Determination of ammonium, nitrate, and urea nitrogen in fertilizer by

Wei Wang , Yi Ding , Jeffrey L. Ullman , Richard F. Ambrose , Yuhui Wang , Xinshan Song , Zhimiao Zhao ... Millicent A. Firestone , Mark L. Dietz , Au...
0 downloads 0 Views 698KB Size
Anal. Chem. 1983, 55, 535-539

homogenous mixture. No attempt was made in this first feasibility study to minimize analysis time.

ACKNOWLEDGMENT The authors are gr'ateful to c' Jochum and de Koven for their helpful discussions and to Kontron Analytical for use of their UVIKON spectrophotometer.

535

(6) Kalivas, J. H.; Kowalski, B. R . Anal. Chem. 1981, 5 3 , 2207-2212. (7) Gerlach, R.; Kowalski, B. R. Anal. Chim. Acta 1982, 134, 119. (8) Moran, M.; Kowalski, B. R., Laboratory for Chemometrics, DepaiZment of Chemistry, University of Washington, unpublished work, June 1981. (9) Renoe, B. W.; O'Keefe, K. R.; Malmstadt, H. V. Anal. Chern. 1976, 4 8 , 661-666. (10) Danielson, J. D. S.; Brown, S. D.; Appellof. C. J.; Kowalski, B. R. Chem., Biomed. Envlron. Instrum. 1979, 9 , 29-47. (11) Naylor, T. H.; Balintfy, J. L.; Burldick, D. S.;Chu, K. "Computer Sirnulation Techniques"; Wiley: New York, 1966; Chapter 4.

LITERATURE CITED Enke, C. G. Science 1982, 215, 785-791. Niemczyk, M.; Ettinger, D. G. Appl. Spectrosc. 1978, 32, 450-453. Saxberg, Bo W. H.; Kowalski, B. R. Anal. Chem. 1979, 51, 1031- 1038. Jochum, c.; JoChUWl, p.; Kowalski, 6. R . Ana/. Chem. 1981, 53, 85-9 2. Kaiivas, J. H.; Kowalski, B. R . Anal. Chem. 1982, 5 4 , 560-565.

'r.

RECEIVED for review August 17, 1982. Accepted November 12, 1982. This work was supported in part by the Office of Naval Research. The authors also gratefully acknowledge supportfrom the , ~ ~ science t i ~ ~~ ~~ l ~under~ ~~~~t d NO. CHE-8004220.

Determination of Ammonium, Nitrate, and Urea Nitrogen in Molecular Absorption Spectrometry Fertilizer by Gas-Phase Vincent C. Anigbogu, Mark

L. Dietz, and Augusta Syty"

Department of Chemistry, Indiana University of Pennsylvania, Indiana, Pennsylvania 15705

Ammonium nltrogen In fertilizer is determlned by Injecting an aliquot of fertilizer solution Into strong base and measuring the translent absorbance exhibited by the evolved ammonia gas. Urea is converted to iimmonlum wlth the aid of urease and then determined as above. Nltrate Is reduced to nitrite by metallic cadmium and determlned by lnjectlng an allquot Into strong acid and measuring the absorbance of the evolved nitrogenous gases. Thie burner head of an atomic absorptlon spectrophotometer Is replaced wlth a flow-through absorptlon cell. A glass reactlon vessel is used to evolve the absorbing specles into tho gas phase. Ammonlum and urea can be determlned down to O.ID05 % and nitrate down to 0.015 % by weight In fertllirer, based on a 103 sample diluted to 500 mL. The reported methods are considerably faster than the corresponding AOAC methods, as they involve no dlstlllatlons. Data for flve commerciial fertllizers are given and the results agree with those obtallned by standard methods.

Nitrogen, phosphorus, and potassium are the main active ingredients of chemical fertilizers. Nitrogen is usually present as inorganic salts of aimmonia or nitrate, or as urea. Government regulation and quality control processes require the availability of rapid, convenient, and accurate methods for the analysis of fertilizers. The basis for the majority of N determination methods is the conversion of each form of N to ammonia. The determination of total N by the Kjeldahl method, or some modification thereof, involves digestion, distillation, and titration. Although the application of atomic absorption/emission spectrophotometry to the determination of fertilizer components such as K, as well as that of micronutrients such as IFe, Zn, Cu, and others, is quite common, the flame spectrophotometer cannot offer straightforward evaluation of nitrogen. However, the atomic absorption instrument can be modified very simply to allow molecular absorption measurements in the gas phase without the flame. This has made possible the determination of I-, Br- ( I ) , S032-(2-4), S2-(5),NO; (6),CN(7), and NH4+ (8) in a variety of materials. The objective of this report is to describe the use of gasphase molecular absorption spectrometry for the rapid de0003-2700/83/0355-0535$0 1.50/0

termination of ammonium, nitrate, and urea nitrogein in commercial water-soluble fertilizers.

EXPERIMENTAL SECTION Approach. After the fertilizer is dissolved in water, the ammonium ion concentration is determined by injecting aliquots of the solution into sodium hydroxide and measuring the absorbance of the evolved NH3 gas. For the determination of urea, the fertilizer solution is first treated with urease to convert urca to NH4+. The total lVH4+ is then measured as above, and the concentration of urea is deduced by subtraction. For the determination of nitrate, the fertilizer solution is passed through a Cd column to reduce NO3- to NOz- and the concentration of nitrite is then determined by injecting aliquots of the reduced solution into hydrochloric acid and measuring the absorbance of the evolved nitrogeneous gases which consist predominantly of nitrosyl chloride. Apparatus. All absorbance measurements were made on a Perkin-Elmer Model 460 atomic absorption spectrophotometer and recorded on a 10-mV strip chart recorder. The instrument was modified for molecular absorption measurements in the gas phase by removing the burner head from the nebulizer/burner and replacing it with a 15 cm long flow-through cell with quartz windows. The deuterium arc lamp, whose normal function in the instrument is to provide for background correction, served as the source of exciting radiation. For the measurement of evolved ammonia, the wavelength was set at 194 nm with a slit corresponding to a bandwidth of 2 inm, as recommended by Muroski and Syty (8). For the measuremlent of the gases evolved from nitrite (believed to be nitrosyl chloride (9) and oxides of nitirogen), absorbance was measured at 195 nm with a bandwidth of 0.7 nm, as suggested by Syty and Simmons (6). A specially designed glass reaction vessel was used for evolution of the absorbing species into the gas phase. The vessel is illlustrated in Figure 1 of ref 5. The cylindrical glass vessel has an internal volume of about 60 mL. A 6-mL aliquot of 10 N NaOH (in the case of ammonium) or of 8 N HC1 (in the case of nitrite) is introduced into the reaction vessel from a 25-ml buret attached to a side arm via a small sleeve of Tygon tubing. Ammoniumor nitrite-containing samples are injected into the basic or acidic solution through the rubber septum covering the injection port by means of a 1-mL Hamilton syringe. The evolved ammonia or nitrogenous gases are swept by a continuously flowing stream of nitrogen carrier gas from the reaction vessel to the flow-through absorption cell and then vented into the hood. The carrier ,gas 0 1963 American Chemical Society

~

t

536

ANALYTICAL CHEMISTRY, VOL. 55, NO. 3, MARCH 1983

enters the solution through a plain glass tip about 1 mm in internal diameter submerged in the acidic or basic reagent. The rate of nitrogen flow was regulated by a Model 620-PBV Matheson flowmeter and maintained at 1.50L/min during the determination of ammonium and at 1.45 L/min during the determination of nitrite. As the evolved gases pass through the absorption cell, a transient absorbance signal is recorded in the form of a peak which rises sharply to a maximum and falls off gradually. The appearance of typical recorded peaks is illustrated in Figure 1 of ref 6. The height of the recorded peak above the base line has been shown to be directly proportional to the concentration of the test species in the injected solution over a substantial concentration range (6, 8). As soon as the maximum of the absorbance peak has been recorded, the recorder is turned to “standby” and the reaction vessel is drained through the stopcock at the bottom of the vessel, without stopping the carrier gas flow. The vessel is then immediately refilled with a fresh aliquot of the NaOH or HC1 causing the recorder pen to return to the original base line. The recorder is turned back “on” and the apparatus is then ready to receive the next injection of sample. Emptying the reaction vessel before the entire absorbance signal has been recorded of course means that a large portion of the test gaseous species is not given a chance to become evolved from the reaction mixture and to pass through the absorption cell but instead, is discarded with the drained reaction mixture. This, however, causes no inaccuracy because it is the peak height of the absorbance signal which is measured and not the area under the entire absorbance trace. In fact, the ability to dispose of the reaction mixture in this rapid and convenient manner after recording just the desired parameter of the absorbance signal represents one of the main advantages of the gas-phase absorbance technique. While it takes more than 2 min to record the entire transient absorbance trace if all of the ammonia or nitrogenous gases are allowed to flush out of the train, it takes less than 25 s for the peak maximum to be reached. Reagents. Aside from common laboratory reagents, a 1%by weight aqueous solution of urease was used. The solution was prepared fresh every week and kept refrigerated. The solid reagent (Sigma Chemical Co.) had an activity of 1200 unita/g. It was also kept refrigerated. Procedure. About 10 g of each test fertilizer (Table I) was accurately weighed into 500-mL volumetric flasks and diluted to volume. As a check of the homogeneity of the weighed solid materials, duplicate samples of all test fertilizers were prepared and carried through all the determinations. The relatively large size of fertilizer samples, with the consequent dilutions, was dictated only by the desire to take nearly representative samples of the products rather than by sensitivity requirements of the proposed method. (Samples on the order of 0.1 g and smaller can readily be used when the fertilizer is clearly homogeneous.) Although all test fertilizers were advertised as “soluble”, some suspended matter was present in several of the prepared stock solutions. Therefore, after being allowed to stand overnight, all fertilizer stock solutions were filtered through Whatman No. 1. These filtered stock solutions were used for the determination of ammonia, urea, and nitrate nitrogens by the proposed method and by the accepted methods. a. Ammonium Nitrogen. For the determination of ammonium N by the proposed method, further preparation of the fertilizer stock solutions involved only dilution. Since it is desirable that the concentration of NH4+in the test solution fall within the linear portion of the ammonium N calibration curve, a suitable dilution factor was estimated using the percent N values indicated on the product labels or obtained by preliminary experiments. In the determination of ammonium N, the concentration range used extended from 25 pg/mL to 150 pg/mL ammonium N. In addition to evaluation by simple comparison to a calibration curve, the method of standard additions was also applied. The exact method of preparation of the spiked and unspiked fertilizer test solutions is indicated in Table I. For analysis, 1-mL aliquots of each of the prepared solutions were injected into fresh 6-mL aliquots of 10 NaOH and the transient absorbance signals recorded. At least five repeated injections were made for each solution.

‘ z s 0

0

0

0

0

8 2 % 0.1

2 8 0

0

I

0.10.1

2 8 8 0

0

0

4 0 . 1

9 0

0

0

s s g % 3 m

8 8 8 0

0

0

0.10.10.1

w 0

U

a,

B

8

w

E

.-0

U

3

! ci

ANALYTICAL CHEMISTRY, VOL. 55, NO. 3, MARCH 1983

537

Table 11. Results of Determination of Ammonium Nitrogen in Commercial Fertilizers by the Proposed Method

fertilizer Siim-u-plant All Purpose Plant Food Schultz Instant Liquid Plant Food Stern’s Miracid Soil Acidifier anld Plant Food Ra,pid.aro Soluble Piant >ood Hyponex Afri’can Violet Food

label information, % standard addition ammonium calibration N curve spike no. 1 spike no. 2

accepted method, % average ammonium N difference

4.5

2.47

2.46

2.45

2.46

2.40

1.0

1.33

1.29

1.37

1.33

1.31

3.0

3.28

3.19

3.38

3.28

3.24

+ 0.06 + 0.02 + 0.04

4.0

3.65

3.67

3.73

3.68

3.65

+ 0.03

1.2

1.86

1.79

1.86

1.84

1.78

t 0.06

b. Urea Nitrogen. A n aliquot of each fertilizer stock solution was placed in a 500-mL volumetric flask, treated with 10.0 mL of 1% urease solution, allowed to stand for an hour at room temperature, and diluted to volume. Since urea N is determined by subtracting the experimentally determined ammonium N from the total ammonium N concentration obtained following enzymatic hydrolysis of urea to ammonia, appropriate volumes of the stock solutions had to be taken to ensure adequate magnitude for the calculated difference and also to ensure that hydrolyzed samples, both spiked and unspiked, fell on the linear portion of the ammonium calibration curve. The exact quantities used are listed in Table I. For analysis, at least five repeated 1-mL injections of each solution into fresh 6-mL aliquots of 10 N NaOH were made. c . Nitrate Nitrogen. The cadmium column used for the reduction of nitrate was prepared as recommended in the literature (10). Briefly, this involved the following. A quantity of metallic cadmium was prepared by immersing zinc rods into a solution of 0.26 M CdS04. These rods were permitted to stand for about 1.5 h. During this time, the cadmium sponge deposited on the rods was periodically scraped off. The lumps of cadmium sponge were then subdivided by blending for 2-3 s in a blender and carefully transferred into a 50-mL buret containing a small wad of glass wool at the bottom to prevent passage of fine Cd particles into the buret tip. A Cd bed about 25 cm long was prepared and thoroughly rinsed first with 0.1 N HC1 and then with distilled water. The Cd column !wasalways kept filled with water. After the passage of each fertilizer sample through the column, the activity of the column was restored by passing about 50 mL of 0.1 N HC1 at a rate of about 3 drops/s and rinsing with distilled water. The reduction was carried out as follows. An aliquot of the fertilizer stock solution was added to the column and passed through at a rate of 1drop/s collecting the effluent in a volumetric flask. As soon as the meniscus of the solution reached the top of the Cd bed, the stopcock was closed and 10 mL of distilled water was added to the column, thereby ensuring that the walls of the tube are rinsed of any remaining sample. This liquid was also passed through the column at a rate of 1drop/s. A second 10-mL water rinse was passed through the column at a rate of about 3 drops/s. Finally, about 25 mL of water was passed through the column at a rate of approximately 15 mL/min until the water meniscus again reached the top of the Cd bed. The contents of the volumetric flask were diluted to volume and tested by making repeated injections of 1-mL aliquots into 6-mL aliquots of tl N HCl. The sizes of the aliquots of the fertilizer stock solution taken to prepare the unspiked and the spiked samples for the determination of nitrate N were selected in such a manner that the NOz- concentration obtained after the Cd reduction step fell on the linear portion of the nitrite calibration curve (up to about 40 pg/mL NO2-). The details are indicated in Table I.

RESULT8 AND DISCUSSION Ammonium Nitroglen in Fertilizer. The results of ammonium N determination in six different commercial fertilizers by the proposed method are given in Table 11. The calibration

curve was prepared by injecting 1-mL aliquots of NH4+ standard solutions into fresh aliquots of 10 N NaOH, recording the transient absorbance signals, and making a plot of recorded peak height vs. concentration of ammonium N. The calibration curve remained linear up t o about 300 yg/mL ammonium N. All results represent the average of duplicate fertilizer samples. As the data in Table I1 indicate, evaluation by simple comparison to a calibration curve yields results which are equivalent to those obtained by the method of standard1 additions, indicating freedom from interferences in the analysis. The results obtained by the proposed gas-phase absorbance method are compared to those obtained by the AOAC-endorsed (11) Magnesium Oxide Method (2.055) which involves distillation of amrnionia followed by titration. The agreement between the two methods is very good, the difference not exceeding 3.3% in any of the cases. If label information on product composition is interpreted to stand for minimum concentrations rounded off to the nearest whole number, then all experimental results for percent ammonium N are in agreement with manufacturers’ information. Urea Nitrogen in Fertilizer. The results of determination of urea N in fertilizers by the proposed gas-phase absorbimce method are given in Table 111. Following treatment with urease, the urea N is first determined in a sum with original ammonium N. Urea N proper is then determined by mbtracting the previously determined percent of ammonium N. Part (a) of Table 111.represents the results of the determination of the sum of ammonium plus urea N in unspiked samples and in those spiked with additional urea. The data indicate good agreement between the results obtained by simple comparison to an ammonium calibration curve and those obtained by the use of standard additions. This further supports the absence of interferences in the proposed method. To obtain the percent urea N (proposed method) listed in the third column of Table III(b), the average percent ammonium N listed in the sixth column of Table I1 is subtracted from the total percent of ammonium N plus urea N listed in the last column of Table III(a). The accepted method used for comparison with the proposed method was the AOACendorsed (11)Kjeldahl method (2.070). The latter involves treatment with urease and distillation of ammonia followed by titration. The data in Table III(b) reveal fair agreement between the proposed and the accepted methods. A small amount of urea was detected by both methods in a fertilizer in which the manufacturer had not listed any. In another fertilizer there was discovered by both methods an approximately 4-fold excess of urea over the modest claim made by the manufacturer (see “Reproducibility” section). Nitrate Nitrogen in Fertilizer. The results of the determination of nitrate N in fertilizers by the proposed gas-

538

ANALYTICAL CHEMISTRY, VOL. 55, NO. 3, MARCH 1983

Table I11 (a) Results of Analysis for Total Ammonium Plus Urea Nitrogen by the Proposed Method

proposed method, % urea plus ammonium N standard addition calibration curve spike no. 1 spike no. 2 average

fertilizer Stim-u-plant All Purpose Plant Food Schultz Instant Liquid Plant Food Stern’s Miracid Soil Acidifier and Plant Food Ra.pid.gro Soluble Plant Food Hyponex African Violet Food

11.18 11.93 31.94

11.05 11.80 29.3 7

10.79 11.99 30.32

11.01 11.91 30.54

20.69

20.67

20.22

20.53 2.10

2.10

( b ) Results of Determination of Urea Nitrogen by the Proposed Method label information, % urea N

fertilizer Stim-u-plant All Purpose Plant Food Schultz Instant Liquid Plant Food Stern’s Miracid Soil Acidifier and Plant Food Ra.pid.gro Soluble Plant Plant Food Hyponex African Violet Food

proposed method, % urea N

accepted method, % urea N

difference

2.1

8.55

8.73

-0.18

8.4

10.58

10.17

+0.41

27

27.26

26.46

t 0.80

14

16.85

15.31

t 1.54

0.26

0.17

+ 0.09

Table IV. Results of Analysis for Nitrate Nitrogen by the Proposed Method

fertilizer Stim-u-plant All Purpose Plant Food Schultz Instant Liquid Plant Food Stern’s Miracid Soil Acidifier and Plant Food Ra.pid.gro Soluble Plant Food Hyponex African Violet Food

proposed method, % nitrate N label information, calibration standard % nitrate N curve addition average

accepted method, % nitrate N

difference

4.5

0.00

0.00

0.00

0.03

-0.03

0.6

0.70

0.79

0.14

0.79

-0.05

0.00

0.00

0.00

0.00

0.00

5.0

4.94

5.14

5.04

5.03

t

5.8

6.20

5.67

5.94

5.99

-0.05

phase absorbance method are given in Table IV. Aliquots of fertilizer stock solutions were passed through the Cd column to reduce NO; to NO;. One-milliliter aliquots of the resulting solutions were injected into aliquots of 8 N HC1 and the transient absorbance signal caused by the evolved nitrosyl chloride and oxides of nitrogen was recorded. To prepare the calibration curve, the peak height of the recorded absorbance signal was plotted vs. the concentration of NO2- in a series of standard nitrite solutions. The calibration curve remained linear up to about 40 pg/mL NOz-. The reduced fertilizer samples were also evaluated by the method of standard additions by spiking the samples with additional NO9- prior to passage through the column. The slight variation between the results obtained by the proposed method using the calibration curve and the standard additions evaluations is attributed to random experimental errors. Good agreement is seen between the average results yielded by the proposed method and those yielded by the accepted method. In the latter, the determination of nitrate N is achieved by first determining the total N (urea, ammonium, plus nitrate) in the sample and then subtracting the other forms of nitrogen determined by the previously described accepted methods. The determination of total N by the accepted method involves treatment with urease t o hydrolyze urea and reduction of nitrate to ammonia by the

0.01

AOAC-endorsed (11)Devarda method (2.005), followed by the usual Kjeldahl distillation of ammonia and titration. In all cases except one, the experimental results confirmed the minimum percent nitrate N listed by the manufacturers on the product labels. In the case of the other fertilizer, the proposed method revealed no nitrate N, the accepted method showed 0.03%, while the label indicated 4.5%. The aliquot of this fertilizer spiked with standard NO3- prior to analysis showed 100% recovery of the spike, indicating the absence of chemical interference by the matrix in the proposed method. The duplicate stock solutions of the fertilizer yielded identical results. The reason for the discrepancy is unclear but may be related to the sampling step (see “Reproducibility” section). Recovery Checks. Prior to applying the proposed gasphase absorbance method to the analysis of commercial fertilizers, the recovery of ammonium, urea, and nitrate from known mixtures of standards was studied a t different relative concentrations. The recovery of urea N a t 100 and 1000 pgjmL levels averaged 99%; the recovery of ammonium N a t 100 and 1000 pg/mL levels averaged 101%, and the recovery of nitrate N at the same concentration levels averaged 95%. The slightly low recovery of nitrate N is attributed to incomplete reduction by the Cd column. The efficiency of the Cd column being used was evaluated and frequently checked by running standard solutions of nitrate N. Com-

ANALYTICAL CHEIMISTRY, VOL. 55, NO. 3, MARCH 1983

parison to nitrite standards revealed a conversion of 9~5.6%~ A factor taking into account this incomplete conversion was applied to all fertilizer solutions passed through the Cd column. Reproducibility. Since the objective of this research was only to show the feasibility and advantages of the proposed analysis step, no particular attention w a paid ~ to the reliability of the sampling step (that is the isolation of a 10-g sample from inhomogeneous solid fertilizer material) beyond preparing all stock solutions of fertilizers in duplicate. The relative deviation from the mean for each of the duplicate determinations of ammonium N in the tested fertilizers ranged from 0.79% for Schultz Instant to 5.6% for Stern's Miracid. Similarly, for urea N, it ranged from 0.00% for Hyponex African Violet to 2.6% for Stirn-U-Plant. And for nitrate N, it ranged from 0.00% for Stim-U-Plant to 0.81% for Hyponex African Violet. These values indicate that the actual variability of composition in the 10-g fertilizer samples was not pronounced. However, since only duplicate iruns were made, it should still be remembered that only comparisons between the results obtained by the proposed and the accepted methods on the same sample are really valid, and not comparisons between either of these and the label information provided by the manufacturers. For evaluation of the reproducibility of the injection technique, 10 repeated injections of one fertilizer solution into fresh aliquota of 10 N NaOH were made and the resulting ammonia absorbance signals wiere recorded. The relative standard deviation of the peak heights was 1.4%. For evaluation of the reproducibility of the enzymatic hydrolysis of urea, 10 alicluots of one fertilizer were treated with urease and the resulting total ammonium content was evaluated by averaging three injections of each solution. A relative standard deviation of 0.70% was observed. Obviously, the hydrolysis step contributes no measurable uncertainty to the results. For evaluation of the reproducibility of the reduction of nitrate N by the Cd column, ten 5-mL aliquots of a fertilizer stock solution (found to contain 1001 Fg/mL nitrate N)were passed through the column (as described under "Procedure") and each diluted to 5010 mL. The activity of the column was restored by the HCl and water rinses between runs. For each reduced solution, threle injections of 1-mL aliquots into 8 N HCl were made and the recorded peak intensities averaged. The resulting ten values exhibited a relative standard deviation of 3.3%. The use of the Cd column is concluded to contribute the geatest c3hare of the uncertainty associated with the proposed method. Detection Limits. Defining the detection limit as that concentration which causes a signal twice the size of standard deviation from the mean, the detection limit of the ammonium N analysis by the proposed method is 1 pg/mL. This corresponds to 0.005% by weight of ammonium N in the fertilizer, based on a 10-g sample diluted to 500 mL and tested without further dilutions. The same detection limit applies to the determination of urea N. The detection limit can obviously be improved by going to a larger sample weight or a smaller volume. The detection limit of the nitrate N determination as nitrite by the proposed method is 0.6 p g / m L This corresponds to a detection limit of 0.015% by weight of nitrate N in fertilizer, based on taking a 10-g s,mple, diluting to 500 mL, and causing a &fold dilution during the reduction step. The latter detection limit can obviously also be improved by going to a larger sample or a smaller initial volume. Advantages of t h e Proposed Method. The main advantages of the proposed gas-phase absorption spectrometric method for the determination of different forms of N in

539

fertilizers are speed and simplicity. Excluding the weighing and dilution time, the typical time required to inject a 1-mL aliquot of ammonium N-conta:ining solution and to record the peak maximum is under 25 s, It takes another 1-2 min to drain and refill the reaction vessel and to reestablish a base line on the recorder, thus matking the apparatus ready to receive the next injection. Therefore, in 45 min, close to 20 injections of samples and/or standards can be made. On the other hand, the accepted Kjel.dah1 method for ammoniacal N evaluation takes an average of 45 min for a single distillation, The distillation must be further followed by a titration (with prior standardization o f the titrant) and rinsin.g of the distillation flask. In the proposed method, on the other hand, all that needs to be done between runs is to rinse the Hamilton syringe with the new solution. The reaction vessel requires no rinsing because the continuously flowing carrier gas flushes away any traces of leftover ammonia gas during the refilling of the vessel and the reestablishment of the base line. The only relatively time-consuming step in the prop'osed method is the passage of the nitrate-containing samples through the Cd column. Each sample takes about 10 .min, with another few minutes needed for rinsing the column ,with 0.1 N HC1 followed by rinsing with water. Standards need not be prepared by passage through the column. Instead, the calibration curve iis established by using standard solutions prepared with nitrite. Although the reduction of nitrate in mixtures was shown to be unaffected by the presence of ammonium and urea, other reducible materials were found to be present in the commercial fertilizers, necessitating, as a precaution, the reactivation of the Cd column by the acid rinse after each run of fertilizer. The conversion of urea to NH,+ is a step common to both the proposed and the accepted methods. While the proposed method offers no saving of time in this step, it does offer better accuracy than the accepted Kjeldahl methods for the determination of urea T\J and ammonium N in mixtures. This is due to the decomposition of urea during the determination of ammonium by the Kjeldahl method caused by heating during distillation. The proposed method, by contrast, is quite selective for ammonium N: there is no interference by urea N since analysis is done at room temperature and urea does not evolve ammonia upon injection into base. Another advantage of the proposed method is the small sample size required. Only 1 mL of solution is needed for a single run and the minimum quantity of N which needs to be present for detection in the 1-mL aliquot is about 1 Fg. In conclusion, thle proposed gas-phase absorption spectrometric method offers a simple and convenient alternatime to the familiar Kjeldahl methods for the determination of water-soluble nitrogen compounds in commercial fertilizers. Registry No. Ammonium, 14798-03-9;urea, 57-13-6;nitrate, 14797-55-8. LITERATURE C I T E D Nicholson, G.; Syty, A. Anal. Chem. 1976, 4 8 , 1481. Syty, A. Anal. Chem. 1973, 4 5 , 1744. Winkier, H. E.; Syty, A. Environ. Sci. Techno/. 1976, 70, 913. Ruschak, M. L.; Syty, A. Anal. Chem. 1952, 5 4 , 1639. Syty, A. Anal. Chem. 1979, 51, 911. Syty, A.; Simmons, R. Anal. Chim. Acta 1980, 720,163. Grieve, S.;Syty, A. Anal. Chem. 1981, 5 3 , 1711, Muroski, C. C.; Syty, A. Anal. Chem. 1980, 52, 143. Coiburn. C. E., Ed. "Developments in Inorganic Nitrogen Chemistry"; Elsevier: New York, 1973; Voi. 2, pp 141-142. (10) Sen, N. P.; Donaldson, E. J . Assoc. Off. Anal. Chem. 1978, 67, 1389. (11) Honitz, W. Ed. "Official Methods of Analysis of the Association of Official Analytical Chemists", 12th ed.; AOAC: Washington, DC, 1!)75. (1) (2) (3) (4) (5) (6) (7) (6) (9)

RECEIVED for review September 2,1982. Accepted Decemiber 17, 1982.