Rapid determination of antioxidant molecules in volatiles of rose tea by gas chromatography–mass spectrometry combined with DPPH reaction

Abstract

Volatiles have been regarded as active substances in many foods, whose chemicals can be analyzed by GC–MS qualitatively and quantitatively. However, the activities of volatiles are often studied as a whole, and it has no an effective method to determine that which molecule is active in volatiles by far. In order to identify the antioxidant molecules in volatiles, a rapid determination method was developed by GC–FID/MS combined with DPPH radical reaction in this study. Three antioxidant molecules were identified and validated among 20 components in rose tea infusion. Their activity validation and the methodological evaluation indicated this method could be used for distinguishing antioxidant molecules in volatiles rapidly and effectively.

Introduction

A large number of researches indicated that overproduction of oxygen free radicals in the human bodies is a major cause in aging, death, and many serious diseases, including arteriosclerosis, diabetes, cancer, and neurodegenerative diseases, etc. (Ani et al. 2006). At a young age, enough endogenous free radical scavengers in vivo can prevent the accumulation of free radicals to damage the body. But those endogenous scavengers will be reduced gradually with age (Harman 1956). Thus, the supplement of exogenous scavengers and inhibitors of oxygen free radicals timely are might be effective ways to prevent those diseases related to oxygen free radicals.

Volatiles is a kind of active substances in many foods, and can produce significant antioxidation (Mohamd et al. 2011; Sarbanha et al. 2011; Wungsintaweekul et al. 2007). And more and more people prefer to using some antioxidant food to protect themselves against some diseases related to oxidative stress (Gülçin 2012; Lorenzo et al. 2018). Among them, tea is a famous one. Previous researches showed that the leaves, buds, and flowers from plants being used as a tea, not only has a pleasant fragrance, but also is rich in volatiles with antioxidant activities (Ahmad et al. 2015; Wei and Shibamoto 2007; Yin et al. 2015).

To date, determination of antioxidant molecules in mixtures usually requires the help of instrumental analysis, chemical reactions, and some bioassay. Generally, the methods to identify and analyze whether nonvolatile molecules having antioxidant activities usually include HPLC–MS analysis, separation, purification, and bioassays for single compound in vitro (Park et al. 2000), such as thiobarbituric acid method (Rio et al. 2005), 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging assays (Brand-Williams et al. 1995). Notably, on-line HPLC-DPPH analysis is a rapid and effective method that could identify the antioxidant molecules in nonvolatile substances directly (Kosar et al. 2017; Li et al. 2012). However, to our knowledge, it has no an effective method to determine the antioxidant molecules from volatiles so far. The chemicals in volatiles are often analyzed by gas chromatography–mass spectrometry (GC–MS) and its activity are studied as a whole, respectively (Belviso et al. 2013; Ud-Daula et al. 2016).

In terms to the principle of GC analysis, in this study, a rapid method to the determination of antioxidant molecules of volatiles in rose tea was proposed, which is a popular drink in many Asia areas and rich in volatiles. That method is developed by the analyzing volatiles before and after its reaction with DPPH using the GC and GC–MS. As results, 20 components were found from rose tea volatiles, amounting to 69.67% of the total volatiles. Among them, three active molecules producing DPPH radical scavenging were confirmed by the activity validation and methodological evaluation. It not only can help to study the antioxidant molecules from volatiles, but also perform the more reasonable quality control of active volatiles in the future.

Materials and methods

Standards and reagents

Eugenol, 4-hydroxyphenethyl alcohol, and isoamyl alcohol were purchased from National Institutes for Food and Drug Control (Beijing, China). BHA and benzyl alcohol both were obtained from Aladdin Industrial Corporation (Shanghai, China). DPPH was purchased from Tokyo Chemical Industry (Tokyo, Japan). And ethyl acetate and methanol were produced by Xilong Scientific Co., Ltd (Chengdu, China).

Samples

The dried rose analyzed in this study was purchased from the local market, and identified as Rosa rugosa Thunb. F. plena (Regel) Byhouwer by assistant Prof. Weifeng Dai (Kunming University of Science and Technology). The rose weighed 4 g accurately, was placed in a conical flask (150 mL) with 50 mL water (80 °C) and covered with a plug. Then, the conical flask was boiled for 20 min at 80 ± 0.2 °C and filtered while hot. The filtrate was collected for further treatment, stored at 4 °C.

Volatile collection

The fifty milliliter of filtrate was placed in a conical flask (150 mL), and volatile in filtrate was extracted with 50 mL, 30 mL, and 20 mL ethyl acetate, respectively (Wang et al. 2012). Ethyl acetate extract in a conical flask was collected, and dried by 5 g anhydrous sodium sulfate. Under reduced press, the volatile, named sample A, was concentrated to 1 mL at 36 °C, and stored at 4 °C for analysis by GC–MS.

Analysis of volatiles

The volatile was analyzed by GC–MS (Agilent GC7890B/MS5977A, Agilent Technologies Inc., USA) fitted with an HP-FFAP capillary column (30 m × 0.32 mm internal diameter, 0.25 μm of film thickness, Agilent). The temperature program of initial oven was as follows: 70 °C, raised at 8 °C/min, up to 150 °C, then raised at 4 °C/min up 220 °C and maintained for 8 min. 1 μL of the organic layer (“Volatile collection” section) was injected. The temperature of the injector was set at 250 °C, and the splitless mode was chosen. Nitrogen was used as carrier gas, whose flow rate was 2.5 mL/min.

The temperature of the transfer line was 250 °C, and the temperature of the quadrupole was 150 °C. The ionization potential of the mass selective detector was 70 eV, and SCAN mode was selected in this study. All compounds isolated were estimated by NIST14 database.

Identification of antioxidant molecules in volatiles

Firstly, 100 μL of sample A was analyzed by GC–FID/GC–MS. Then, 3 mg DPPH was added to another 100 μL of sample A and incubated for 30 min at room temperature (Molyneux 2004). That sample A after the reaction, renamed it as sample B, was analyzed by GC–FID directly. Meantime, 0.1 mg isoamyl alcohol, used as an internal standard, was added into samples A and B, respectively. The analytic conditions of samples A and B, on the GC, were the same as the above description in a section of analysis of volatiles, and the temperature of FID was set at 300 °C. The ratio of each chromatographic peak area with internal standard peak area was expressed as R, which was used to evaluate the changes of the peak area of each component before and after the reaction with DPPH. In the case of some constituent with R-value reduced in sample B, it would be regarded as the active molecule with the antioxidant function of DPPH radical scavenging.

Activity validation

The activity validation of the molecules was carried out in this part. The molecules, with R-value reduced, R-value unchanged, and no antioxidant reports, all need to be tested by DPPH radical scavenging assay in vitro using their corresponding standards, respectively. The steps were as follows: 1 mL DPPH-methanol solution (200 μg/mL) was placed in 1 mL standard-methanol solution (1 mg/mL), and its absorbance was measured on Ultraviolet spectrophotometer at 517 nm after reaction for 30 min. The standard-methanol solution was replaced by methanol in the blank group, and each group has repeated three times. The radical-scavenging activity of standard was evaluated by the value of P, which was calculated by absorbencies of two groups.

Further, in order to validate that the area peak of active compounds in GC will reduce in reacted sample with DPPH, the standard samples were made by the standards of eugenol, 4-hydroxyphenethyl alcohol, 3(2)-tert-butyl-4-hydroxyanisole (BHA) and benzyl alcohol for confirmation of the changes of peak areas in samples before and after reaction with DPPH, in which BHA was selected as positive control, and benzyl alcohol was selected as negative control. In a standard work solution, the concentration of each compound was 1 mg/mL. The standard sample, before reaction with DPPH, named sample C, was determined by GC. Meantime, 10 mg DPPH was added into another standard work solution. The reacted standard work solution, named sample D, was also determined by GC after reaction for 30 min.

Methodological evaluation

The performance parameters used to assess methodological evaluation were as follows: linearity, precision, recovery, limit of detection (LOD) and Limit of quantity (LOQ).

Correlation peak area and concentration were used as analytical signals for study linearity. Six concentration levels of eugenol and 4-hydroxyphenethyl alcohol were set at 0, 12.5, 25, 50, 100, 200 μg/mL and 0, 6.25, 12.5, 25, 50, 100 μg/mL, respectively.

Precision was expressed as repeatability, and the value was expressed as relative standard deviation (RSD), which was obtained by analysis of the same sample for eight times (Gao et al. 2012). The sample was taken from the filtrate of rose tea, and the sample was treated according to the above conditions in sections of analysis of volatiles and activity validation. The value of repeatability was calculated by areas of correlation peaks.

For the recovery study, three samples were analyzed. The additive amount of sample was set as 1.25, 2.5, 5 mg/L, respectively, and each level was repeated three times. All samples were treated on the basis of steps described in sections of analysis of volatiles and activity validation. Correlation peak areas were used to calculate the value of recovery.

Furthermore, the limit of detection (LOD) and limit of quantity (LOQ) were determined by dilutions of working standard solution subsequently. Their values were expressed as the corresponding concentration of working standard solution when S/N (signal–noise ratio) was established at 3 and 10, respectively.

Results and discussion

Identification of compounds in volatiles

Sample A was analyzed by GC–MS, and 20 components, accounting for 69.67% of the total peak areas, were identified from it by matching their mass spectra with NIST14 database (Table 1). The main types of those constituents are alcohols, aldehydes, phenols, acids, and hydrocarbons. The content of phenethyl alcohol was 29.41% of the total. The common constituents, benzaldehyde, benzyl alcohol, phenethyl alcohol, methyl eugenol, and eugenol, found in rose essential oil (Xue and Li 1989; Umezu et al. 2002), account for 46.44% in the volatiles from rose tea, and the antioxidant activity of eugenol has been reported (Arenas et al. 2011), that indicating those also are potential antioxidants in rose tea.

Table 1.

Identified compounds and their average area in volatiles from rose tea infusion

RT (min)	Name	Area (%)	CAS no.	Formula	Q1 (m/z)
3.670	Tetradecane	1.29	629-59-4	C14H30	57
4.590	Furfural	0.56	98-01-1	C5H4O2	96
5.282	Benzaldehyde	0.42	100-52-7	C7H6O	106
6.095	Hexadecane	0.31	544-76-3	C16H34	57
7.392	undecylcyclopentane	0.31	6785-23-5	C16H32	69
9.806	Benzyl alcohol	3.81	100-51-6	C7H8O	108
10.242	Phenethyl alcohol	29.41	60-12-8	C8H10O	91
11.067	1-(1H-Pyrrol-2-yl)ethan-1-one	1.26	1072-83-9	C6H7NO	94
11.624	Methyl eugenol	2.86	93-15-2	C11H14O2	178
13.914	Eugenol	9.94	501-19-9	C10H12O2	164
15.667	4H-Pyran-4-1,2,3-dihydro-3,5-dihydroxy-6-methyl-	6.93	28564-83-2	C6H8O4	144
16.243	1-Ethenyl-3-nitrobenzene	0.71	586-39-0	C8H7NO2	142
18.431	Benzoic acid	1.29	65-85-0	C7H6O2	95
19.893	5-Hydroxymethylfurfural	3.01	67-47-0	C6H6O3	97
23.024	Dibutyl phthalate	0.65	17851-53-5	C16H22O4	149
26.011	3-Phenylpropenoic acid	2.52	621-82-9	C9H8O2	147
26.895	n-Hexadecanoic acid	0.94	57-10-3	C16H32O2	73
28.118	Nicotinamide	0.52	98-92-0	C6H6N2O	122
28.825	4-Hydroxyphenethyl alcohol	2.33	501-94-0	C8H10O2	107
29.551	Squalene	0.6	111-02-4	C30H50	69

Identification of antioxidant compounds

The gas chromatograms of samples A and B were shown in Fig. 1. Compared the R values of 23 chromatographic peaks in sample A with those in sample B, the R values of those peaks numbered 10, 13, 14, 15, 23 were significantly reduced in sample B by 25.85 ± 4.27%, 33.04 ± 0.83%, 72.40 ± 2.86%, 41.34 ± 9.47%, 24.13 ± 3.43%, respectively (Fig. 2). Further comparison of retention times and mass spectra of those peaks between samples A and B indicated three of them were eugenol, 4-hydroxyphenethyl alcohol, 4H-pyran-4-1,2,3-dihydro-3,5-dihydroxy-6-methyl (Table 2). Unfortunately, another two components, whose area peaks reduced in sample B, failed to identify because of their poor separations.

Fig. 1.

The gas chromatograms of the samples A and B, standard work solutions C and D

Fig. 2.

The change of peak areas in gas chromatograms of samples A and B (*P < 0.05 as compared with sample A)

Table 2.

The details of the compounds in standard samples, whose peak areas were reduced in reacted samples

Name (peak no.)	Standard samples
	RT (min)	GC peak area (C)	GC peak area (D)	Area reduction percentage (%)	MS fragments (m/z)
Benzyl alcohol (7)	10.515	139509221	134237276	3.03 ± 0.85	108, 79, 77, 51
	10.513	138687346	134227840
	10.482	139015855	136092517
Eugenol (13)	14.775	117014259	31830749	73.31 ± 0.48	164, 149, 131, 103, 77
	14.803	116006034	30461757
	14.792	116505723	30996487
4-Hydroxyphenethyl alcohol (23)	30.074	60425812	34001560	43.57 ± 0.71	138, 108, 77
	30.085	60397473	34547452
	30.022	60891347	33987481
BHA	21.128	115523943	1171300	99.01 ± 0.03	180, 165, 137
	21.116	119742072	1205223
	21.175	113522639	1087862

In the previous literatures (Arenas et al. 2011; Cechovská et al. 2011; Wang 2009), the antioxidant effects of eugenol, 4-hydroxyphenethyl alcohol, and 4H-pyran-4-1,2,3-dihydro-3,5-dihydroxy-6-methyl have been reported. Thus, it speculated R values of the chemicals with antioxidant effects in rose tea should be reduced in sample B. In order to further verify that speculation, activity validation was carried out. Standard samples C and D were made for confirmation of the area peak of active compounds reduced in the reacted sample with DPPH. As results, the peak area of BHA reduced by 99.01 ± 0.03%, but that of benzyl alcohol has no significant change (Table 2 and Fig. 1). In vitro antioxidant experiment, the value of P > 0.05 also suggested that benzyl alcohol has no antioxidant activity (Table 3). The results of the validation of activity indicated that components with R-value reduced have antioxidant activities, but components with R-value unchanged have no antioxidant activities.

Table 3.

The result of antioxidant experiment of benzyl alcohol in vitro

Analyte	Absorbance			P
	1	2	3
Blank group	1.097	1.101	1.104	0.598 (> 0.05)
Benzyl alcohol	1.104	1.100	1.102

Furthermore, the reasons for the decrease of peak areas in reacted samples with DPPH were speculated. As we known, there was one simultaneous step in the reaction of antioxidants with DPPH. Taking an example of phenol, first was H-atom transfer (HAT), and the second was electron transfer (ET). Their reaction equations were listed as (1) and (2), respectively (Foti et al. 2004). But in polar solvents, ET was the important mechanism, which capable of forming strong hydrogen bonds with the ArOH molecules (Villaño et al. 2007). Based on that, the physical properties and chemical structures of active compounds in volatiles would be changed after the reaction, such as vaporization temperature, polarity. The vaporization temperature of products might be higher than those of antioxidants, resulted in non-vaporization of products and the different retention time of products in gas chromatogram. On the other hand, the R-values of some peaks, which was labeled as 1, 21, 22, were significantly increased in sample B, compared with those in sample A (Fig. 2), speculating some new products produced after reaction with DPPH radical in sample B, giving rise to increase of these peak areas. It indicated indirectly that the physical properties and chemical structures of active compounds in volatiles would be changed after reaction. However, it needed more researches and validations in depth

$${\text{ArOH}}^{+}+\text{DPPH}\to {\text{ArO}}^{·}+\text{DPPH}·\text{H}$$ 1

$${\text{ArO}}^{·}+\text{DPPH}\to \text{products}.$$ 2

Methodological evaluation

The results for methodological evaluation were presented as below. The values of LOQ and LOD of two compounds were shown in Table 4. It revealed that the values of repeatability were less than 3.5%, suggesting the present method had good precision.

Table 4.

The data of methodological evaluation

Analyte	Repeatability (RSD %, n = 8)	LOD (mg/L)	LOQ (mg/L)	Linearity range (μg/mL)	R2	Recovery (%, n = 3)
						1.25 mg/L	2.5 mg/L	5 mg/L
Eugenol	1.37	0.051	0.016	0–200	0.99918	85.01	87.74	91.07
4-Hydroxyphenethyl alcohol	3.17	0.036	0.091	0–100	0.99392	87.7	89.37	86.02

Six concentration levels of eugenol and 4-hydroxyphenethyl alcohol were set to studying the linearity. When the value of R² was greater than 0.99 (Meersche et al. 2015), the linearity of the calibration curve was considered to be acceptable. In this part, their linearity ranges were set to 0–200 μg/mL and 0–100 μg/mL, respectively. The value of R² of eugenol and 4-hydroxyphenethyl alcohol were obtained as 0.99918 and 0.99392 (Table 4), respectively, both greater than 0.99, indicating this method has nice linearity.

Further, the recovery values of two compounds were greater than 70% and lower than 120% (Table 4), respectively, which were considered to be acceptable (Lopez et al. 2015), indicating this method was suitable for the extraction of the target compound.

All the above results of the methodological evaluation have proved this method was an appropriate analysis of the target compound.

Conclusion

The aim of our study was to develop a rapid method to determine the antioxidant molecules in volatiles. In the current study, the main components in the essential oil of rose have been proved to exist in the volatiles of rose tea. By the development of this method, the antioxidant molecules in volatiles can be studied more targeted, and the quality control and standard of volatiles can be formulated more scientifically in the future.

Compliance with ethical standards

Conflict of interest

All authors declare that they have no conflict of interest.