Talanta 51 (2000) 599604

FT-Raman spectroscopy analytical tool for routineanalysis of diazinon pesticide formulations

Stavroula G. Skoulika, Constantinos A. Georgiou *,1, Moschos G. PolissiouChemistry Laboratory, Agricultural Uni6ersity of Athens, 75 Iera Odos, 11855 Athens, Greece

Received 7 July 1999; received in revised form 23 November 1999; accepted 25 November 1999

Abstract

FT-Raman spectroscopy based on band intensity and band area measurements, was used for the quantitativedetermination of diazinon in pesticide formulations. Bands at 554, 604, 631, 1562 and 2971 cm1 were used forcalibration. Spectra were acquired by averaging 100 scans at a resolution of 4 cm1. Calibration curves were linear(correlation coefficients, 0.9920.9992 and 0.990.999 for band intensity and band area measurements, respectively)in the range of 0.23.5 M for 554 and 2971 cm1, 0.33.5 M for 604 cm1, 0.63.5 M for 1562 cm1 and 1.03.5M for 631 cm1 bands. Normalization of calibration curves against the 802 cm1 cyclohexane band improved theirlong term stability and minimized the effect of laser beam power fluctuations. No interference was found bycommonly used surfactants and the proposed method was applied to the analysis of diazinon formulations. Resultsobtained compare well as indicated by the t-test, with those obtained by the HPLC reference method. Precisionranged between 0.27.8 and 0.17.2% RSD, (n4) for band intensity and band area measurements, respectively. Theproposed method is rapid, simple and safe, as toxic samples are analyzed as received without sample pre-treatment,permiting the routine analysis of pesticides formulations. 2000 Elsevier Science B.V. All rights reserved.

Keywords: Diazinon; FT-Raman; Pesticide formulations; Determination

www.elsevier.com:locate:talanta

1. Introduction

Diazinon (O,O-diethyl O-6-methyl-2-1-methyl-ethyl-4-pyrimidinyl phosphorothioate) is a non-systemic insecticide and acaricide with contact,stomach and respiratory action, that is used forthe control of sucking and chewing insects and

mites on a very wide range of crops. It is alsoused as a veterinary ectoparasiticide [1]. Severalmethods based on non-aqueous titrimetry [24],gas liquid chromatography [5,6] and high perfor-mance liquid chromatography [7] have been usedfor the quantitative analysis of diazinon in pesti-cide formulations. These methods are time con-suming and require extensive manual handling oftoxic samples with various solvents. As diazinonis a toxic cholinesterase inhibitor, alternative fastanalytical methods requiring minimal sample han-dling are highly desirable.

* Corresponding author. Tel.: 30-1-5294248; fax: 30-1-5294265.

E-mail address: cag@aua.gr (C.A. Georgiou)1 URL: http:::www.aua.gz:georgiou

0039-9140:00:$ - see front matter 2000 Elsevier Science B.V. All rights reserved.

PII: S0 039 -9140 (99 )00336 -7

S.G. Skoulika et al. : Talanta 51 (2000) 599604600

In this study, a novel FT-Raman method fordiazinon determination is presented and evaluatedin the analysis of pesticide formulations. Thiswork is part of an ongoing project in our labora-tory concerning pesticide formulation analysis [8].

2. Experimental

2.1. Sample preparation and chemicals

Diazinon of 98.7% purity, xylene of technicalgrade, the Xelmadix 60 EC diazinon formulationand two proprietary surfactants were a kind offerof Xelafarm, Hellas. The analytical diazinon stan-dard of certified 99.8% purity was a kind offer ofCiba Geigy, Hellas. Benzene pro analysi was pur-chased from Merck. Amok 60 EC, Efdiazon 60EC and Diazol 60 EC formulations were a kindoffer of Agroza Hellas, Eythimiadis Hellas andAlfa Georgika Efodia Hellas, respectively. Allchemicals were used without further purification.While preparing standards and handling formula-tions, extreme care should be exercised to avoiddiazinon contact with the skin and eyes, due todiazinon toxicity. Diazinon that is slightly solublein water at 20C (60 mg l1) is toxic, the acuteoral LD50 for rats being 1250, for mice being80135 and for guinea-pigs being 250650 mgkg1 [1]. Standard diazinon solutions in the rangeof 0.23.5 M were prepared by weighing theappropriate amount of the 98.7% diazinon anddissolving in xylene.

2.2. Apparatus

FT-Raman spectra were recorded with a Nico-let 750 FT-Raman spectrometer equipped with aNd:YAG laser source that emits at 1064 nm. A

CaF2 beamsplitter, an IndiumGalium Arsenide(InGaAs) detector and 180 backscattering ge-ometry are used in the spectrometer. A motorizedpositioner focuses the laser beam to the sampleand a manual side-to-side adjuster allows sampleadjustment for maximum optical efficiency. Toensure that the spectrometer is fine tuned and thedetector signal maximized, an optical bench align-ment was performed before each batch of mea-surements. Sample cells used were cut to 6 cmfrom Wilmad WG-5M NMR tubes of 4.97 mmouter diameter and 0.38 mm wall thickness. Spec-tra were accumulated from 100 scans collectedduring 3 min at a resolution of 4 cm1.

High performance liquid chromatographicanalysis was carried out using a Jasco PU-98Pump and a Jasco UV-970 UV:VIS Detectorequipped with a Waters Spherisorb S5 ODS24.6250 mm column. The flow rate was 1.0 mlmin1 and detection was at 254 nm.

3. Results and discussion

3.1. Raman spectrum of diazinon

The Raman spectrum of diazinon has beenpreviously reported using a Bruker FT-Ramanspectrometer with 300400 mW Nd:YAG laserexcitation [9]. Scattering frequencies are summa-rized in Table 1 where the classification to strong,medium, weak and very weak bands is based onthe % relative intensities compared to the 802cm1 cyclohexane band that was measured rightafter the acquisition of diazinon spectra retainingthe same experimental parameters (laser power,sample holder position, resolution and number ofscans).

Table 1Diazinon Raman bands (cm1)a

2927 (65) 2971 (38) 2873 (22)Strong130 (11)1101 (10)Medium 1456 (11) 554 (12)992 (17)

227 (6.6)1562 (5.2) 174 (9.9)Weak 3077 (8.5) 604 (7.9)631 (9.5)1380 (8.2)1296 (5.1)1587 (4.0) 506(3.9)817 (2.9)Very weak 865 (4.0)951 (3.5)2724 (4.5)

a % Relative intensities to the 802 cm1 cycloxehane band are shown in parentheses.

S.G. Skoulika et al. : Talanta 51 (2000) 599604 601

Fig. 1. The FT-Raman spectrum of diazinon recorded at 1.01, 0.81 and 0.57 W laser excitation power in the range of 32002800,17001200 and 700450 cm1 (A) before and (B) after normalization against the 802 cm1 cyclohexane band.

Molecules with PS bonds display a strongband in the 600700 cm1 region [10]. This bandis clearly distinguishable for compounds withoutbenzene rings, but several of the vibrations due tothe benzene ring overshadow it as in diazinonspectra, where two overlapping bands appear at631 and 604 cm1. The dCNC bands are ob-served at 604 and 554 cm1 [11,12] and the CCstretching mode at 1562 cm1 [13]. The intensebands appearing in the 28003100 cm1 rangeare due to aliphatic and aromatic CH stretchingmodes [10,1417].

Raman spectra of diazinon (99.8% purity) ac-quired during a time span of 15 days and under0.57, 0.81 and 1.01 W excitation intensities areshown in Fig. 1(A). Normalization was achievedby dividing with the Raman intensity of the 802cm1 cyclohexane band measured right after theacquisition of each diazinon spectrum, retaining

the same experimental parameters (laser excita-tion power, sample holder position, resolutionand number of scans). As is shown in Fig. 1(B),normalization eliminates differences. During di-azinon quantitative analysis, normalization wasachieved by dividing the slope of calibrationcurves by the intensity or the area of the cyclohex-ane 802 cm1 band and multiplication by 100.Normalization resulted in increased long termstability of calibration curves compensating alsodifferences in excitation intensities.

3.2. Quantitati6e analysis of diazinon formulations

For quantitative analysis of diazinon formula-tions the 554, 604, 631, 1562 and 2971 cm1

bands were used. These were chosen as the solvent(xylene) and commonly used surfactants do notinterfere spectrally. Furthermore these bands are

S.G. Skoulika et al. : Talanta 51 (2000) 599604602

characteristic for diazinon [9]. Band intensitiesand band areas were calculated using a two pointbaseline correction using the range 29582994cm1 for the 2971 cm1 band, 15481571 cm1

for the 1562 cm1 band, 620651 cm1 for the631 cm1 band, 593613 cm1 for the 604 cm1

band and 545564 cm1 for the 554 cm1 band.Calibration data based on the normalization

procedure at 1.01 W excitation intensity are pre-sented in Table 2. The calibration curves presentexcellent linearity with correlation coefficients inthe range of 0.9920.9992 and 0.990.999 forband intensity and band area measurements, re-spectively. Formulation analysis is achieved with-out sample dilution, minimizing manual handlingof toxic samples.

The linear range was found to be 0.23.5 M for

554 and 2971 cm1, 0.33.5 M for 604 cm1,0.63.5 M for 1562 cm1 and 1.03.5 M for 631cm1 bands. Detection limits are in the range of0.070.3 and 0.090.4 M for band intensity andband area measurements, respectively. The pro-posed method has no procedural steps and diazi-non is thermally stable under the laser beamradiation. Therefore, precision is affected only bysample positioning changes and possible cell vari-ations. Precision values presented in Table 2 weredetermined by removing the sample tube from thesample holder and then placing it again, afterturning by 60. For safety reasons, when openingthe sample compartment, an interlock mechanismcuts the laser beam. A short study was conductedto determine the time needed for laser powerstabilization after closing the sample compart-

Table 2Diazinon calibration data normalized against the 802 cm1 cyclohexane band

Equation of calibration graph Detection limita % RSDb (n4)ConcentrationBand (cm1) Correlation coeffi-range (M) (M)cient

Band intensity calibration cur6e2971 0.73.80.23.5 BI (0.990.2)(6.690.1)C 0.998 0.09

0.63.5 0.992 0.31562 3.07.8BI (0.490.1)(1.0090.06)

CBI (0.090.2)(1.8090.07)1.03.5631 0.995 0.3 0.95.8

CBI (0.0890.05)0.33.5 0.993 0.2 0.34.9604

(0.6790.03)C0.23.5 BI (0.0490.04)(1.890.02)554 0.24.70.9992 0.07

C

Band area calibration cur6e0.23.52971 BA (291)(23.090.7)C 0.995 0.1 0.74.00.63.5 BA (1.090.2)(1.890.1)1562 0.99 0.3 1.07.2

C1.03.5 BA (2.190.6)(4.190.2)C 0.993 0.4 1.24.06310.33.5 BA (2.790.1)(1.290.6)604 0.991 0.14.10.2

C0.23.5 BA (0.3390.07) 0.999 0.09 0.95.5554

(2.4090.04)C

a Calculated as three times the standard deviation of the intercept divided by the slope.b Calculated by changing sample position.

S.G. Skoulika et al. : Talanta 51 (2000) 599604 603

Table 3Results of the analysis of diazinon formulations by band intensity and area measurement

Diazinon concentrationa (M)9SD (n3) found during the analysis of four formulations accompanied by the experimentalt-valueb in parentheses

Band (cm1) AMOK XELMADIX EFDIAZON DIAZOL

Band intensity calibration cur6e2971 1.8590.05 (2.252)1.9390.05 (0.680) 1.990.1 (1.529) 1.8190.08 (1.891)

1.9890.05 (1.930) 1.990.1 (1.529)1562 1.9590.08 (1.050)2.0090.08 (1.934)1.9990.08 (1.471) 1.990.1 (1.529)1.9590.07 (0.980) 1.890.1 (1.699)631

2.0090.07 (2.205)604 2.090.1 (1.359) 2.0090.07 (0.238) 1.890.1 (1.699)1.8590.06 (1.917) 1.990.1 (1.529) 1.890.1 (1.699)554 1.9790.05 (2.039)

Band area calibration cur6e1.8390.09 (1.691) 1.990.1 (1.529) 1.890.1 (1.699)2971 1.890.1 (2.104)2.090.1 (1.359) 1.990.1 (1.529)1.9790.07 (1.470) 1.890.1 (1.699)1562

1.9590.05 (0.965) 1.9390.08 (1.261)631 1.890.1 (1.699)1.8590.08 (1.289)1.990.1 (0.340) 1.8890.08 (2.311)1.9690.04 (2.101) 1.890.1 (1.699)604

1.8690.07 (1.428) 1.990.1 (1.529)554 1.8190.07 (2.142)1.9890.07 (1.715)

a Diazinon concentration, 1.9190.01 M (AMOK), 1.9290.02 M (XELMADIX), 1.9990.02 M (EFDIAZON) and 1.9090.02M (DIAZOL) (n3) as determined using a HPLC reference method [8].

b ttheoritical2.776 for 95% confidence level and four degrees of freedom.

ment. In contrast to a previous study [18] where:2 min were required for laser power stabilization,we found that the analytical signal is stabilized rightafter closing the sample compartment. Precisionranged from 0.27.8 and 0.17.2% RSD (n4) forband intensity and band area measurements, re-spectively.

The use of the cyclohexane external standardcould lead to a Raman equivalent to the UV visibleabsorption coefficient [8,18,19]. Such coefficientsare the slopes of the calibration curves presented inTable 2.

Results of the analysis of diazinon formulations,by band intensity and band area measurements areshown in Table 3. As shown by the experimentalt-values, results obtained by the proposed methodare identical to those acquired through the HPLCreference method [8]. Relative differences betweenresults obtained through the FT-Raman and theHPLC method are in the range of 5.34.7 and5.54.2% for band intensity and band areameasurements, respectively. For quantitative anal-ysis, the use of band areas is believed to give moreaccurate results than the use of band intensities [20],as parameters influencing the shape and position ofRaman bands are more likely to affect the intensity

than the area of a band. However, as shown inTable 3 the use of intensity measurements givesresults of equal accuracy.

4. Conclusions

The proposed analytical method based on eitherband intensity or band area measurements is accu-rate and precise. The method is safe, as it eliminatessample pre-treatment and keeps handling of toxicsamples to a minimum. The proposed method isalso fast, requiring just 30 s for the analysis (usinga single band measurement) and simple so it can beperformed by minimally trained personnel. For thequantitative analysis of formulations, any of thestudied bands can be used. The FT-Raman methodcan be used to replace existing methods that requireextensive manual sample handling and is faster.The proposed method could also be used for theanalysis of diazinon formulations without evenopening the container, provided that the containeris made of glass or transparent plastic. The advan-tages of the FT-Raman technique can be furtherexploited by application to other pesticide formula-tions.

S.G. Skoulika et al. : Talanta 51 (2000) 599604604

Acknowledgements

We gratefully acknowledge support, in the formof a scholarship to S.G.S., from the Greek StateScholarship Foundation.

References

[1] C.D.S. Tomlin, in: C.D.S. Tomlin (Ed.), The PesticideManual, 11th ed., The British Crop Protection Council,London, 1997, p. 354.

[2] R. Suter, R. Delley, R. Meyer, Anal. Chem. 147 (1955)173.

[3] A. Margot, K.H. Stammabach, in: G. Zweing (Ed.),Pesticides, Plant Growth Regulators and Food Additives,vol. 2, Academic Press, New York, 1964, p. 109.

[4] G.R. Raw, in: G.R. Raw (Ed.), CIPAC Handbook, Ana-lysis of Technical and Formulated Pesticides, vol. 1, W.Heffer and Sons, Cambridge, 1970, p. 315.

[5] K. Helrich, Method 971.08, in: K. Helrich (Ed.), OfficialMethods of Analysis of the AOAC, 15th ed., AOACInternational, Arlington, VA, 1990, p. 222.

[6] P. Cunniff, Method 7.4.08, in: P. Cunniff (Ed.), OfficialMethods of Analysis of the AOAC, 16th ed., AOACInternational, Gaithersburg, MD, 1996, p. 40.

[7] C.J. Stafford, E.S. Greer, A.W. Burns, D.F. Hill, Diazi-non EPA-1 method, in: C.J. Stafford, E.S. Greer, A.W.

Burns, D.F. Hill (Eds.), US EPA Manual of ChemicalMethods for Pesticides and Devices, 2nd ed., AOACInternational, Arlington, VA, 1992.

[8] S.G. Skoulika, C.A. Georgiou, M.G. Polissiou, Appl.Spectrosc. 53 (1999) 1470.

[9] P.A. Tanner, K.H. Leung, Appl. Spectrosc. 50 (1996) 565.[10] M.L. Nicholas, D.L. Powell, T.R. Williams, S.R. Huff, J.

AOAC Int. 59 (1976) 1071.[11] M. Mylrajan, J. Mol. Struct. 348 (1995) 257.[12] S. Sanchez-Cortes, J.V. Garcia-Ramos, J. Raman Spec-

trosc. 21 (1990) 679.[13] G.A. Neville, H.F. Shurvell, J. Raman Spectrosc. 21

(1990) 9.[14] M.L. Nicholas, D.L. Powell, T.R. Williams, R.Q.

Thompson, N.H. Oliver, J. AOAC Int. 59 (1976) 1266.[15] J.C. Bowman, D.P. DeSalvo, Nicolet Application Note

no. 9250 (1992).[16] E.A. Cutmore, P.W. Skett, Spectrochim. Acta 49A (1993)

809.[17] H. Sadeghi-Jorabchi, R.H. Wilson, P.S. Belton, J.D. Ed-

wards-Webb, D.T. Coxon, Spectrochim. Acta 47A (1991)1449.

[18] C.J. Petty, P.J. Hendra, T. Jawhari, Spectrochim. Acta47A (1991) 1189.

[19] T. Jawhari, P.J. Hendra, H.A. Willis, M. Judkins, Spec-trochim. Acta 46A (1990) 161.

[20] D.R. Lombardi, C. Wang, B. Sun, A.W. Fountain III,T.J. Vickers, C.K. Mann, F.R. Reich, J.G. Douglas, B.A.Crawford, F.L. Kohlasch, Appl. Spectrosc. 48 (1994) 875.

.