Development and diffusion of practical technologies for pesticide residue analysis

Abstract

We often encounter difficulties in analyzing pesticide residues in food, since the residue level to be measured is infinitesimal but the food samples may contain a variety of matrices, including interferences, for chromatographic analysis. Thus it is important to find out the key points to be checked in the extraction, purification, and determination steps of analysis. We have attempted to develop practical technologies for the analysis of pesticide residues in raw agricultural commodities based on the following studies: The effect of processing and cooking on pesticide residue levels in several crop samples (rice, wheat, soybean, and sesame) was investigated. The processing factor is useful to estimate the amount of exposure to each pesticide residue for risk assessment and is helpful in setting the maximum residue limits of processed foods. In addition, residue levels in the peel and pulp of watermelon, melon, and kiwi fruit samples were examined to confirm the differences in the distribution of residues. As a basic study, the effect of water-soaking pretreatment of powdered dry cereal on extraction efficiency was examined, and an optimal time for water-soaking was found. A recent study examined the ability of several new types of solid-phase extraction columns to remove matrices in brown rice samples and verified the effective purification method for each sample.

Introduction

The principal procedures for the analysis of pesticide residues in food generally consist of the following four steps: 1) sample preparation, 2) extraction, 3) purification, and 4) determination. This principle is the same now as formerly. However, the details of each step have been changed in accordance with the development of new analytical instruments and purification tools.

Before the 1980s, the main analytical instrument was the gas chromatograph (GC), due to the chemical properties of pesticides and the performance of instruments. In the case of pesticides that could not be directly measured with a GC, derivatization was conducted before the analysis. After that, the separation column of the GC was changed from a packed column to capillary one. As to the general-purpose type of GC, a mega-bore column (0.53 mm I.D.) that can be used at the same gas flow rate as a packed column had become effective. In addition, the gas chromatograph-mass spectrometer (GC-MS) and high-performance liquid chromatograph (HPLC) were introduced into practical use. Consequently, the HPLC was used to analyze chemicals such as polar pesticides that cannot be directly measured with a GC. For the process of purification, a commercial solid-phase extraction column (mini-column) had come to be used more frequently than an open column packed with sorbent in a chromatographic tube. This led to a significant reduction in the volume of solvent used and a shortening of the time required for purification. In the 1990s, the liquid chromatograph-mass spectrometer (LC-MS) was developed, and the basis for its practical use was established along with the appearance of atmospheric pressure ionization (API) methods, including electrospray ionization (ESI), atmospheric pressure chemical ionization (APCI), and atmospheric pressure photoionization (APPI). Thereafter, rapid progress in the development of measuring instruments was made, and the LC-MS(/MS) and GC-MS(/MS) became the major instruments for pesticide residue analysis in food in the 2000s. Furthermore, an ultra-high-performance liquid chromatograph (UHPLC) with a high-pressure liquid transfer ability has been developed. As a result of this and the increased sensitivity of measuring instruments, the sample amount required for analysis has been reduced to about 1/5 of that in the previous analysis, which is due to the increase in sensitivity of measuring instruments.

Under these circumstances of pesticide residue analysis, we have conducted various studies with special attention to the following issues: 1) the effect of processing and cooking on pesticide residue levels in raw agricultural commodities (RACs), 2) the distribution of pesticide residues in the peel and pulp of watermelon, melon, and kiwi fruit samples, and 3) check points to be considered for extraction and purification in pesticide residue analysis. This manuscript describes the results of these studies, providing new information on pesticide residue analysis.

1. Effect of processing and cooking on pesticide residue levels

Although agricultural, stock farm, and marine products can be eaten raw, they are usually cooked or processed for eating. Accordingly, it is important to determine the actual residue levels of pesticides in not only RACs but also cooked and processed foods and to provide information on the precise estimation of residue intake through foodstuffs. The precise estimation of pesticide residue intake through foodstuffs would be helpful to set the maximum residue limits (MRLs) for pesticides in agricultural products at harvest and also to monitor the residue levels in those products when marketed.

The crop samples were examined for changes in the residue levels of pesticides during processing and/or cooking process after harvest. The present study was aimed at the estimation of pesticide residue intake through food and the effects of processing and cooking on the pesticide residue levels in foodstuffs after harvest.

The residue levels of pesticides in several crop samples at harvest were compared with those after processing and/or cooking, and the residual amount ratio (RAR, %) to the initial value was calculated on the basis of the total amount of pesticide residues for each crop sample. In addition, the processing factor (Pf) was determined by dividing the concentration (mg/kg) of pesticide residues in processed and/or cooked products by the initial concentration in the original crop sample at harvest. The Pf is useful to estimate the exposure amount of each pesticide residue for risk assessment. It is also helpful to set MRLs for processed foods. The correlation of these RARs and Pfs with the physical and chemical properties of pesticides was also examined. The results of these investigations on each crop sample were as follows.

1.1. Rice sample⁵⁾

Test samples were prepared after the preharvest application of pesticides in Japan and the USA.

The effects of processing and cooking on the pesticide residue levels in rice samples were investigated for 12 pesticides. The processing and cooking methods are shown in Fig. 1. In the polishing process, the RARs (%, total pesticide residue amount in the product/that in brown rice) of polished rice ranged from 9.7 to 65%. These values varied from pesticide to pesticide. The Pfs [pesticide concentration (mg/kg) in the product/that in brown rice] of polished rice ranged from 0.11 to 0.73 (Table 1). The loss of pesticides during processing and/or cooking did not correlate to any single physical or chemical property. Investigation of the changes in pesticide residues during processing and/or cooking is useful not only to establish MRLs but also to recognize actual levels of pesticide residues in food. The concentrations of pesticide residues were determined using a GC(NPD), an HPLC, a GC-MS, or an LC-MS.

Fig. 1. Processing and cooking methods for the rice sample.

Table 1. Processing factors (Pf) of rice product samples

Pesticide	ftsb)	arc)	Concentration (mg/kg)	Pfa)
			Brown rice	Brown rice	Polished rice	Rice bran	Washed brown rice	Washed polished rice	Cooked brown rice	Cooked polished rice
Etofenprox	JP	5X	0.178	1.00	0.15	10.6	0.81	0.034	0.31d)	<0.17d)
									0.34e)	0.022e)
Fenitrothion	JP	5X	0.988	1.00	0.27	6.96	0.78	0.11	0.23	0.017
	USA1	1X	0.137	1.00	0.29	6.74	0.66	0.094	0.073	0.029
	USA1	5X	0.296	1.00	0.30	7.43	0.93	0.10	0.022	0.020
Diazinon	JP	5X	0.365	1.00	0.23	6.47	0.91	0.10	0.42	0.033
Parathion-methyl	USA1	1X	0.135	1.00	0.27	7.04	0.58	0.038	0.12	<0.022
	USA1	5X	0.241	1.00	0.26	7.76	0.61	0.066	0.10	<0.012
Malathion	JP	5X	0.029	1.00	<0.17	8.52	0.76	<0.17	0.17	<0.10
	USA1	1X	0.502	1.00	0.15	6.95	0.59	0.043	0.10	0.0078
	USA1	5X	5.34	1.00	0.26	9.23	0.68	0.029	0.13	0.0048
	USA2	1X	0.253	1.00	0.12	8.50	0.46	0.020	0.10	<0.020
	USA2	5X	1.21	1.00	0.19	9.34	0.59	0.030	0.15	0.0058
Carbaryl	JP	5X	4.11	1.00	0.45	5.50	0.83	0.25	0.0046	0.010
	USA1	1X	1.36	1.00	0.43	5.26	0.62	0.13	<0.0022	0.013
	USA1	5X	4.70	1.00	0.40	6.38	0.33	0.095	0.0017	0.010
Carbofuranf)	USA1	5X	0.03	1.00	0.33	4.67	0.47	<0.33	<0.20	<0.20
	USA2	5X	0.16	1.00	0.44	6.88	0.56	0.19	<0.038	<0.038
Phthalide	JP	5X	1.12	1.00	0.14	8.75	0.62	0.062	0.18d)	0.017d)
									0.20e)	0.025e)
Phosphamidong)	USA1	1X	0.105	1.00	0.29	4.00	0.60	0.097	0.076	0.019
	USA1	4X	0.453	1.00	0.66	3.09	0.49	0.33	0.13	0.059
	USA3	1X	0.212	1.00	0.39	6.46	0.50	0.086	0.052	0.019
	USA3	4X	3.68	1.00	0.38	6.55	0.54	0.30	0.23	0.083
Imidacloprid	JP	5X	0.329	1.00	0.58	5.53	0.73	0.46	0.27d)	0.13d)
									0.37e)	0.26e)
Dimethoate	USA1	5X	0.011	1.00	0.73	<1.82	0.82	<0.45	<0.27	<0.27
Diquat	USA1	1X	0.17	1.00	0.24	9.29	0.82	0.18	0.24	<0.12
	USA1	5X	1.37	1.00	0.11	9.27	0.93	0.044	0.33	0.022

^a) Residue concentration in a product (mg/kg)/residue concentration in brown rice (mg/kg). ^b) Field trial test site: JP; Japan Plant Protection Association, Ibaraki, Japan, USA1; Mid-South Ag Research, Crittenden County, Proctor, AR, USA, USA2; Research 2000, Glenn County, Chico, CA, USA, USA3; Research 2000, Glenn County, CA, USA. ^c) Aplication rate as a relative ratio of the GAP maximum rate for pre-harvest treatment. ^d) Using a rice cooker. ^e) Using a pressure cooker. ^f) Carbofuran+3-OH-carbofuran. ^g) Phosphamidon (cis+trans)+N-desethyl phosphamidon (cis).

1.2. Wheat sample^6–8)

Test samples were prepared after the preharvest application of pesticides in Japan, the USA, and Australia. The effects of processing and cooking on the pesticide residue levels in wheat samples were investigated for 10 pesticides. The processing products includes primary products (milling process: bran, shorts, flour, and rest flour) and secondary products (making bread [using flour and whole wheat], lumps of udon [Japanese wheat noodle] and Chinese noodle). In the milling process, the RARs (%, total pesticide residue amount in the product/that in wheat grain) of wheat bran were greater than 70%. The RARs of flour ranged from 1.7 to 23%. The RARs and processing factors for the primarily processed products showed no significant differences among the pesticides investigated, but those for the secondarily processed products exhibited a marked difference between pesticides with and without susceptibility to thermolysis/hydrolysis. Ethylene thiourea (ETU) was detected in all samples containing moncozeb residues. The degradation from moncozeb to ETU was confirmed to be accelerated during the process of baking bread by heating (Fig. 2). The Pfs [pesticide concentration (mg/kg) in the product/that in the wheat grain] of flour ranged from 0.030 to 0.46 (Table 2). The pesticide residue concentrations were determined using a GC(NPD), an HPLC, a GC-MS, or an LC-MS/MS.

Fig. 2. RARs (%) of mancozeb and ETU in primary and secondary products from the wheat grain sample.

Table 2. Processing factors (Pf) of wheat product samples

Pesticide	ftsb)	arc)	Concentration (mg/kg)	Pfa)
			Wheat grain	Wheat grain	Flour	Rest flour	Bran	Shorts	Bread (white)	Bread (whole meal)	Udond)	Chinese noodlee)
Disulfotonf)	USA1	5X	0.007	1.00	0.29	0.57	5.00	1.70	—	—	—	—
	USA2	5X	0.007	1.00	<0.219	0.71	4.00	<0.86	—	—	—	—
Fenitrothion	JP	5X	0.830	1.00	0.17	0.42	3.54	2.84	0.073	0.72	0.10	0.10
	USA1	1X	0.319	1.00	0.085	0.15	3.54	1.37	0.044	0.65	0.053	0.066
	USA1	5X	2.45	1.00	0.073	0.16	3.46	1.07	0.033	0.62	0.043	0.049
Parathion-methyl	USA1	1X	0.151	1.00	0.19	0.28	3.15	1.35	0.093	0.54	0.11	0.13
	USA1	5X	2.35	1.00	0.19	0.32	2.83	1.27	0.073	0.51	0.10	0.12
Malathion	JP	5X	0.084	1.00	0.18	0.63	3.82	4.21	<0.060	0.61	<0.060	<0.060
Carbofurang)	USA1	5X	0.13	1.00	0.077	<0.069	5.62	0.77	<0.069	0.48	<0.069	<0.069
	USA3	5X	0.18	1.00	0.085	0.25	4.66	3.31	<0.050	0.97	0.11	0.11
Benomyl	JP	5X	1.70	1.00	0.46	0.49	2.09	1.58	0.27	0.68	0.29	0.31
Phosphamidonh)	AUS1	1X	0.019	1.00	0.16	0.16	5.16	1.47	<0.16	<0.16	<0.16	<0.16
	AUS1	5X	0.112	1.00	0.057	0.075	4.43	1.09	0.028	0.15	0.028	0.038
	AUS2	1X	0.135	1.00	0.030	0.081	3.76	1.08	0.030	0.24	0.022	0.030
	AUS2	5X	0.462	1.00	0.041	0.089	3.85	1.49	0.015	0.23	0.022	0.022
Dimethoate	USA1	1X	0.308	1.00	0.12	0.15	3.22	1.24	0.039	0.35	0.065	0.038
	USA1	5X	2.23	1.00	0.17	0.22	3.03	0.97	0.049	0.40	0.086	0.090
Mancozebi)	USA1	1X	0.18	1.00	0.28	0.33	3.72	0.56	0.17	0.67	0.22	0.17
	USA1	5X	1.46	1.00	0.30	0.35	2.88	0.44	0.16	0.53	0.18	0.16
	USA3	1X	1.05	1.00	0.40	0.42	2.38	1.09	0.23	0.49	0.30	0.27
	USA3	5X	1.98	1.00	0.26	0.19	1.59	0.71	0.14	0.47	0.17	0.15
Diquat	USA1	1X	0.30	1.00	0.23	0.43	3.70	1.33	0.12	0.61	0.15	0.11
	USA1	5X	4.73	1.00	0.082	0.17	4.52	1.57	0.042	0.52	0.050	0.038

^a) Residue concentration in a product (mg/kg)/residue concentration in wheat grain (mg/kg). ^b) Field trial test site: JP; Japan Plant Protection Association, Ibaraki, Japan, USA1, 2; Northern Plains Ag Research, Cass County, Gardner, ND, USA, USA3; Bennett Ag Reseach, Adair County, Kirsville, MO, USA, AUS1, 2; Martin Collett, Agriasearch Services Pty. Ltd., Bathuurst, NSW, Australia. ^c) Application rate as a relative ratio of the GAP maximum rate for pre-harvest treatment. ^d) Lump of udon. Left for 2 hr at 25°C, after 10 g of sodium chloride and 90 mL water added to 200 g of wheat flour and kneaded well. ^e) Lump of Chinese noodle. Left for 1 hr at ambient temperature, after 80 mL of water and 5 mL of an aqueous solution which contained potassium carbonate and sodium carbonate at concentrations of 20% and 3.3%, respectively, were added to 200 g of wheat flour and kneaded well. ^f) PSSO₂+POSSO₂. ^g) Carbofuran+3-OH-carbofuran. ^h) Phosphamidon (cis+trans)+N-desethyl phosphamidon (cis). ⁱ⁾ Mancozeb+ETU.

1.3. Soybean sample^6,9,10)

Test samples were prepared after the preharvest application of pesticides in Japan and the USA. The effects of processing and cooking on the pesticide residue levels in soybean samples were investigated for 19 pesticides. The processing and cooking methods are shown in Fig. 3. In the soaking process, the RARs (%, total pesticide residue amount in the product/that in soybeans) of soaked soybeans were greater than 60% for most of the pesticides investigated. The RARs of soymilk ranged from 37 to 92%, and that of tofu ranged from 7 to 63%. The Pfs [pesticide concentration (mg/kg) in the product/that in soybeans] of tofu ranged from 0.026 to 0.33 (Table 3). These values varied among pesticides. There was a high correlation between the log P_ow and the residual ratio for tofu. The test described here should be useful to obtain the residual ratios of pesticide residues in the processing and/or cooking steps. The pesticide residue concentrations were determined using a GC(NPD, ECD), a GC-MS, an HPLC, or an LC-MS.

Fig. 3. Processing and cooking methods for the soybean sample.

Table 3. Processing factors (Pf) of soybean product samples

Pesticide	ftsb)	arc)	Concentration (mg/kg)	Pfa)
			Soybean (dry)	Soybean	Soaked soybean	Soymilk	Okara	Tofu	Boiled soybean	Natto
Esfenvalerate	USA1	5X	0.042	1.00	0.45	<0.12	<0.12	<0.12	—	—
	USA2	5X	0.043	1.00	0.56	<0.12	<0.12	<0.12	—	—
Permethrind)	JP	5X	0.04	1.00	0.45	<0.025	<0.25	0.33	<0.25	0.33
Disulfotone)	USA2	5X	0.027	1.00	0.22	<0.037	<0.074	<0.074	—	—
Fenitrothion	USA2	1X	0.018	1.00	0.56	<0.27	<0.27	<0.27	—	—
	USA2	5X	0.244	1.00	0.41	0.14	0.18	0.28	0.20	—
	JP	5X	0.43	1.00	0.50	0.098	0.16	0.25	0.11	0.12
Procymidone	JP	5X	1.29	1.00	0.39	0.90	0.16	0.22	0.32	—
	JP	5X	1.99	1.00	0.41	0.10	0.060	0.21	0.23	0.42
Parathion-methyl	USA2	1X	0.039	1.00	0.36	<0.13	<0.13	0.18	—	—
	USA2	5X	0.415	1.00	0.35	0.089	0.14	0.22	0.10	—
Azoxystrobin	JP	5X	0.06	1.00	0.25	0.070	0.08	0.15	0.23	0.23
Carbaryl	JP	5X	0.03	1.00	0.43	<0.17	<0.17	0.14	—	—
Clethodimf)	USA3	1X	0.183	1.00	0.35	0.077	0.071	0.082	—	—
	USA3	5X	0.893	1.00	0.41	0.10	0.10	0.10	0.043	—
	USA3	1X	0.653	1.00	0.63	0.12	0.12	0.13	—	—
	USA1	5X	3.12	1.00	0.69	0.14	0.11	0.12	0.028	—
	USA3	5X	0.044	1.00	0.52	0.14	0.11	0.14	<0.11	—
Carbofurang)	USA1	5X	0.099	1.00	0.26	<0.091	<0.091	0.11	<0.091	—
	USA2	5X	0.182	1.00	0.40	0.038	0.033	0.049	<0.049	—
Phosphamidonh)	USA2	5X	0.084	1.00	0.083	<0.024	0.060	0.048	—	—
	USA2	5X	0.064	1.00	0.047	<0.031	<0.047	<0.047	—	—
Oxadixyl	JP	5X	0.16	1.00	0.36	0.094	0.10	0.12	0.21	—
Dimethoate	IA	5X	0.185	1.00	0.24	0.086	<0.027	<0.027	0.016	—
Glyphosatei)	USA1	1X	72	1.00	0.26	0.056	0.069	0.026	0.083	—
Paraquat	USA2	5X	0.050	1.00	0.29	0.074	0.21	0.14	—	—
	USA3	5X	0.094	1.00	0.32	0.10	0.42	0.12	—	—
Diquat	USA2	5X	0.045	1.00	0.31	0.091	0.23	0.10	—	—
	USA3	5X	0.077	1.00	0.40	0.13	0.24	0.049	—	—
Dinotefuran	JP	5X	1.36	1.00	0.40	0.10	0.10	0.092	0.26	0.32

^a) Residue concentration in a product (mg/kg)/residue concentration in soybean (mg/kg). ^b) Field trial test site: JP; Japan Plant Protection Association, Ibaraki, Japan, USA1; Mid-South Ag Research, Crittenden County, Proctor, AR, USA, USA2, 3; Bennett Ag Research, Jefferson County, Richland, Iowa, USA. ^c) Aplication rate as a relative ratio of the GAP maximum rate. ^d) cis+trans. ^e) PSSO₂+POSO₂. ^f) CLSO₂. ^g) Carbofuran+3-OH-carbofuran. ^h) Phosphamidon (cis+trans)+N-desethyl phosphamidon (cis). ⁱ⁾ Glyphosate+AMPA.

1.4. Edamame (immature soybean) sample⁷⁾

The effects of processing and cooking on the residue levels of benomyl, which was applied preharvest and determined to be as carbendazim (degradation product of benomyl), were investigated. In the case of edamame, the RARs (%) of whole raw seeds as the edible portions were reduced to 1% after removing the pods. Although more than 70% of the initial residues were eliminated by transfer to the boiling wastewater, the RARs in boiled seeds increased slightly to 2.6–3.6%. This increase indicated that part of the residue was transferred from the pod to the seed portion during the boiling process. Removing edamame pods effectively reduced the benomyl residue. The benomyl residue concentrations were determined using an LC-MS/MS.

1.5. Sesame sample¹¹⁾

There are differences in the method of producing sesame oil between Japan and other countries. In Japan, roasted sesame seeds are used to produce sesame oil, and thus Japanese sesame oils are commonly dark brown in color. In order to verify these differences, sesame oils were prepared using raw sesame and roasted sesame. The effects of processing on the pesticide residue levels in sesame samples were investigated for 3 pesticides (A: pyrethroid, B: fungicide, C: insecticide). A and B are not used for this application in Japan. The RARs of all investigated pesticides in roasted sesame were A: 50%; B: 69%; and C: 3%. The value of C was especially low because it is thermally decomposable. These values varied among pesticides. The Pfs in roasted sesame, sesame oil (roasted), and sesame oil (raw) are shown in Table 4. The pesticide residue concentrations were determined using a GC-MS, an LC-MS, or an LC-MS/MS.

Table 4. Processing factors (Pf) of sesame product samples

Pesticide	log Powb)	Concentration (mg/kg)	Pfa)
		Raw sesame	Raw sesame	Roasted sesame	Sesame oil (roasted)	Sesame oil (raw)
Pyrethroid Ac)	6.1	0.389	1.00	0.52	0.32	1.72
Fungicide Bd)	3.0	13.6	1.00	0.72	0.82	1.74
Insecticide Cd)	0.704	9.91	1.00	0.028	0.01	0.032

^a) Residue concentration in a product (mg/kg)/residue concentration in raw sesame (mg/kg). Field trial test site: Japan Plant protection Association, Miyazaki, Japan. ^b) The data were taken from the Pesticide Manual, 14th edition. ^c) Application rate (5X) as a relative ratio of the GAP maximum rate for pre-harvest treatment. ^d) No application in Japan (reference data).

2. Distribution of pesticide residues in the peel and pulp of watermelon, melon, and kiwi fruit samples

Conventionally, the MRLs for RACs were basically adopted for the edible food portions in Japan, which was not the case in other countries. This inconsistency in MRL standards among countries was regarded as a trade obstacle for RACs. Therefore, international harmonization has been attempted to remove such obstructions. Recently, there have been some changes in analytical portions of RAC samples. In consideration of these circumstances, we conducted a study on the pesticide residue concentrations of whole watermelon, melon, and kiwi fruits, as well as distributions in the pulp and peel, and analyzed the obtained data. In the case of melon and kiwi fruit samples, the surface shape of the plant body appears to be a rate-limiting factor for the RARs because the abundance ratio of pulp could be influenced by the peel shape.

2.1. Watermelon sample¹³⁾

As shown in Fig. 4, the pesticide residue levels in watermelon samples were investigated for 16 pesticides (large type samples) and 18 pesticides (small type samples). The pesticide residue concentration in small type samples was higher than that in large type samples. The peel of large type samples was thicker than that of small type samples. The RARs in the pulp tended to be higher in the case of lower log P_ow pesticides with systemic action, although the ratio varied among chemicals. There was no noticeable difference in RARs between large and small type samples, but the values in the small type samples tended to be slightly higher than those in large type samples since the peel of the small type is thinner than that of the large type.

Fig. 4. Distribution of pesticide residues in the pulp and peel of the watermelon sample (large and small types). The upper graph shows data for the large type sample. The lower graph shows data for the small type sample. The hatching of log P_ow indicates that a pesticide has systemic action. The data are taken from the Pesticide Manual, 13th edition.

2.2. Melon sample¹³⁾

As shown in Fig. 5, the pesticide residue levels in melon samples were investigated for 18 pesticides (netlike skin type sample) and 16 pesticides (non-netlike skin type sample). The pesticide residue concentration in netlike skin type samples was higher than that in non-netlike skin type samples. The RARs of lower log P_ow pesticides with systemic action in the pulp tended to be higher than those of higher log P_ow pesticides. This is similar to the result for watermelon. There was no noticeable difference in RARs between the two types. In the case of netlike skin type samples, the migration of high log P_ow pesticides from the peel to the pulp after application might be inhibited.

Fig. 5. Distribution of pesticide residues in the pulp and peel of the melon sample (netlike and non-netlike skin types). The upper graph shows data for the netlike skin type sample. The lower graph shows data for the non-netlike skin type sample. The hatching of log P_ow indicates that a pesticide has systemic action. The data are taken from the Pesticide Manual, 13th edition.

2.3. Kiwi fruit sample¹³⁾

As shown in Fig. 6, the pesticide residue levels in kiwi fruit samples were investigated for 15 pesticides. Both the RARs (0.4–2.7%) and Pfs (0.0078–0.041) of all pesticides in the pulp were consistently low, suggesting that the migration of these pesticides from the peel to the pulp could be inhibited.

Fig. 6. Distribution of pesticide residues in the pulp and peel of the kiwi fruit sample. The lower graph shows an extended version of the upper graph. The hatching of log P_ow indicates that a pesticide has systemic action. The data are taken from the Pesticide Manual, 13th edition.

3. Considerations for extraction in pesticide residue analysis

3.1. Extraction efficiencies of pesticide residue analysis in rice and wheat samples¹⁴⁾

The effects of water-soaking pretreatment of powdered dry cereal using brown rice and wheat samples on extraction efficiency were examined. The recovery of four pesticide residues (carbaryl, malathion, diazinon, and fenitrothion) detected in brown rice by acetone extraction without water-soaking was as low as 6–68% of recovery after a 30 min water-soaking pretreatment. The optimal soaking time in water was 15 to 30 min for the powdered cereal samples. The relative recovery of spiked pesticides and residual pesticides was not always parallel between the two extraction conditions, i.e., acetone extraction with and without a water-soaking pretreatment. Pretreatment produced higher measured values in many cases but also resulted in lower values for some pesticides, such as malathion and diazinon. Soaking dry samples in water for 1.5–30 min as a pretreatment will result in a good residual pesticide extraction.

3.2. Liquid–liquid extraction versus a macroporous diatomaceous earth column

Recently, a macroporous diatomaceous earth (MDE) column was used as an alternative to liquid–liquid extraction (LLE). Both technical capabilities were compared with the n-hexane of recoveries for 206 pesticides. The recoveries are shown in Fig. 7. Using LLE, the pesticides with high water solubility tended to have low recoveries. This is because the polarity of n-hexane is low for polar pesticides. On the other hand, the pesticides with low water solubility tended to have low recoveries using an MDE column. This is because pesticides in the aqueous phase retained on the surface of the MDE column were not well partitioned. Three hydrolyzed pesticides (dithianon, phoxim, and fluoroimide) showed no recoveries, since the MDE column is slightly alkaline.

Fig. 7. Comparison of recoveries via liquid–liquid extraction and a macroporous diatomaceous earth column. The upper graph shows the results using liquid–liquid extraction. The lower graph shows the results using a macroporous diatomaceous earth column. The notes in the lower graph indicate three alkaline hydrolyzed pesticides: a) dithianon, b) phoxim, and c) fluoroimide.

4. Decrease in the matrix-enhancement effect on GC-MS determination¹⁶⁾

The matrix removal ability of several new types of solid-phase extraction (SPE) columns was tested. E-HyCu is made of a new type of material, chemically modified carbon fibers, and can remove food components such as monoacylglycerols, tocopherols, and sterols. Z-Sep+ and Z-Sep/C18 are also new materials that contain zirconium dioxide. Z-Sep+ and Z-Sep/C18 are used for high lipid samples as the dispersive SPE in the QuEChERS method. We evaluated these sorbents as filled columns. In our previous studies, it was shown that when the multi-residue analytical method was applied, monoacylglycerols were the most significant components that caused a matrix-enhancement effect on pesticides in food when using a GC-MS. These new types of SPEs could remove not only monoacylglycerols but also fatty acids, tocopherols, flavonoids, and sterols. The matrix-enhancement effects were dramatically reduced in cases of approximately 260 pesticides spiked in brown rice extracts pretreated with these SPEs.

Conclusions

Several purification steps in pesticide residue analysis are no longer necessary, since the sensitivity of the instruments has significantly improved. However, problems with extraction efficiency remain, and problems with the matrix effects on measurement are more difficult to solve.

The basic data obtained from our studies should be helpful for the precise estimation of pesticide-residue exposure levels from food intake and for setting MRLs. Furthermore, several investigations in our studies have suggested important points to be checked in the extraction and purification process for analysis. Processing and cooking test models for rice, wheat, soybean, edamame, and sesame samples were established and verified on a laboratory scale. The Pfs of the edible portions of products from each sample were obtained. These data are useful for exposure estimation. Data on the concentration of pesticide residues in whole and various portions (pulp and peel) of watermelon, melon, and kiwi fruits could be utilized as the basic data to establish new MRLs in the future.

It is important to verify and recognize the key points to be checked in the process of pesticide residue analysis. In addition, special attention should be paid to warning points that could be easily overlooked. We hope these data from our studies will be helpful to colleagues involved in pesticide residue analysis.

Acknowledgment

This work was supported in part by the grant of Ministry of Health, Labor and Welfare of Japan, and Ministry of Agriculture, Forestry and Fisheries of Japan. I am grateful for comments and advice from Dr. Tamiji Sugiyama (Former professor, Meiji University), Dr. Yasuhiro Kato (Advisor, IET) and Dr. Takanori Harada (President, IET). I would like to thank Dr. Kuniyo Sugitate (Agilent Technologies, Inc.) for my co-researcher. Finally, I wish to thank officers and members in Chemistry and Study Management Division of the IET.

References

• 1) M. Saka: CHROMATOGRAPHY 33, 157–166 (2012) (in Japanese).
• 2) http://www.mhlw.go.jp/stf/seisakunitsuite/bunya/kenkou_iryou/shokuhin/zanryu/zanryu3/siken.html (Accessed 11 Jun., 2018).
• 3) M. Saka: J. Pestic. Sci. 35, 580–586 (2010) (in Japanese).
• 4) http://www.env.go.jp/chemi/anzen/lcms/ma01.pdf (Accessed 11 Jun., 2018).
• 5) M. Saka, K. Iijima, M. Nishida, Y. Koma, N. Hasegawa, K. Sato and Y. Kato: Food Hyg. Saf. Sci. 49, 141–149 (2008) (in Japanese).
• 6) M. Saka, K. Iijima, M. Nishida, M. Fujita and Y. Kato: Abstr. 5th European Pesticide Residue Workshop, p. 182, 2004.
• 7) M. Saka, K. Iijima, M. Nishida, Y. Koma, N. Hasegawa, K. Sato and Y. Kato: Food Hyg. Saf. Sci. 49, 150–159 (2008) (in Japanese).
• 8) M. Saka, N. Tomiyama, Y. Koma, K. Iijima and Y. Kato: J. Pestic. Sci. 41, 11–17 (2016) (in Japanese).
• 9) M. Saka, K. Iijima, M. Nishida, Y. Koma, N. Hasegawa, K. Sato and Y. Kato: Food Hyg. Saf. Sci. 49, 160–169 (2008) (in Japanese).
• 10) M. Saka, Y. Koma and K. Sato: Abstr. 35th Annu. Meeting Special Committee on Pesticide Residues Analysis of the Pestic. Sci. Soc. Jpn., 156–163, 2012 (in Japanese).
• 11) M. Saka, Y. Wakasone, Y. Koma and K. Sato: Abstr. 100th Annu. Meeting of Japanese Society for Food Hygiene and Safety, p. 45, 2010 (in Japanese).
• 12) Notification of Ministry of Health, Labor and Welfare of Japan, No. 499, Nov 29, 2005.
• 13) M. Saka, Y. Koma and K. Sato: Abstr. 37th Annu. Meeting Special Committee on Pesticide Residues Analysis of the Pestic. Sci. Soc. Jpn., 105–111, 2014 (in Japanese).
• 14) M. Saka, K. Iijima, Y. Koma, M. Fujita, Y. Odanaka and Y. Kato: J. Pestic. Sci. 31, 431–434 (2006) (in Japanese).
• 15) K. Iijima, M. Saka, Y. Odanaka and O. Matano: J. Pestic. Sci. 22, 17–26 (1997).
• 16) K. Sugitate and M. Saka: J. Pestic. Sci. 40, 87–91 (2015).
• 17) K. Sugitate, M. Saka, T. Serino, A. Toriba and K. Hayakawa: J. Agric. Food Chem. 60, 10226–10234 (2012).
• 18) K. Sugitate, S. Nakamura, N. Orikata, K. Mizukoshi, M. Nakamura, A. Toriba and K. Hayakawa: J. Pestic. Sci. 37, 156–163 (2015).