View stunning SlideShares in full-screen with the new iOS app!Introducing SlideShare for AndroidExplore all your favorite topics in the SlideShare appGet the SlideShare app to Save for Later — even offline
View stunning SlideShares in full-screen with the new Android app!View stunning SlideShares in full-screen with the new iOS app!
YANG ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002 1241AGRICULTURAL MATERIALSDetermination of Phosphorus in Fertilizers by InductivelyCoupled Plasma Atomic Emission SpectrometryWEI MIN YANG, RHONDA L. BOLES, and THOMAS P. MAWHINNEY1University of Missouri-Columbia, Experiment Station Chemical Laboratories, Rm 4, Agriculture Bldg, Columbia, MO 65211An inductively coupled plasma atomic emission tion curve, the weight of the fertilizer samples is adjusted tospectrometry (ICP-AES) method was developed for control final sample phosphate concentrations.the determination of phosphorus in fertilizers. To- A major advantage of inductively coupled plasma atomictal phosphorus, direct extraction available phos- emission spectrometry (ICP-AES) is its ability to determinephorus (EDTA), and water-soluble phosphorus, re- several elements simultaneously in a wide range of concentra-ported as phosphorus pentoxide (P2O5), in tions. For determination of phosphorus, the sensitivity of the15 Magruder check fertilizers were measured by ICP-AES analysis method exceeds that of the AOAC officialICP-AES, and the results were compared with methods; in addition, the ICP-AES has a calibration curve of athose obtained by the AOAC official method. very wide linear range. As a result, no sample weighing adjust-Five analytical wavelengths of phosphorus, ment is required for the ICP-AES method. Because of these fea-177.499, 178.287, 213.618, 214.914, and 253.565 nm, tures, the determination of phosphorus by the ICP-AES methodwere tested for the determination of phosphorus in has attracted the attention of several scientists.fertilizers, and their detection limits were obtained. Hamalova et al. (5) determined phosphorus, potassium,Acid effects of perchloric acid and possible matrix and magnesium in 12 fertilizers by ICP-AES. For phosphorus,effects of aluminum, calcium, magnesium, potas- the results of ICP-AES were compared with those obtained bysium, and sodium were negligible for phosphorus the standard gravimetry method. Generally, precision and ac-determination. Wavelength 213.618 nm was the curacy of the ICP-AES method were satisfactory. Ardis andbest analytical wavelength for phosphorus deter- Baker (6) used an ICP spectrometer to monitor fertilizer plantmination by all 3 sample preparation methods for effluents for phosphorus, sulfur, and metals. The detectionthe selected Magruder fertilizers. The results dem- limit for phosphorus was 0.08 µg/mL at 213.618 nm.onstrated that the accuracy and precision of the Jones (7, 8) determined the concentrations of major, micro,ICP-AES method were comparable with those of and trace elements by ICP-AES in fertilizer standards. It wasthe official methods. determined that analysis by ICP-AES of fertilizer materials with wide ranges of elemental contents is a rapid technique that can yield results comparable with those obtained by n fertilizers, phosphorus is one of the major nutrients that AOAC protocols.I is most commonly determined. Accurate analysis of fertil- izers is necessary both for the quality control of their pro-duction and for determining their effective use in the field. This study was undertaken to standardize the ICP-AES method for determination of phosphorus in several sample preparation methods in the Magruder fertilizer check samplesPresently, the gravimetric determination with quimociac re- by the use of the same calibration curve.agent and the spectrophotometric molybdovanadophosphate METHODprocedures are the AOAC official reference methods for de-termination of phosphorus in fertilizers (1–3). The gravimetric Apparatusprocedure is both tedious and time-consuming. The spectro-photometric molybdovandate phosphate method, on the other (a) Spectrometer.—Atomic emission measurements ofhand, measures only orthophosphate (4). Because of this, fer- phosphorus were made with an ARL 3410+ sequential ICPtilizer samples that possess phosphate in other forms must be spectrometer (Fisons, Dearborn, MI) with a minitorch.converted to orthophosphate before determination. Further- (b) Nebulizer and pump.—The sample solutions weremore, to control the citrate interference which is used for di- nebulized by a pneumatic nebulizer (Precision Glassblowing,rect available extraction, appropriate matrix match is neces- Englewood, CO) with a Gilson peristaltic pumpsary for the calibration standards. Finally, to ensure that the (Worthington, OH). The main operating conditions are listedmeasured signals fall within the working range of the calibra- in Table 1. Reagents Received October 25, 2001. Accepted by EB June 5, 2002. 1 Author to whom correspondence should be addressed; e-mail: (a) Standard solution of phosphorus(1000 mg/L).—Pre-mawhinneyT@missouri.edu. pared from dried (2 h at 110°C) KH2PO4 (NIST Standard Ref-
1242 YANG ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002Table 1. ICP-AES instrumental parameters and previously been folded, rinsed, and dried on aoperating conditions polymethylpentene filter. Insert each filter into a numberedInstrumentation Operating condition 500 mL volumetric flask for collection of filtrate. Wash ac- tively until ca 250 mL filtrate has been collected. Bring filtrate to final volume and mix. The phosphorus concentration in theInstrument ARL 3410+ sequential ICP spectrometer dissolved samples of 9703 and 9504 are ca 1.0 and 4.0 ppm,RF generator frequency 27.12 MHz respectively, and run directly by ICP-AES without dilution.Plasma torch Minitorch Dilute remaining 13 Magruder dissolved fertilizers samplesPump Gilson peristaltic pump 10-fold before ICP-AES analysis. The phosphorus concentra-Nebulizer Meinhard pneumatic nebulizer (Type K) tions in these diluted solutions range from 5.5 to 41.0 ppm.Forward power 650 W Acid and Matrix EffectsReflected power <3 W Phosphorus solutions (5 mg/L) with different concentra-Viewing height 9.0 mm above load coil tions of perchloric acid were prepared for acid effects study.Coolant gas flow rate 10.5 L/min The final concentration of perchloric acid varied from 0.0 andAuxiliary gas flow rate 0.8 L/min 0.0001 mol/L to as high as 2.0 mol/L. Phosphorus solutionsCarrier gas flow rate 0.8 L/min (5 mg/L), mixed with different concentrations of possible ma-Integration time 1s trix elements (Al, Ca, Mg, K, Na), were used for the matrix ef-Sample uptake rate 2.5 mL/min fects investigations. The individual concentrations of these 5 matrix elements were 50, 100, 150, and 200 mg/L. All the ICP-AES measurements for acid effects and matrix effects were triplicate determinations, and relative standard devia- tions (RSD) were within ± 3.0%. Results and Discussionerence Material), and 20 mL perchloric acid was added beforediluting to volume. Acid Effects (b) Calibration solutions.—Prepared to contain 0.0, 1.0, 5.0,10.0, 50.0, and 100.0 mg P/L in 4% HClO4. The concentrations After sample preparation as described in the Method sec-of phosphorus in the samples were calculated from a linear re- tion, the solution of total P2O5 is acidic, containing aboutgression equation of the calibration curve. Required correlation 0.4 mol/L HClO4, but the solutions of direct available P2O5coefficient of this equation was set to not less than 0.9990. and water-soluble P2O5 are both neutral. As a result, it was necessary to investigate the acid effects of HClO4 on the emis- Sample Preparation sion signal of phosphorus. (a) Total P2O5.—Weigh triplicate 0.50 g subsamples into As shown in Table 2 , the emission intensities ratio for all250 mL boiling flasks. Add 15 mL HNO3 and 10 mL HClO4. 5 phosphorus analytical wavelengths varied from 94 to 105%,Boil gently on medium-heat hot plate for ca 1 h until solution is while being measured in the presence of 0.0001–2.0 mol/Lcolorless, or nearly so, and dense white fumes of HClO4 appear HClO4. Because the dissolved samples were diluted 10-foldin the flask. After digestion, cool to room temperature, add ca before determination, 4% (v/v) HClO4 was used as the diluent50 mL deionized water, and boil gently for ca 15 min. Remove in all cases. As a result, the negligible influence of acid effectssample from hot plate, cool to room temperature, dilute to vol- on phosphorus determination was further minimized in allume, stir, and stopper. Dilute dissolved samples 10-fold before ICP-AES assays.ICP-AES determination. The phosphorus concentrations in the Matrix Effectdiluted solutions range from 2.5 to 45.0 ppm. (b) Direct extraction available P2O5 (EDTA).—Weigh Five elements, K, Na, Mg, Al, Ca, were selected as matrixtriplicate 0.50 g subsamples into 250 mL boiling flasks. Add elements in this investigation. Table 3 shows that the phospho-100 mL ammonium citrate–EDTA solution that has been rus emission intensity ratio at various wavelengths was withinpreheated to 65°C. Place flask in 65°C shaker bath. Close the range of 99–104%, clearly demonstrating that matrix inter-flask tightly with rubber stopper. Vigorously shake flask for ference was minimal for analysis of phosphorus. Notably, con-1 h at 65°C. Shaking should be vigorous enough to ensure centrations of those elements in diluted dissolved samples werewetting of sample and flask walls. Remove samples from well below the matrix concentrations presented in Table 3.bath, and immediately cool to room temperature with 4°C dis- Therefore, matrix effects can be considered negligible fromtilled water. Dilute to volume, and mix thoroughly. Dilute dis- these elements when phosphorus in fertilizers is determined.solved samples 10-fold before ICP-AES determination. The Detection Limitphosphorus concentrations in the diluted solutions range from2.0 to 45.0 ppm. A blank solution (8%, v/v, HClO4) and a solution of (c) Water soluble P2O5.—Weigh triplicate 1.00 g P (5 mg/L) were used to measure the detection limit. Detectionsubsamples on 11 cm Whatman No.4 filter circles that have limit (3σ) of the wavelengths 177.499, 178.287, 213.618,
1244 YANG ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002Table 4. Linear regression analysis of ICP phosphorus results with average values of Magruder fertilizers Total P2O5 Direct available P2O5 Water-soluble P2O5Wavelength, nm Regression line (a Regression line (a Regression line (a177.499 y = 1.007x – 0.2529 0.9998 y = 1.010x – 0.5845 0.9994 y = 0.8678x + 0.5807 0.9022178.287 y = 1.013x – 0.2403 0.9997 y = 1.004x + 0.0604 0.9997 y = 0.8797x + 0.3319 0.9082213.618 y = 1.001x – 0.1247 0.9999 y = 1.002x – 0.0864 0.9999 y = 0.8872x + 0.9076 0.9084214.914 y = 1.010x – 0.1571 0.9998 y = 1.008x – 0.3590 0.9998 y = 0.8843x + 0.6544 0.9150253.565 y = 0.9913x – 0.0829 0.9948 y = 0.9875x – 0.5438 0.9984 y = 0.9849x + 0.3170 0.8987a Linear regression correlation coefficients.214.914, and 253.565 nm, which were calculated from the (mono-ammonium phosphate 9803B). The NPK values instandard deviations of 11 measurements of the blank solution, Table 5 represent the percentages of nitrogen, available phos-were 40, 42, 23, 55, and 88 µg/L, respectively. phorus (P2O5), and soluble potassium (K2O). Except for 9811A, which is clear liquid fertilizer, the other 14 Magruder Magruder Fertilizer Analysis by ICP-AES samples are solid. Triple superphosphate sample In 1922, E.W. Magruder, chief chemist of the F.S. Royster 0998A (0-46-0) contains only one nutrient: phosphorus; theGuano Co. (Norfolk, VA) began the fertilizer check sample remaining 14 Magruder samples contain 2 or 3 nutrients: ni-program that bears his name today. Magruder fertilizer check trogen, phosphorus, potassium. With the 213.618 nm wave-sample programs are widely used to ensure consistency length used to determine P2O5, the ICP-AES recoveries wereamong laboratories and methods. The results can identify sta- compared with the Magruder average standard values (Ta-tistically different repeatabilities or reproducibilities among ble 5). The recovery was defined as the ratio of the ICP resultsdifferent methods or among laboratories using the same to the Magruder average.method. In this study, 15 Magruder fertilizer check samples The values of total P2O5 and direct available P2O5 mea-were selected to study the performance of ICP-AES for the de- sured by ICP-AES compared closely with the reportedtermination of phosphorus. Magruder average. Two notable exceptions were observed, Generally, total P2O5, direct extraction available P2O5 both for water-soluble P2O5 analyses, where significant dis-(EDTA), modified extraction available P2O5 (non-triple crepancies between the values of ICP results and Magrudersuperphosphate), and water-soluble P2O5 are often deter- average were determined. Specifically, these anomalies in-mined in laboratories. Because the matrixes of direct extrac- volved Magruder samples 9504 and 9811A. The remarkablytion available P2O5 (EDTA) and modified extraction available large standard deviation of the Magruder average on the solidP2O5 are similar, this investigation used the sample prepara- fertilizer sample 9504 makes the low recovery understand-tion procedure of direct extraction available P2O5 (EDTA). able. This is in contrast to the high recovery (329.4%) of phos- Table 4 summarizes the linear regression analysis of the phorus by ICP-AES for the liquid Magruder fertilizer sam-ICP results with the Magruder average at various wave- ple 9811A, which is totally water-soluble. This is reasonablylengths. A main focus of this study was to determine which explained by the fact that all phosphorus in the solution iswavelength for phosphorus was suitable. For total P2O5 and determined by the ICP spectrometer, whether it is present in itsdirect available P2O5, the overall agreement for phosphorus ortho-, poly-, organic-, or low-oxide forms, and the Magruderwas excellent over the wide concentration range found within average presented in Table 5 reports only the orthophosphatethe 15 Magruder fertilizers. The intercept was small and not concentration in the sample. The determined total P2O5 ofsignificantly different from zero, and the slope was very close 33.47% in this sample by ICP-AES is further substantiated byto 1.0. This demonstrated that the ICP-AES method possessesneither constant nor proportional systematic errors compared the fact that sample 9811A’s content as determined by totalwith the Magruder average. The best linearity was achieved at P2O5 and direct available P2O5 are in agreement. If sam-213.618 nm. In addition, 213.618 nm is the most sensitive ple 9811A is excluded from the linear regression (Table 4), thewavelength for phosphorus. Therefore, among the 5 analytical regression equation of water-soluble P2O5 will be y = 0.9733x +wavelengths tested, wavelength 213.618 nm was the optimal 0.7855 at 213.618 nm, with a correlation coefficient of 0.9983.analytical wavelength for determining phosphorus in fertiliz- Analysis of total phosphorus by ICP-AES is simpler anders by ICP-AES. faster than the traditional procedures using digestion and spec- To represent the assorted commercial fertilizer used in the trophotometric determination. Significant advantages of theagriculture, the contents of the total phosphorus (P2O5) in the ICP-AES method are also demonstrated by its wide dynamicselected Magruder fertilizer samples varies from several per- range of calibration and high sensitivity. Furthermore, by thecent, as in organic fertilizers 9703 and 9504, to as high as 52% ICP-AES method, the same amount of fertilizer sample can be
1246 YANG ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002weighed regardless of the phosphorus concentration, and no general, the precision and accuracy of ICP-AES determina-spike experiment is required for low phosphorus content. tions were close to the average values of Magruder standard The precision and accuracy of the methodology developed check fertilizer samples. Care must be taken in the determina-were tested by analyzing total P2O5 and direct available P2O5. tion of total P2O5 in fertilizer samples in which phosphorusFor short-term precision testing (7 consecutive determinations), may exist in several forms. By the ICP-AES method, all phos- phate forms are analyzed, yielding a true total P2O5 value.the RSD of total P2O5 and direct available P2O5 were in the rangeof 0.6–2.2 and 0.8–2.6%, respectively. For long-term precision Referencestesting (7 determinations over 7 consecutive days), the RSDs oftotal P2O5 and direct available P2O5 were in the range of (1) Official Methods of Analysis (1990) 15th Ed., AOAC,0.7–3.5 and 0.8–3.9%, respectively. The precision of total P2O5 Arlington, VA, sec. 962.02and direct available P2O5 determination was also excellent. (2) Dahlgren, S.W. (1962) Z. Anal. Chem. 189, 243–256 (3) Official Methods of Analysis (1990) 15th Ed., AOAC,Conclusions Arlington, VA, secs. 958.01, 963.03(a) (4) Fujiwara, K., Mignardi, M.A., Petrucci, G., Smith, B.W., & Winefordner, J.D. (1989) Spectrosc. Lett. 22, 1125–1140 The ICP-AES method enables the determination of total (5) Hamalova, M., Hodslavska, J., Janos, P., & Kanicky, V.J.P2O5, direct extraction available P2O5 (EDTA), and wa- (1997) J. AOAC Int. 80, 1151–1155ter-soluble P2O5 in fertilizers by the same calibration curve. (6) Ardis, J.D., & Baker, A.M. (1997) J. AOAC Int. 72, 857–859Wavelength 213.618 nm proved to be the best analytical (7) Jones, J.B. (1982) J. Assoc. Off. Anal. Chem. 65, 781–785wavelength for phosphorus determination in fertilizers. In (8) Jones, J.B. (1983) Spectrochim. Acta 38B, 271–276