Simple Diffusion Methods To Determine Inorganic Nitrogen In Soil And Water

Illinois Fertilizer Conference Proceedings
January 27-29, 1997

1997 Index

Simple Diffusion Methods To Determine Inorganic Nitrogen In Soil And Water

S.A. Khan¹, R.L. Mulvaney, C.S. Mulvaney, and W.B. Stevens spacer

Introduction

Numerous methods have been used for quantitative determination of N in soil extracts and water samples. Most of these methods require expensive and specialized equipment, have a high skill and labor requirement, or are prone to error from chemical or physical interference. Recent work in our laboratory has led to the development of novel diffusion methods for quantitative inorganicN analysis of soil extracts and water. These methods are simple, inexpensive, and reliable. They are also subject to fewer interferences than are most colorimetric methods, and are a convenient alternative to steam distillation for speciating inorganic N in ¹⁵N-tracer investigations. This paper provides a complete description of the methodology, and also discusses the findings from extensive evaluations involving recovery, specificity, and interference tests.

Materials and Methods

Soils

The soils used (Table 1) were surface (0 -15 cm) samples from Illinois selected to represent diverse soils with respect to their physico-chemical characteristics. The Plainfield and Catlin soils were collected from fields under soybean (Glycine max L.) production, while the Houghton soil was taken from a permanently waterlogged site. Each sample was air-dried and crushed to pass through a 0.15 mm screen prior to analyses. The analyses reported in Table 1 were performed as described by Mulvaney and Kurtz (1982).

Soil extracts were obtained by shaking 10 g of soil with 100 mL of 2 M KCl in a 125-mL widemouth polyethylene bottle for 1 h, and filtering the suspension through filter paper (Whatman 42) under vacuum (Mulvaney, 1996). Ten extracts were obtained from each soil and were combined in a 1-L flask to ensure uniform composition before use.

Apparati

Diffusion Unit: The unit used (Fig. 1) consisted of a 1-pint wide-mouth Mason jar (cat. # 7061000518, Kerr Glass Mfg. Corp., Los Angeles, CA; or Ball brand, cat. ,# 14400-66000, Alltrista Corp., Muncie, IN) with 86 mm dome lid (Ball brand only) modified to support the bottom for a 60 mm (dia.) Pyrex Petri dish (55 mm dia., 17 mm high). The lid was modified by: (i) drilling a 9/64 inch hole 5 mm from the inside edge of the sealing band; (ii) attaching a 6-32 x 38 to 64 mm (1.5 to 2.5 inch) stainless-steel machine screw with two O-rings (Viton, 2.9 mm i.d., 6.4 mm od) and two stainless-steel nuts; and (iii) fastening a screw-down mounting base for a nylon cable tie to the stainless-steel screw via a 1/8 inch cable clamp, a 6-32 x 1/2 inch nylon machine screw and nut, and four stainless-steel nuts. To improve the fit of the lid, the cable clamp was ground flush with the nut. The cable tie for mounting the Petri dish was tightened sufficiently that the dish could not slip from the mount, yet could be easily removed and replaced. Double stainlesssteel nuts were employed to prevent loosening.

Microburette or Automatic Titrator: Manual titrations were carried out using a 5-mL microburette graduated at 0.01-mL intervals and equipped with a three-way stopcock for rapid refilling from a reservoir. Automatic titrations were done using a Metrohm Model 678 EP/KF Processor equipped with Model 665 Dosimat (Metrohm, Herisau, Switzerland) and a microelectrode.

Orbital Shaker (optional): A heavy-duty unit (Lab-Line Model 3590) was equipped with a plywood box (91 cm long, 66 cm wide, 30 cm high) to serve as a platform for shaking up to 144 samples. Before use, the shaking speed was adjusted to 75-85 rev min^-1.

Electric Hotplate: The commercial griddle manufactured by the West Bend Co., West Bend, WI (Model # 76208), is satisfactory. Before use, the heat control was adjusted, such that a temperature of 45-50°C was obtained when a thermometer was immersed in 100 mL of deionized water in a Mason jar placed in the center of the hotplate. The hotplate was inclined at an angle of approximately one degree to avoid the accumulation of condensed moisture on the inner side of the lid.

Reagents

Magnesium oxide: Either light or heavy powder obtained from Fisher Scientific, Pittsburgh, PA, was used.

Devarda's alloy: The alloy produced by Merck, Darmstadt, Germany (# 5341), is satisfactory. Boric-acid indicator solution (4 %): To 18 L of deionized water in a 20-L Pyrex bottle were added 800 g of reagent-grade H₃BO₃. The H₃BO₃ was dissolved by vigorous stirring with a motorized stirrer. Then 0.099 g of water-soluble bromocresol green and 0.066 g of methyl red (as a watersoluble sodium salt) were added (satisfactory indicators are available from Merck), and the volume was brought to 20 L with deionized water. With continuous stirring, the pH of this solution was adjusted to 4.2-4.3 by adding single NaOH pellets. When an aliquot of the H₃BO₃-indicator solution was diluted with an equal volume of deionized water, a pH of 4.8-5.0 was obtained.

Sulfamic acid solution (0.2 M): Two g of certified reagent (Fisher Scientific or Merck) were dissolved in 100 ml, of deionized water. The solution was stored in a refrigerator (5°C). Sulfuric acid (0.0025 M standard): To 10 L of deionized water in a Pyrex bottle was added 1.4 mL of concentrated (18 M) H₂SO₄. After thorough mixing, this solution was standardized against primary-standard grade tris (hydroxymethyl) aminomethane (TRAM).

Methanol: Anhydrous grade was used.

Procedures

To determine NH₄⁺-N, an aliquot (10-100 mL) of soil extract or water was pipetted into a Mason jar. A Petri dish was attached to the jar lid with a cable tie, and 5 mL of H₃BO₃indicator solution were dispensed into the dish. Approximately 0.2 g of MgO was added to the jar with a calibrated spoon, and the jar was swirled to mix the contents. A period of 15-30 s was provided so that any MgO dust would settle; then the lid was placed on the jar and sealed with a screw band. The jar was transferred to a room maintained at 25°C for diffusion with or without orbital shaking, was placed in an incubator maintained at 20 or 30°C, or was placed on a hotplate maintained at 45-50°C. After a sufficient period for completion diffusion of NH₃ into the H₃BO₃ solution (Table 2), the Petri dish was removed from the jar, 5 mL of deionized water were added to the dish, and NH₄⁺-N in the H₃B0₃ solution was determined by titration with 0.0025 M H₃S0₄ using an automatic titrator. At the endpoint, the color change was from green to a permanent, faint pink. The amount of N liberated by diffusion was calculated from the expression, (S - C) x T, where S is the volume of H₂SO₄ used in titration of the sample, C is the volume used in the titration of a control (obtained by diffusing the same volume of extractant or deionized water taken for analysis of the sample), and T is the titer of the titrant (for 0.0025 M H₂S0₄, T = 70 µg N mL^-1).

To determine (NO₃^- + N0₂^-)-N, the jar was opened following diffusion of NH₄⁺-N, and the Petri dish was removed from the jar lid. Another Petri dish containing 5 mL of H₃BO₃-indicator solution was attached to the lid. The NH₄⁺-N liberated was determined as previously described. To determine (NH₄⁺ + N0₃^-+ N0₂^-)-N, the procedure described for determination of NH₄⁺-N was modified by adding 0.2 g of Devarda's alloy to the sample in the jar after the addition of Mg0. To determine N0₃^--N or (NH₄⁺ + N0₃^-)-N in the presence of N0₂^- , the sample in the jar was treated with 1 mL of sulfamic acid solution, and the jar was swirled for a few seconds to ensure complete destruction of N0₂^-, before the addition of MgO and Devarda's alloy.

For ¹⁵N analysis of the diffused NH₃ -N by the Rittenberg process, the titrated sample was acidified with 1 M H₂SO₄ ( 0.1 µL µg N^-1) and evaporated to dryness on a hotplate (90°C). The residue was treated with 4 mL of methanol to eliminate H₃BO₃, and excess methanol was removed by heating to dryness at 90°C. Four ml, of deionized water were then added, swirled gently to dissolve any residue on the side of the Petri dish, and the water was removed by heating to dryness at the aforementioned temperature. The dried residue in the dish was dissolved in 0.5 mL of deionized water, and a 0.15-to 0.2-mL aliquot (containing approximately 50 µg of N) was transferred to a plastic microplate and evaporated to dryness (70°C) in a gravity-convection oven for ¹⁵N analysis with an automated mass spectrometer (Mulvaney and Liu, 1991).

Results and Discussion

The Mason jar used in the methods described is a 1-pint wide-mouth design widely available from either Kerr or Ball. The Kerr jar is somewhat larger in diameter and somewhat shorter than the Ball jar, and is a better choice because diffusion is slightly more rapid and cleaning is easier. Regardless of which jar is used, a Ball lid is strongly recommended. The Ball lid seals more reliably than the Kerr lid when reused and may be used for many analyses. To ensure a gas-tight seal, two Viton O-rings are used in attaching the stainless-steel screw that suspends a Petri dish from the lid.

Studies showed that the capacity of H₃BO₃ solution for absorption of NH₃ increases with the concentration of H₃BO₃, and in the methods described diffusions are performed using 5 mL of a reagent that contains 4 % H₃BO₃, instead of the 2 % concentration employed for inorganic-N analyses by steam distillation. With the higher concentration of H₃BO₃, 4 mg of N were quantitatively recovered from 20 mL of 2 M KCl in about the same period needed to recover 300 µg of N with 2 % H₃BO₃ solution. To avoid a loss of sensitivity from the increased H₃BO₃ concentration, the solution in the Petri dish is diluted with 5 mL of deionized water before titrimetric determination of the absorbed NH₃-N.

The 0.2-g additions of MgO and Devarda's alloy specified in the methods described need not be weighed accurately, and are more easily made using calibrated glass or metal spoons. Care should be taken in the use of MgO, because contamination of H₃BO₃ indicator solution by MgO dust will lead to overestimation of diffused NH₃-N. A second addition of MgO with the addition of Devarda's alloy is necessary for complete recovery when analyses are performed on soil extracts to determine N0₃^--N and (N0₃^- + NO₂^-)-N (Saghir et al., 1993).

The minimum periods for the Mason jar diffusion methods (Table 2) were established by diffusing 2-4 mg of N from 10-100 mL of deionized water, 0.5 M K₂SO₄, or 1, 2, or 4 M KCl. A longer period was needed for analysis with a larger sample volume because of a lower surface-to-volume ratio, whereas complete recovery was achieved in a shorter period when the analysis was performed at a higher temperature, with orbital shaking, or in the presence of salt (Table 2). Studies showed that ambient temperatures higher than 35°C reduce the capacity of H₃BO₃ indicator solution for absorption of gaseous NH₃, and thereby lead to low analyses. For this reason, a maximum temperature of 30°C is recommended when diffusions are carried out under ambient conditions. A higher temperature may be employed by using a hotplate as a bottom heat source, thereby producing a temperature gradient within each jar. A study to compare different methods of heating (Table 3) showed that complete recovery was quickly achieved using a hotplate adjusted to 45-50°C as specified. The latter technique allows quantitative recovery of NH₄⁺-N or N0₃^--N from soil extract or water in a single working day, and is particularly useful for soil and plant N testing.

The accuracy and precision of the methods described are illustrated by Table 4, which summarizes the data obtained when replicate analyses were performed to determine the recovery of NH₄⁺-N, N0₃^-N, or NO₂^--N added to soil extracts. In all cases, recovery was quantitative, and averaged between 98.6 and 100.7%. The coefficient of variation for six replicate analysis did not exceed 2%.

Table 5 shows the results of a study to estimate the extent of interference from decomposition of two alkali-labile organic N compounds (glucosamine and glutamine) during diffusions with MgO or MgO plus Devarda's alloy. No liberation of NH₄⁺-N occurred with glutamine. Very little, if any, decomposition of glucosamine was observed when diffusions were performed from 10, 20, or 50 mL of 2 M KCl, whereas appreciable decomposition did occur when the sample volume was 100 mL, owing to a longer period for diffusion. Such decomposition is unlikely to cause significant interference, except with samples having an unusually high content of dissolved organic matter, such as extracts from organic horizons of forest soils or water samples that contain manure or septic waste. In such cases, the aliquot for analysis should not exceed 50 mL, and analytical accuracy should be checked by measuring the recovery of N added as NH₄⁺, N0₃^-, and/or NO₂^-.

Besides their use for quantitative determinations, Mason jar diffusion methods allow isotope-ratio analysis of diffused NH₃^-N in ¹⁵N-tracer investigations. Experience showed that H₃BO₃ residue sometimes interfered with automated ¹⁵N analysis by the Rittenberg technique (Mulvaney and Liu, 1991). In the method described, H₃BO₃ is removed by utilizing its reaction with methanol to form a volatile ester (Steinberg and Hunter, 1957). To ensure complete removal of H₃BO₃, the methanol treatment must be performed in the absence of water, so prior to this treatment the titrated sample is acidified with 1 M H₂SO₄ to prevent loss of NH₃, and then evaporated to dryness.

The Mason jar diffusion units must be cleaned thoroughly before each use. The jars, lids, and Petri dishes may be cleaned by thoroughly rinsing them with warm tap water immediately after use, and can be reused after they are rinsed with deionized water and then dried. Further cleaning is needed when diffusions are being done for N-isotope analysis (see Mulvaney et al., 1996).

Summary

The methods described are so simple and convenient as to be suitable for routine testing, yet they provide the accuracy and precision needed for N-tracer research. As with any analytical technique, these methods should be thoroughly evaluated in one's own laboratory before routine use is undertaken. This evaluation is best accomplished by checking the completeness of recovery when known amounts of N are diffused.

Tables and Figures

Table 1: Analyses of soils

Table 2: Minimal diffusion periods for quantitative analysis

Table 3: Effectiveness of different methods of heating for recovery of NH₄⁺-N

Table 4: Recovery of NH₄⁺-N, NO₃^--N, and NO₂^--N added to soil extracts

Table 5: Extent of liberation of NH₄⁺-N from alkali-labile organic-N compounds by the methods described

Figure 1. Mason-jar diffusion unit

Footnotes and References

¹ S.A. Khan is Research Associate and R.L. Mulvaney is Professor, Dept. of Natural Resources and Environmental Sciences, Univ. of Illinois; C.S. Mulvaney is Consultant, Monticello, IL; and W.B. Stevens is Jonathan Baldwin Turner Fellow, Dept. of Crop Sciences, Univ. of Illinois.

Mulvaney, R.L. 1996. Nitrogen - Inorganic forms. In: Sparks D.L. et al. (eds) Methods of soil analysis: chemical methods. Soil Science Society of America, Book Series No. 5:1123-1184 Madison, Wisconsin.

Mulvaney, R.L., and L.T. Kurtz. 1982. A new method for determination of 15N-labeled nitrous oxide. Soil Science Society of America Journal 46:1178-1184.

Mulvaney, R.L., and Y.P. Liu. 1991. Refinement and evaluation of an automated mass spectrometer for nitrogen isotope analysis by Rittenberg technique. Journal of Automatic Chem 13:273-280.

Saghir, N.S., R.L. Mulvaney, and F. Azam. 1993. Determination of nitrogen by microdiffusion in Mason jars. 1. Inorganic nitrogen in soil extracts. Conunun Soil Sci Plant Anal 24:1745-1762.

Steinberg H., and D.L. Hunter. 1957. Preparation and rate of hydrolysis of boric acid esters. Ind Eng Chem 49:174-181.

Back to Table of Contents