1182

Determination of Nitrate in Soil Extracts by Dual-wavelength Ultraviolet Spectrophotometry1 RICHARD J. NORMAN, JEFFREY C. EDBERG, AND JOSEPH W. Sruciu2 ABSTRACT A rapid method is described for determining nitrate concentration in a soil extract solution based on its UV absorbance at 210 nm. The interference of nonnitrate species is accounted for by subtracting an empirically-determined multiple of the absorbance of the extract solution at 270 nm from its absorbance at 210 nm. The value of the multiplication factor, R, is calculated from the 210: 270 nm absorbance ratio of the extract solution after it has been treated with Raney-Nickel catalyst to remove nitrate. The method was tested using 10 Illinois soils. The composite value of R for these soils was 3.05 ± 0.22. This value, as well as each individual value of R for the respective soils, was used to calculate the nitrate concentration. The results were then compared to one another and to results obtained by the steam distillation method. Correlation among these various methods was very high, giving values of r2 of 0.9987 and 0.999. Large errors occurred when no correction for nonnitrate species was made. " Additional Index Words: soil testing, Raney-Nickel NO3, analysis, UV spectroscopy. Norman, R.J., J.C. Edberg, and J.W. Stucki. 1985. Determination of nitrate in soil extracts by dual-wavelength ultraviolet spectrophotometry. Soil Sci. Soc. Am. J. 49:1182-1185.

N

ITRATE (NOf) is one of the most important ions to plant growth, and its measurement has received much attention (Bremner, 1965; National Research Council, 1978). In a previous paper we described a method to assay soil extract solutions quantitatively for NOj and nitrite (NOi~) using ultraviolet (UV) spectrophotometry, which has the advantages of measuring both ions, is rapid, and avoids many of the interferences common to other sensitive NOj methods (Norman and Stucki, 1981). However, for routine analyses where many samples are involved, one disadvantage of this method is that in order to correct for nonnitrate species absorbing at 210 nm, the UV spectrum of the soil solution must be obtained before and after chemical treatment to remove the NOf which involves two steps in the procedure and increases the analysis time for each sample. Other sensitive NOf methods also have a similar or more time consuming step in their procedures. If NOf is the only 1 Contribution from the Illinois Agricultural Experiment Station and Dep. of Agronomy, Univ. of Illinois, Urbana, IL 61801. Received 15 Feb. 1984. Approved 18 Mar. 1985. 2 Former Graduate Research Assistant, now Assistant Professor of Agronomy, Univ. of Arkansas Rice Research and Extension Center, Stuttgart, AR 72160; Undergraduate Student; and Associate Professor of Soil Science, respectively.

ion of interest, a more rapid method would be advantageous, especially in cases where many samples are to be analyzed. Goldman and Jacobs (1961) measured NOf in natural water effluents using dual-wavelength, UV spectrophotometry. They measured the concentration of NOj by subtracting an empirically-determined multiple of the absorbance of the sample solution at 270 nm from its absorbance at 210 nm. This method was based on the principle that if the type and relative proportions of nonnitrate species are relatively uniform within a given set of samples, the absorbance ratio for these species between any two wavelengths will be similarly constant. Since at 270 nm NOj has no absorbance but various other nonnitrate species absorb significantly (Bastian et al, 1957;; Armstrong, 1963), they were able to correct the total absorbance of a given sample at 210 nm using the absorbance of the same sample at 270 nm. Such a dual-wavelength approach avoids the necessity of chemical treatment to obtain a nitrate-free solution. The purpose of the present study was to determine whether the concept of Goldman and Jacobs (1961) could be applied successfully to soil extracts and increase the rate of sample analysis for NOj without sacrificing accuracy and precision. MATERIALS AND METHODS Apparatus A Beckman model 5230 single monochromator, double beam in space and time, spectrophotometer was used in conjunction with a deuterium lamp. Samples were introduced to the beam in a 1-cm, quartz flow cell. Measurements were

recorded at 210 and 270 nm with spectral bandwidths of 1.8 and 1.03 nm, respectively. Reagents Sulfuric Acid Solution (20% v/v)—Dilute one part of 98%, reagent-grade sulfuric acid solution (specific gravity = 1.84) with four parts of deionized water and store in a glass-stoppered flask. Sulfamic Acid Solution (2% w/v)—Dissolve 2 g of sulfamic acid crystals in 100 mL of deionized water and store in a glass-stoppered flask in a refrigerator. Raney Nickel Catalyst Powder—Use Ni-Al (1:1) powder such as available from Sargent-Welch Scientific (Raney Catalyst) or Aldrich Chemical Co. (Al-Ni Catalyst). Nitrate Stock Solution—Dissolve 7.221 g of analytical reagent-grade potassium nitrate (KNO3) in 1 L of H2O to give a solution containing 1000 ^g N (in the form of NOf)/mL.

1183

NORMAN ET AL.: DETERMINATION OF NITRATE IN SOIL EXTRACTS .8

Table 1. Series and subgroup descriptions of soils used in this study.

A = WITH NO£ B = WITHOUT NO§

Soil no.

1

2 3

4 5 6 7 8

9 10

Soil series

Subgroup

Cisne, silty loam Drummer, silty clay loam Elliot te, silt loam Flanagan, silty loam Hartsburg, silty clay loam Miami, silt loam Strip mine material (high pH) Strip mine material (high pH) Sidel, silty clay loam Sable, silty clay loam

Mollic Albaqualfs Aquic Argiudolls Aquic Argiudolls Aquic Argiudolls Typic Haplaquolls Typic Haplaquolls Typic Udorthents Typic Udorthents Typic Argiudolls Typic Haplaquolls

HI

.6

o GQ EC O » ffl

.4

.2

0 I— 180

Nitrate Standard Solution—Dilute 5 mL of NOj stock solution to 100 mL with H2O, giving 50 ng NOf-N/mL. Preparation of Soil Extract Place 10 g of dry soil in a 50-mL Erlenmeyer flask, add 0.1 g of calcium sulfate anhydrite (CaSO4) and 30 mL of H2O, stopper, and shake on a mechanical shaker for 15 min. After the suspension has settled for 3 to 5 min, decant and filter through a dry filter paper (Whatman no. 42 or equivalent). If the filtrate is clear, but colored from the organic matter in the soil, refiltration is unnecessary. Turbid filtrates, however, must be refiltered using a new filter paper and receiving flask. One molar potassium chloride (KC1) may be substituted for CaSO4 in the extractions. If alkaline soils of an appreciable carbonate content are to be extracted, the extract should be neutralized to remove all carbonates. Nitrate Determination Determine NO^ by transferring 1 mL of the extract to a 125- by 16-mm test tube, dilute to 10 mL with H2O, and measure the absorbance of the resulting solution at 210 nm (Ai) and 270 nm (A2).

Nitrite will interfere with the analysis for NOf, but its concentration in soils is usually sufficiently low to be ignored. If appreciable levels are suspected, however, add 1 mL of 2% w/v sulfamic acid to the extract solution and swirl to dispel the NOf before determining NOf. Prepare a CaSO4 reagent solution by dissolving 0.1 g of CaSO4 crystals in 30 mL of H2O. Then prepare a reagent blank by diluting 1 mL of the CaSO4 reagent to 10 mL with H2O and determine the absorbance of the resulting solution at 210 nm (A3) and 270 nm (A4). Note that if the soil extract is treated with sulfamic acid to dispel NOf, then the CaSO4 reagent solution must also be treated with the same amount of sulfamic acid in order to obtain the appropriate matrix in the reagent blank solutions. Calculate the concentration of NOf-N (m^ using Eq. [1]:

i = k(A{ -

- R(A2 -

[1] In this equation, the value of a, = 0.555 cm2//*g NOs~-N, which is the absorption coefficient of NOf at 210 nm obtained directly from the slope of the NOf calibration curve (Norman and Stucki, 1981). The value of k = 30 g H2O/g soil, which is the soil dilution ratio. R is an empirical factor used to calculate the absorbance of nonnitrate species at 210 nm. The value of R is determined by measuring the absorbance of the soil extract solution in the absence of both NOf and NO^, according to the following procedure. Into a test tube (125 by 16 mm) place in order 0.3 g of Raney Nickel catalyst powder, 5.0 mL of the soil extract solution (or the soil extract solution treated with sulfamic acid if NOj was dispelled), and 0.5 mL of 20% v/v H2SO4. Place the unstoppered tube in an oven at 55 to 60°C for 35 min. Filter the solution in a 5-mL funnel tube using 7- to 9-cm i

210

240

270

300

WAVELENGTH (nm)

Fig. 1. Absorption spectra of a soil extract solution before (A) and after (B) removal of NOf with Raney-Nickel catalyst.

diam. filter paper (Whatman no. 42 or equivalent). Transfer 1 mL of the filtrate to a second tube, dilute to 10 mL with H2O, and measure the absorbance of the diluted solution at 210 nm (A5) and 270 nm (A6). Prepare a reagent blank solution using the same procedure, except substitute 5 mL of the CaSO4 reagent solution for the soil extract, then determine the absorbance of the resulting solution at 210 nm (A-,) and 270 nm (^8). Calculate the value of R using the equation = (A5- A1)I(A6 [2] This individual value of R may then be used in one of two ways. First, it may be used in Eq. [1] to calculate the NOf -N concentration in a given soil sample. Once the value of R for that soil has been determined by the method described above, it can be used repeatedly. However, if the NOf-N concentration in a particular soil is to be determined only once, then the use of the individual value of R provides no advantage over the method of Norman and Stucki (1981) because the second step in the procedure, i.e., treatment of the sample with Raney-Nickel to remove NOf, must be employed to determine R. The second way to use the value of R is to obtain a composite mean value of all the individual values, which can then be used in Eq. [1] for all samples in the sample set. As will be discussed later, however, the use of the composite value of R instead of the individual values undoubtedly will result in some loss of accuracy. RESULTS AND DISCUSSION The proposed method was tested using 10 different soils from various parts of Illinois (Table 1). In Fig. 1 are shown UV absorption spectra typical of a soilextract solution before (spectrum A) and after (spectrum B) Raney-Nickel treatment to remove NO^~. Notice that the absorption of the untreated extract is maximum at about 203 nm and approaches a minimum of near-zero slope well below 270 nm. Notice also that the absorbance of the treated solution is the same as the untreated solution at 270 nm. The absorbance of the treated soil extract (spectrum B) increases two to fourfold with decreasing wavelength, but lacks the intense peak observed in the untreated solution (spectrum A). This gradual increase in absorbance is due to organic matter and other nonnitrate species in the soil extract solution (Norman and Stucki, 1981), and contributes to the total absorption envelope of an untreated solution. Subtraction of the treated from the untreated spectrum in Fig. 1 produces a spec-

1184

SOIL SCI. SOC. AM. J., VOL. 49, 1985

Table 2. NOi-N concentrations in 10 Illinois soils determined by the proposed dual-wavelength UV absorbance method, and values of R for the same soils, t

1 2 3 4 5 6

7 8 9 10

II NOj-NI

I Ǥ

Soil} 4.66 2.27 2.79 3.15 2.56 2.92 3.39 2.47 2.72 3.61

Method

III NO;-N#

————— NO.--N, liglg soil —————— 57.82 ± 0.60 56.67 ± 0.44 16.60 ± 0.14 16.31 ± 0.23 30.80 ± 0.11 31.08 ± 0.08 26.40 ± 0.06 26.37 ± 0.05 27.02 ± 0.15 24.57 =t 0.24tt 4.28 ± 0.14 3.66 ± 0.07tt 11.42 ± 0.22 11.31 ± 0.15 3.82 ± 0.33 4.09 ± 0.20 90.51 ± 1.10 91.92 ± 0.76 70.80 ± 1.22 69.60 ± 0.78

± 0.62 ± 0.39 ± 0.08 ± 0.20 ± 0.03 ± 0.03 ± 0.48 ± 0.43 ± 0.18 ± 0.36

Table 4. Comparison of NOi-N (/ig/g soil) values obtained using two methods: steam distillation, and UV spectrophotometry with individual R values and the composite-R value.

t Values reported are the mean ± std. dev. based on three trials. J See Table 1 for description of soils. § The overall mean or composite value of R for these 10 soils is 3.05 ± 0.22. 1 Calculated using Eq. [1] and the individual mean values of R (column I) for the respective soils. it Calculated using Eq. [1] and the composite value of R (i.e., 3.05 ± 0.22). tt NOj-N concentration calculated with the composite R value (column III) is significantly different (0.05 level, t-test) from that calculated with the individual R value.

UV spectrophotometryt

Soil

X

SD

1 2 3 4 5 6

56.12 16.75 31.21 26.43 28.21 4.50 11.28 4.24 92.59 68.99

0.84 0.32 0.32 1.26 1.45 0.12 0.24 0.84 1.21 1.36

7 8 9 10

Soil

1 3 7

Incubated!

Cropped§

Control, air-dried

Air-dried

Moist

Air-dried

Moist

4.66a 2.79b 3.39c

4.44a 2.74b 3.50c

4.52a 2.70b 3.43c

4.71a 2.83b 3.47c

4.50a 2.76b 3.33c

t Means followed by the same letter in the same row are not significantly different at the 5% level, t-test. t Aerobic incubation for 30 d at —0.05 MPa moisture tension and 25 °C. § Cropped to ryegrass (four cuttings).

CV

X

SD

1.49 67.82 ±0.60 1.89 16.31 ±0.23 1.02 30.80 ±0.11 2.24 26.40 ±0.06 5.15 24.57 ±0.24 3.66 ±0.07 2.67 2.13 11.42 ±0.22 3.82 ±0.33 19.80 1.26 90.51 ± 1.10 1.98 70.80 ± 1.22

CV 1.03 1.39 0.36 0.24 0.96 1.91 1.94 8.75 1.21 1.72

Individual R values

X

SD

56.67 ±0.44 16.60 ±0.14 31.08 ±0.08 26.37 ±0.05 27.02 ±0.15 4.28 ±0.14 11.31 ±0.15 4.09 ±0.20 91.92 ±0.76 69.60 ±0.78

CV 0.89 0.95 0.42 0.35 0.51 0.77 0.76 4.37 1.37 1.88

t Reported values are the mean (X), standard deviation (SD), and coefficient of variation (CV) based on three trails. Table 5. NOi-N (ng/g soil) determined by UV spectrophotometry corrected (composite .R value) and uncorrected for non-nitrate species, and the ratio (A) of UV-corrected to UV-uncorrected results.

Table 3. Effect of incubation, cropping, and air-drying on the values of R for three soils. Values of .Rt

Composite R value

DistiUationt

Method UV-correctedt

UV-uncorrectedt

X

SD

CV

X

SD

CV

A

57.82 16.31 30.80 26.40 24.57 3.66 11.42 3.82 90.51 70.80

0.60 0.23 0.11 0.06 0.24 0.07 0.22 0.33 1.10 1.22

1.03 1.39 0.36 0.24 0.96 1.91 1.94 8.75 1.21 1.72

60.00 17.42 33.95 27.14 39.86 18.00 12.38 5.26 103.34 77.38

0.53 0.16 0.14 0.10 0.20 0.14 0.09 0.23 1.41 1.46

0.89 0.95 0.42 0.35 0.51 0.77 0.76 4.37 1.37 1.88

0.96 0.94 0.91 0.97 0.62 0.20 0.92 0.70 0.88 0.92

T See footnote, Table 4.

trum of identical shape to that of a NO^" standard solution. The values of .R for all soils studied are reported in Table 2 (column I), and range from 2.27 for the Drummer soil to 4.66 for the Cisne soil. Each value is the mean absorbance ratio determined from the spectra (such as the one reported in Fig. 1) of three RaneyNickel-treated replicate samples using Eq. [2]. The composite mean of the entire set of samples was 3.05. Also reported in Table 2 (columns II and III) are the concentrations of NO^"-N found in the various soil

extracts using the proposed, dual-wavelength UV

method. The values listed in column II were obtained by substituting the individual values of R, given in column I for the respective soils, into Eq. [1]; the values in column III were also obtained from Eq. [1] except the composite mean value of R (3.05) was used in every case. The high correlation (r2 — 0.999) between the results in these two columns verifies that the composite mean of R can be substituted for the individual values in this sample set, with only a small sacrifice in accuracy and precision. For soils 5 and 6, however, the mean NOf-N concentrations calculated with the composite R value varied significantly (5% level) from those calculated with the individual R values. Thus, the use of the composite R value for these two soils is questionable, and illustrates that this substitution of the composite for

the individual R values of a given soil may result in some loss of accuracy. The extent of this variability obviously depends on how much the composite R value deviates from the individual R value, which will be determined by the concentrations of NOj" and nonnitrate species in the soil and by the homogeneity of the soil sampling process. In the case of soil 5, background absorbance due to nonnitrate species was relatively large. Soil 6 had a large nonnitrate background coupled with a low NO^"-N concentration. Changing soil conditions due to cropping, incubation, and air-drying could affect the relative proportions of nonnitrate species in the soil and, thus, could affect the value of R for the soil. Reported in Table 3 are the individual R values for three soils that were subjected to a 30-d aerobic (25°C) incubation and four cuttings of ryegrass (Lolium multiflorum Lam.), then analyzed, air-dried and moist. No effect on the value of R was observed due to incubation, cropping, or whether the soil was moist or air-dried prior to analysis. This suggests that the R value is rather insensitive to moderate changes in these soil conditions. However, the nonnitrate background could be affected significantly by organic amendments (e.g., manure, sewage sludge), the presence of transition metals, or the mixing of soil layers as occurs in some strip-mine soils. If these conditions exist, caution is advised in

TAN: SCANNING ELECTRON MICROSCOPY OF HUMIC MATTER AS INFLUENCED BY METHODS OF PREPARATION

using the UV dual wavelength method, particularly if

the background intensity is high and the individual R values are dissimilar to the composite R value. In the soils studied, the best results using the composite R value were obtained when the nonnitrate background absorbance was < 20% of the total absorbance at 210 nm. Further, as indicated by soils 5 and 6 above, the composite R value should be similar to the individual R value. A good test for similarity is to calculate the difference in NOj-N obtained with each of the two values, then determine whether this difference is within the tolerance limits and requirements of the particular study for which the analyses are being performed. In Table 4 the results reported in columns II and III of Table 2 (i.e., results obtained by dual wavelength UV using the individual and composite R values, respectively) are compared with the results obtained by the conventional steam distillation method (Bremner, 1965). While visual inspection suggests that the individual R values more closely correlate with steam distillation results than does the composite R value, statistical analysis revealed a linear correlation coefficient of 0.999 in both cases. Regardless of the value of R, however, the precision of the dual-wavelength UV method exceeded the steam distillation method in every case, and by as much as tenfold in some cases. To test the importance of the background correction, NOj-N concentrations were calculated using Eq.

1185

[1] with the measured values of all variables (corrected

results), and assuming (A2 ~ A4) = 0 and A3 is unaffected by nonnitrate species (uncorrected result). Comparison of these corrected and uncorrected values (Table 5) illustrates the extreme importance of accounting for the potential absorption by nonnitrate species when using UV spectrophotometry to measure NOj directly. The ratio, A, of corrected to uncorrected concentrations varied from a high of 0.97 to a low of 0.20 (Table 5), indicating that some soils have much greater quantities of nonnitrate material than others. Hence, the correction is essential if consistent and accurate results are to be obtained.