholine were added to the four matrices at final concentrations of 25, 100, and 400 μg/kg for 3 days. Accuracy and precision were validated in three different laboratories for cross-lab validation according to ICH guidelines [
30].
Measurement uncertainty was also estimated for the morpholine analysis method in apple peel, apple pulp, orange peel, and orange pulp using the combined standard uncertainty based on the Guide to the Expression of Uncertainty in Measurement Guide by KRISS [
31] and EURA CHEM Guide [
32]. The intra-laboratory data of the reference material, calibration curves, repeatability, and sample preparation were used for estimation. To obtain measurement uncertainty (U), a coverage factor of ~ 95%, where k = 2, was used [
33].
2.5. Morpholine Analysis by GC-MS
Morpholine analysis was performed on a GC-MS instrument (7890A GC-5975 MSD, Agilent Technologies, Santa Clara, CA, USA). The analysis method was modified from that described by Cao et al. by changing the analytical column and split ratio, etc. [
28]. Separation was carried out on a DB-1701 column (30 m × 0.25 mm i.d., 0.25 µm film thickness, Agilent Technologies). A DB-wax column (30 m × 0.25 mm i.d., 0.25 µm film thickness, Agilent Technologies) was compared to optimize the method; the DB-1701 column resulted in better separation with a better response (data not shown). The injection volume was 1 μL and the sample was vaporized at 250 °C with a 1:7 split ratio. The GC oven temperature was operated as follows: 100 °C for 4 min, heating to 120 °C at 10 °C/min and held for 3 min, and then heating to 250 °C at 20 °C/min and held for 5 min. The electron energy was 70 eV. The flow rate was 2.0 mL/min of He (99.999%). The transfer line temperature, quadrupole temperature, and electron impact ionization source temperature were held at 280 °C, 150 °C, and 230 °C, respectively. The scan rate was 3.2 scans/s. Four different ions were selected to detect and quantify
N-nitroso-morpholine (qualifier ion: m/z 86; quantifier ion: m/z 116) and its isotope (qualifier ion: m/z 64; quantifier ion: m/z 124) in selected ion monitoring (SIM) mode.