Rapid Photometric Determination of Total Nitrogen, Phosphorus, and Potassium in Plant Material BENJAMIN WOLF,The G.
L. F.-Seabrook
Farms Raw Products Research Division, Bridgeton,
Rapid photometric methods For the determination of total nitrogen, phosphorus, and potassium in plant material are presented. The plant material is rapidly ashed b y means of sulFuric acid and hydrogen peroxide. Tests for the three elements are run on separate aliquots of the ash extract b y means of a photoelectric colorimeter. Comparison with A.O.A.C. methods shows fairly good agreement. Although not so accurate as the A.O.A.C. methods, these rapid methods provide for a very rapid determination of nitrogen, phosphorus, and potassium with suHicient accuracy for many routine purposes.
T
H E ash analysis of plant tissue yields much information as to the nutritional history of plants, affords an estimate as to the rate and total amount of nutrients removed from the soil, and helps the agronomist t o meet this nutrient removal. I n recent years, the accurate ash analysis of specified plant tissue has been used to ascertain the nutritional status of the plant (6, ’7). A bibliography covering recent investigations in ash analysis has been prepared (4). The results of many more ash analyses of plants would be of great importance to agriculture. However, the use of this type of analysis of plant tissue has been somewhat limited by the length of time required for such determinations. The ashing of plant material can be speeded up considerably by the use of sulfuric acid and hydrogen peroxide. This method has been used for the determination of phosphorus (5) and nitrogen, and has been suggested for other nutrients ( 2 ) . Rapid tests for soluble plant nutrients have been used t o practical advantage (8),but are of limited value. They give the amounts of soluble nutrients only; they do not yield a complete history of plant nutrition, nor give the amounts or rates of nutrient removal. The need for knowing the total constituents in plant material, and for obtaining this information for a large number of samples in a short time, has led to the development of the rapid tests described in this paper. These methods are based on a rapid system of wet-ashing ( 2 ) and rapid photometric methods. They have been used with success on peas, beets, spinach, crimson clover, and rye. Their chief advantage lies in the speed of determination (it is possible for an analyst to determine the total amounts of nitrogen, phosphorus, and potassium in 24 ground samples in 1 day). Added advantages are in the economy of reagents and labor. A large number of reagents prepared for the analysis of soluble nutrients in soils and in plant extracts (8) may also be used in these methods. A laboratory equipped to determine soluble nutrients can also determine ash constituents of plants without great additional expense for equipment or reagents.
DETERMINATION OF KITROGEN.Graves’ reagent (Q), 80 grams of sodium chloride dissolved in 130 ml. of water, to which are added 100 ml. of a cold saturated solution of mercuric chloride (7%) with shaking. The salt is almost dissolved and 70 ml. of a saturated solution of lithium carbonate (1%) are added in small quantities and with continued shaking. Five grams of talc are added to the solution which is filtered, stored in a brown bottle, and kept stoppered, The solution will keep for several weeks. Shake before using. Gum arabic, 0.25% solution. Sodium hydroxide, 15%. Standard nitrogen. Ammonium chloride in extracting- solution to supply IO p.p.G. of nitrogen. Detailed directions for preparing the solutions listed below have been given ( 8 ) . DETERMINATION OF PHOSPHORUS. Ammonium molybdate, 2.5%in 6 N sulfuric acid. Aminonaphthol sulfonic acid. Standard phosphorus. Monosodium phosphate monohydrate, dissolved in extracting solution to give 20 .p.m. DETERMINATION OF POTASSIUM. !odium cobaltinitrite. Isopropyl alcohol (no formaldehyde). Gum arabic solution (as for nitrogen). Standard potassium. Potassium chloride in extracting solution to supply 50 p.p.m. METHODS
ASHING. 4 0.200-gram sample of finely ground, well-mixed, plant material is placed in a 50-ml. Erlenmeyer flask, marked a t 50 ml., and 3 ml. of concentrated sulfuric acid are added. The flask is rotated to mix the plant material with the acid. A small, short-stem funnel is placed in the neck, and the flask is heated on a hot plate for about 5 minutes after fuming starts. The temperature of the hot plate is adjusted so that fumes are given off, but are not driven from the flask. The flask is removed from the hot plate and allowed to cool for a few minutes, and 1 ml. of 30% hydrogen peroxide is slowly added, dropwise, to the sides of the funnel and flask. Slow addition in this manner avoids spattering and washes down charred material adhering to flask or funnel. The flask is reheated for about 2 minutes. If the material is still dark, the flask is cooled, rotated, and 5 additional drops of hydrogen peroxide are added as before. The flask is again heated. This process of adding 5 drops and reheating is repeated until the solution is colorless, then the solution is heated slowly for 5 minutes to expel exceqs hydrogen peroxide. Removing the funnel prior to the last heating will aid in the expulsion of the hydrogen peroxide, but care in heating must be exercised to avoid loss of the liquid. The flasks are cooled, extracting solution is added to the mark, and the contents are filtered on a Whntman No. 2 filter paper to remove silica. Aliquots of the filtrate (referred to as ash extract) are used for the determination of nitrogen, phosphorus, and potassium. (Where many samples are being run it is convenient to allow the silica to settle out in the flask. Aliquots are taken of the supernatant liquid without disturbing the precipitate on the bottom.)
Table 1.
REAGENTS AND SOLUTIONS
All reagents are of C.P. grade. Wherever possible, Baker’s analyzed rea ents are used. ASHING. kulfuric acid, concentrated, nitrogen-free. Hydrogen peroxide, 30%. PREPARATION OF ASHEDMATERIAL. Mor an’s Universal extracting solution ($), 0.5 N acetic acid buffere! a t p H 4.8 with sodium acetate, hereafter referred to as extracting solution.
Determination of Nitrogen, Phosphorus, and Potassium for Standard Curves and in Plant Material Null
APPARATUS
I n addition t o the apparatus suggested for the determination of soluble nutrients in soil and plant extracts (8),the following are needed: plant mill, hot plate, and Erlenmeyer flasks, 50 ml. marked a t 50 ml.
N. 1.
Nutrient Determined Nitrogen
Material
Standard solution Ash extract PhosStandard phorus solution Ash extract Potassium Standard aolution Ash extract
121
Useful Range P.p.m. 0 to 8
Volume of Aliquots Ml. 0 to 16
Diluted to
Filtei Used
Adjustment with Blank to
M1. 20
425 (blue)
100
20 t o 160 1 0 t o 1 2 . 5 0 t o 12.5
20
425
0 (log
1 t o 50 0 t o 20
10
650 (red)
100
6 0 to 8
50 t o 200 1
(blue)
scale)
122
INDUSTRIAL AND ENGINEERING CHEMISTRY
CALIBUTIONOF STANDARD CURVES. Standard curves for nitrogen, phosphorus, and potassium are prepared by adding a series of aliquots of standard solutions and proper amount of the blank to photometer tubes, diluting to appropriate levels (Table I), and treating as for the determination of the elements in the ash extract. Photometer readings are taken with a Fisher electrophotometer using the appropriate filters (Table I). Concentration deflection curves are drawn from the resultant readings. A suitable blank is prepared by ashing 0.200 gram of pure sucrose in the same manner as the plant material, and the ash is diluted to 50 ml. with extracting solution. Aliquots of this blank are added to each of the standards as follows: 1.0 ml. for the nitrogen, 5 ml. for the phosphorus, and 1.0 ml. for the potassium determination. DETERMINATION OF NUTRIENTS. Nitrogen. To a series of photometer tubes 1.0-ml. aliquots of the ash extracts are added, the contents are diluted to 20 ml. with extracting solution, and 0.2 ml. of gum arabic solution is added to each. The contents are mixed by means of a flat-bottomed rod and 0.5 ml. of Graves' reagent is added. The contents are again mixed by means of the flat-bottomed rod and 5 ml. of 15% sodium hydroxide are added, mixing again after the addition of the hydroxide. Photometer readings are taken in exactly 15 minutes. A 425 blue filter is used and the null is adjusted to 100% transmission with the blank. Phosphorus. Aliquots of the ash extracts ( 5 ml.) are pipetted into a series of photometer tubes, the contents are diluted to 20 ml. with extractin solution, and 4 ml. of ammonium molybdate solution and 2 m& of aminonaphthol sulfonic acid solution are added. The contents are stirred and allowed to stand for I5 minutes. Photometer readings are taken, using a 425 blue filter and adjusting the blank to 0 (log scale). Potassium. Aliquots of the ash extract (1ml.) are pipetted into a series of photometer tubes, the contents are diluted to 10 ml. with extracting solution, and to this is added 0.5 ml. of gum arabic solution. The tubes are rotated, avoiding any loss of the liquid, fhen 2 ml. of sodium cobaltinitrite solution are pipetted directly into the contents of each tube. The tubes are rotated to mix the contents, allowed to stand for 5 minutes, and 10 ml. of isopropyl alcohol are run directly into the solution. (The isopropyl alcohol should always be added from a uniform dispensing unit, such as an automatic pipet, and from a uniform height.) The tube is stoppered, inverted 3 times, and allowed to stand for 15 minutes. The stopper is then removed and photometer readings are taken, using the 650 red filter and adjusting the blank to 100% transmission. CALCULATIONS. Using the photometer readings obtained in the tests, the concentrations in p.p.m. can be read directly from the standard curves or from charts prepared from such curves. DISCUSSION OF METHODS
SELECTION AXD PREPARATION OF SAMPLES. Samples dried a t 40.56 ' C.(105OF.) should be finely ground (100% to pass a 0.5mm. sieve) and thoroughly mixed. As in all cases where small aliquots are taken from large samples, it is important that the sample should be finely ground and uniformly mixed. Selection of the portion of the plant t o be used for testing depends upon the purpose of the investigation and the type of plant tested. Investigations as to the status of nutrition have employed the use of leaf material (6, ?'), The entire plant should be analyzed for determining the amounts and rates of removal of nutrients by crops. CALIBRATION OF STANDARD CURVES. The reactions for the tests are influenced considerably by pH changes. Use of the buffered extracting solution helps to maintain a more uniform p H value. However, since the addition of the ash extract to extracting solution does cause a change in the pH, it is best compensated for by adding a similar aliquot of the blank extract in the calibration of the standard curves-for example, a 5-ml. aliquot of the ash extract is used in the phosphorus determination. A similar aliquot of the blank extract should be added to each aliquot of the standard and the contents diluted to 20 ml. TESTS.Aliquots of the ash extract of the blank should be of the same volume and treated in the same manner as the aliquots of the plant material. A blank should be run with all determination. The ashing by means of sulfuric acid and hydrogen peroxide effects solution of the elements in question, with the nitrogen present in ammonium form. Silica precipitates and can be removed by filtration or ali-
Vol. 16, No. 2
quots of the ash extract may be removed from the supernatant liquid without disturbing the precipitate. In the nitrogen determination, care should be taken to heat the sample initially for about 5 minutes after fumes appear to prevent loss of nitrate nitrogen (2) and aid in decomposition of the complex organic nitrogen compounds. Baking should be avoided, since it will drive off nitrogen. A large excess of hydrogen peroxide in the ashing will cause an appreciable loss of nitrogen. The precipitate formed by addition of the Graves' reagent and sodium hydroxide is greatly influenced by the time of standing and pH. Repeatable results can be obtained by taking readings exactly 15 minutes after the addition of the sodium hydroxide. Changes in the p H value due to addition of the sample are compensated for by similar additions of the blank to the standards in the preparation of the standard curves.
Table
II. Determination of Total Nitrogen, Phosphorus, and Potassium in Plant Material
[Compariwn of rapid photometric methods with A.O.A.C. (f) methods] Nutrient Material Rapid Methods A.0..4.C. Methods F /O a
Nitrogen
Wheat straw 0.4 Alfalfa 2.4 Cottonseed meal 7.0 Soybean meal 7.6 Phosphorus Starter mash 1 1.01 Starter mash 2 0.66 Starter mash 3 0.82 Starter mash 4 0.68 0.81 Starter mash 5 Starter mash 6 1 .oo Potassiumb Starter mash 1 1.1 Starter mash 2 0.8 Starter mash 3 0.7 Starter mash 4 0.7 Starter mash 5 0.7 Starter mash 6 0.9 4 Per cent of plant material on a dry basis. b Corrections made for ammonium ion present.
%4
0.308 2.82 7.26 7.69 1.08 0.70 0.77 0.69 0.77 1.04 0.95
0.76 0.69 0.75 0.86 0.82
The gum arabic gives a more uniform dispersion, and also P I I ables the determination of larger amounts of nitrogen. The readings are most accurate for nitrogen in amounts from 1 to 8 p.p.m. With a 1.0-ml. aliquot, this represents from 0.5 to 473 of total nitrogen in the dry tissue; most plant samples mill fall within this range. For material containing less than 0.57, total nitrogen, a standard amount of nitrogen can be added to the ash extract aliquot and subtracted from the results. The phosphorus determination is only slightly affected by pH changes. Such changes are compensated for by adding a similar quantity of blank to the standard in preparation of standard curves. The phosphorus test allows for the determinations of phosphorus in amounts from 0.25 to 12.5 p.p.m. This represents 0.025 to 1.25% total phosphorus in the dry plant material, if a 5-ml, aliquot of the ash extract is used. Most plant material will contain phosphorus in amounts between these figures. The potassium determination is influenced by changes in pH value, by the rate of mixing the alcohol with the cobaltinitrite, by temperature changes, and ammonia present. The influence of pH is nullified by taking the same amount of aliquots in all cases and by adding similar amounts of the blank to the standards. The colloidal precipitate formed by the addition of the alcohol is greatly influenced by the speed with which the alcohol is added and subsequent shaking thereof. Gum arabic aids in the formation of a more uniform precipitate. However, constantly to obtain uniform precipitates, the alcohol should be delivered from the same apparatus a t a constant height above the contents in the photometer tube. By adding the alcohol from a 3-way, 10-rnl. automatic pipet directly into the contents and immediately stoppering and inverting 3 times. fairly uniform precipitates were oh-
ANALYTICAL EDITION
February, 1944
rained. Previously (8) alcohol was added slowly down the sides of the tube and rotated. The present method is more accurate, since there is less dependence on the individual operator. If large changes in temperature take place in the laboratory, separate standard curves should be drawn for every 5 ” C. This scems simpler than cooling to a standard temperature. In previous methods (g), formaldehyde was used to avoid intcrference of ammonia. Formaldehyde lessens interference from ammonia but also reduces the sensitivity for the determination of potassium in lower concentrations. In the present method, there is no provision for avoiding the interference of ammonium ion in the test but a correction for the ammonium ion is made. Curves ran be drawn showing the influence of adding standard amount of :tmmonium ions to definite amounts of potassium and deductions made according to such graphs, For practical purposes an accurate correction can be made by use of the following equation: X = 0.16 ( A ) @ )where X = yo potassium to be deducted, A = % total nitrogen, and B = % potassium (uncorrected), all on a dry weight basis. This correction is b a e d on a 1-ml. sample of the ash extract. The potassium test is satisfactory for potassium in amounts of r? to 20 p.p.m. If a 1-ml. sample is used this represents 1.25 to $5.0%total potassium in plant material. For amounts less than 1.25Yr potassium, a standard amount of potassium is added to the aliquot and later deducted from the results. ACCURACY. The nitrogen test can be repeated with an accui’acyof 0.2 p.p.m., providing the final ammonium nitrogen content is between 0.5 and 4 p.p.m. This represents 0.2% of total nitrogen on a dry weight basis for amounts between 0.5 and 470 o f nitrogen. Comparison of the rapid method (Table 11) with the A.O.A.C. method shows fairly good agreement. Phosphorus determinations can be repeated within 0.2 p.p.m.
123
Based on a 5-ml. aliquot, this representa 0.02% total phosphorus in the plant material. Phosphorus determinations of starter mash agreed very closely with A.O.A.C. values (Table 11). The potassium test is perhaps the least accurate of the thrre. primarily because it is influenced by so many factors. However. if precipitated under uniform conditions and effect of temperature and ammonium ion are taken into consideration, fairly accurate results can be obtained. The test can be repeated within 1 p.p.m. in a range of 5 to 20 p,p.m. Based on a 1-ml. sample, this represents 0.25% potassium in a total potassium content of >I plant material from 1.25 to 5.07c. Agreement with the A.O.A.C. method (Table 11)was fair. ACKNOWLEDGMENT
The author wishes to thank A. L. Prince, New Jersey Agrirultural Experiment Station, for the analysis of the plant and starter mash samples by A.O.A.C. methods. LITERATURE CITED (1) hssoc. Official Xgr. Chem., Official and Tentative Methods of Analysis, 1940. (2) Lindner, R. C., and Barley, C. P., Science, 96,565-6 (1942). (3) Morgan, M.F.,Conn. Agr. Expt. Sta., Bull. 450 (1941). (4) Schmidt, C. M., and Jameson, D. H., “Bibliography of Literature on Analysis of Leaf and Other Plant Tisaues, with Special Reference t o Content of Mineral Nutrients”, 1935 through 1940, Washington, D. C., American Potaah Institute, 1941 ( 5 ) Snell, F. D., and Snell, C . T., “Colorimetric Methods of -4nalysis”, Vol. 1, p. 499, New York. D.Van Nostrand Co.. 1941. (6) Thomas, Walter, Plant Physiol., 12,571-99 (1937). (7) Ulrich, Albert, Soil Sci.,55, 101-12 (1943). ( 8 ) Wolf, Benjamin, ISD.ENG.CHEM.,ANAL.ED.,15, 248-51 (19431. (9) Yoe, J . H., “Photometric Chemical Analysis”, Vol. 1. p. 307, Yew York. John Wiley & Sons, 1928
Photoelectric Photometry
An
Analysis of Errors at High and at Low Absorption
ROBERT HOUSTON HAMILTON, Department of Physiological Chemistry, Temple University School Mathematical and experimental proofs show that errors in setting the zero point and /, on the galvanometer, and in reading the galvanometer defledion for transmitted light, produce high relative errors when a photoelectric photometer is used with solutions of high or of low absorption. For maximum accuracy, conditions should b e so chosen that readings of transmitted illumination fall on the central portion of the scale.
N
UMEROUS photoelectric devices are in use in chemistry
laboratories throughout the country. Most of them (“photoelectric colorimeters”) are used for measuring the relative amount of light transmitted by a given depth of colored liquid in order to determine the concentration of solute which absorbs light of a certain wave length. The light used in measurement is restricted to more or less narrow bands by means of filters, prisms, or gratings. Measurement is effected by reading the deflection of a galvanometer caused by a photoelectric cell, or by use of a potentiometer to balance the circuits of two photoelectric cells. Several articles dealing with errors involved in the use of these instruments have appeared. Two such articles have dealt with errors produced by wide transmission bands and other sources of stray light (1, 4). The most satisfactory means of meeting this deficiency and controlling the errors arising from it has been to construct empirical analytical curves relating galvanometer
of Medicine, Philadelphia,
Pa,
readings or logarithms thereof to respective concentrations of solute. Elimination of errors of this type, the result of instrument limitations, remains a problem for instrument designers and manufacturers. There remain, however, several sources of error which can be minimized by proper use of available instruments, provided the user is aware of the existence of these errors and of the precautions necessary to make them as small as possible. Most chemists know in a general way that maximum accuracy cannot be achieved %-hen galvanometer readings are either very high or very low. The magnitude of the errors which may occur a t each end of the scale makes an analysis of them desirable. When approximately monochromatic light of a wave length in the region of light absorption of the solution is used, the BeerLambert law applies closely to most colored solutions. Let us assume conditions such that the law is valid, and further assume that galvanometer deflection is proportional to the intensity of light striking the photocell. Then
where y is the ratio of the intensity of transmitted light, I , to the intensity of incident light, IO,e is the Napierian base, c is the concentration of light-absorbing molecules, and k is a positive constant determined by the nature of the solute molecule, the wave
