U.S. Department of Agriculture, Agricultural Research Service, National Laboratory for Agriculture and the Environment, 2110 University Blvd, Ames, IA 50011

Description of the Proposed NSM

Chemicals

All chemicals were ACS reagent grade.

(a)

NaOH and HCl.—Sigma-Aldrich (St. Louis, MO).

(b)

Nitrogen gas.—(UN1066) 99.99% purity (Praxair, Danbury, CT).

(c)

Supelco Supelite DAX-8 resin 21567-U.—Sigma-Aldrich.

(d)

Amberlite IR120 strong cation exchange resin.— Hydrogen form 10322 (Sigma-Aldrich).

Equipment

(a)

Analytical balance with glass draft guard.—Capacity 210 g with readability to 0.0001 g (Ohaus PA214, Parsippany, NJ).

(b)

Drying oven.—Precision ± 3°C (Isotemp, Fisher Scientific, Fairlawn, NJ).

(c)

Centrifuge.—Minimum relative centrifugal force 1500 × g (International Equipment Co., Chattanooga, TN).

(d)

Polyethylene centrifuge tubes (50 mL).—VWR Scientific (Batavia, IL).

(e)

Rotary evaporator.—Rotavapor R-210/R-215 with Heating Bath B-491 (Buchi, New Castle, DE).

(f)

Magnetic stir plates and stir bars.—Dataplate 721 (Barnstead-Thermolyne, Dubuque, IA).

(g)

pH Meter and electrode.—WD-35618-03 (Oakton Instruments, Vernon Hills, IL).

(h)

Electrical conductivity meter.—HM Digital EC-3 (Amazon.com).

(i)

Spectrophotometer.—Dual beam 200 to 900 nm, with wavelength accuracy of ±1 nm and reproducibility of ±0.5 nm (Ultrospec II, G.E. Healthcare Sciences, Pittsburgh, PA).

(j)

Peristaltic pump and tubing.—Masterflex 7518-00, Pharmed 06508-17 (Cole Parmer, Vernon Hills, IL).

(k)

Combustion oven, i.e., muffle furnace.—Thermolyne Type 47900 Furnace (Fisher Scientific).

(l)

Rotating shaking mixer.—GlasCol 099A RD 4512, (A. Daigger & Co., Vernon Hills, IL).

(m)

Desiccator.—Vacuum type, KIMAX 21200-250, 250 mm (Capitol Scientific, Austin, TX).

Glassware

(a)

Erlenmeyer flasks.—1000 and 2000 mL.

(b)

Graduated cylinder.—1000 mL.

(c)

Glass chromatography columns.—4×25 cm for DAX-8 resin; 5×60 cm for IR120 cation exchange resin.

(d)

Ceramic combustion crucibles.—Sigma-Aldrich.

Analytical Procedure

Step 1. Alkaline Extraction

If the material is a dry solid, prepare analytical samples by crushing them to a fine powder using a mortar and pestle so that 100% of the crushed analytical sample passes through a U.S.

0.1 M NaOH. The test portion should result in a concentration of 200 to 600 mg/L HA plus FA after dilution with 0.1 M NaOH. The aliquot volume will be based on the HA and FA product concentrations that were claimed by the manufacturer. Determine the density (g/mL) of the liquid material by weighing 10 mL of well-mixed analytical sample in a pretared graduated cylinder (D1). Determine the weight of the liquid test portion using Equation 3 (Step 7).

0.015 absorbance units) to that of the deionized water used to wash the column, using deionized water to zero (i.e., blank) the spectrophotometer. A wavelength of 350 nm gives strong absorbance by FA and allows the use of a spectrophotometer that only measures visual wavelengths.

Desorb the FA by back elution (i.e., influent introduced into the bottom of column) by pumping 0.1 M NaOH using the peristaltic pump. Most of the FA is adsorbed to the very top of the DAX-8 resin. Desorption from the column bottom uses a minimal amount of 0.1 M NaOH to fully desorb the FA. All the FA has been desorbed when the absorbance of the column effluent is equal to the absorbance of influent at 350 nm. Use

0.1 M NaOH as the spectrophotometric blank. Add the effluent taken to check absorbance of the desorbed FA solution.

Protonate and de-ash the FA by passing repeatedly (by gravity feed) through Amberlite IR120 hydrogen form ion exchange resin contained in a 5×50 cm column until the electrical conductivity of the effluent is <120 uS/m as measured with a conductivity meter. To ensure that all the FA is removed from the resin after the final pass, wash the column with deionized water until the absorbance of effluent at 350 nm is the same (e.g., within 0.015 absorbance units) as the deionized water used to wash the column. Use deionized H2O as the spectrophotometric blank. Add the wash and any effluent portions taken to check absorbance to the purified FA solution. To help with removal of all FA, the resin can be agitated (e.g., using a long glass or plastic rod) several times.

Concentrate the FA to a volume of approximately 15 ± 2 mL by using a rotary evaporator at 55°C. Completely transfer the 15 mL fulvic acid concentrate to a 50 mL plastic centrifuge tube and dry at 90°C to constant dryness in a drying oven. Freeze-drying is an alternative to oven drying. After drying, as described for the HA above under Step 4, place the tube in a desiccator to cool. Remove FA from the tube by complete scraping of the tube sides and bottom with a spatula, and weigh it on pretared weigh paper (W8). This material is the “Extracted FA”. Determine the residual ash content of extracted FA as described under Step 4 for HA and calculate the ash ratio (Equation 4, Step 7). Finally, determine the weight of the extracted FA without ash using Equation 6 in Step 7.

Step 6. Column Regeneration

Regenerate the DAX-8 resin by pumping 0.1 M HCl [8.33 mL concentrated HCl/1000 mL final volume deionized (DI) water] through the bottom of the column until the pH of the effluent is equal to the pH of the influent. Use the peristaltic pump to pump all reagents through the DAX-8 column during regeneration. Next rinse the column with DI water by pumping it into the top of the column until the pH of the effluent equals the pH of the influent (i.e., DI water).

Regenerate the H⁺form cation exchange resin in a batch process by pouring the resin into a large beaker (e.g., 4 L plastic beaker), pour off the water, and cover the resin with 1 M HCl

(83.3 mL concentrated HCl/1000 mL final volume DI water). Let stand for a minimum of 30 min with occasional stirring (e.g., once every 5 min). Remove the excess acid from the resin by pouring off the acid and covering the resin with DI water. Stir vigorously with a stirring rod for 15 s, then let the resin and

rinse water sit for 5 min. Repeat the process until the pH of the rinse water equals the pH of the DI water.

Load the regenerated resin back into the column. Once loaded, rinse the resin with DI water and check the pH of the effluent. If it is still lower than the pH of the DI water before it is passed through the column, continue to rinse the column with DI water until the pH of the effluent is within 0.1 pH units of the pH of the influent.

Step 7. Calculations

Dry test portion dry weight = (W3) (1 – moisture ratio) (2)

Weight of extracted HA without ash (g) = (W4HA)(1 – Ash Ratio) (5)

Weight of extracted FA without ash (g) = (W4FA)(1-Ash Ratio) (6)

where W1 = weight of test portion taken for moisture plus ceramic dish before drying, g; W 2 = weight of test portion taken for moisture plus ceramic dish after drying, g; Wt1 = weight of ceramic dish, g; W3 = test portion weight, g; V1 = volume of liquid test portion, mL; D1 = weight/unit volume of liquid test portion, g/mL; W4HA = weight of extracted HA, g; W4FA = weight of extracted FA, g; Wt2 = weight of ceramic dish used for ashing, g; W5 = weight of extracted HA or FA taken for ash plus ceramic dish before combusting, g; and W6 = weight of ash plus ceramic dish after combusting, g.

(a) For solid materials, determine the percentages of FA and HA (dry weight basis) as follows:

(b) For liquid materials:

Materials and Methods

Validation Materials

Validation materials for this study included three commercial liquid humic products supplied by humic product vendors and four solid humate ores supplied by humate mining companies. The products were typical of commercial humic products that are distributed worldwide and raw materials for preparation of

Liquid test materials consisted of commercial products referred to as L16, L17, and L2. These liquid materials were analyzed in triplicate (data not presented) at an early stage of the NSM development, and their HA and FA concentrations were reported as percentage of initial sample mass. They were selected for their reported relative differences in concentrations of both HA and FA for the purpose of validating the analysis of liquid humic materials across a range of concentrations using the proposed method. The concentrations of HA and FA, respectively, in materials from preliminary analyses were: 16.5 and 1.7% (L16), 7.7 and 6.3%, (L17), and 3.9 and 9.8% (L2). L16 represented a product with relatively high concentration of HA with relatively low concentration of FA. L17 represented a product with medium concentrations of HA and FA, and L2 represented a product with relatively low concentration of HA with a relatively high concentration of FA.

IHSS Gascoyne Leonardite, which is rich in HA, was from a mine near Gascoyne, ND (IHSS Cat. No. 1BS104L). It was used as the HA standard and was supplied by the IHSS. The IHSS Gascoyne Leonardite is a mined ore that was ground, sifted, and dried. The FA standard was a commercial FA manufactured through a proprietary process by Hangzhou Dayangchem Co. Ltd, Hangzhou Zhejiang, China. The composition of the FA material was verified by ¹H-NMR and ¹³C-NMR spectrometry analysis (data not shown). In addition, IHSS Pahokee Peat standard FA (2S103F) and IHSS Leonardite standard HA (1S104H) were provided by the IHSS. They were prepared by the IHSS according to protocols described in http://www. humicsubstances.org/sources.html and verified by FTIR and

Table 1. Method precision in recovery of HA and FA from liquid commercial materials^a

a

Extraction conditions were 1 g which contained approximately 500 mg HA + FA/L 0.1 M NaOH.

effect on recovery of HA and FA from the D1 ore, the Gascoyne Leonardite, and the L16 liquid product. Extraction conditions were 1 g test portion in a final volume of 1 L 0.1 M NaOH for the dry samples and 0.600 g/L for the L16 liquid product.

(d) Effect of time and temperature on ash content.—In order not to overestimate the HA or FA content of any product that contains inorganic substances, ash content must be determined (8). The effects of temperature (400 and 500°C) and time (4, 8, 12, and 24 h) of ashing on the ash content determination were evaluated for HA generated from the experiment on pH variation. Humic acid extracted from the D1 ore, Gascoyne Leonardite, and L16 liquid were used.

Selectivity for HA and FA

Dark colored nonhumic materials are reportedly sold in the marketplace as humic products, and it is important that a method of analysis be specific enough to discriminate HA and FA from adulterants in commercial products. Potentially fraudulent materials tested were liquid seaweed (kelp) extract, NPK fertilizer, coal, molasses, and lignosulfonates.

(a)

Seaweed, NPK fertilizer, coal, and molasses.—The effects of these adulterants on the determination of HA and FA were studied using the Gascoyne Leonardite solid reference standard. The Gascoyne Leonardite reference standard was added to the matrix blank, i.e., 0.1 M NaOH, to produce a mixture with a final concentration of 2.5 g/L. Adulterants were tested individually and prepared by adding to the Gascoyne Leonardite/NaOH mixture in a final volume of 1 L 0.1 M NaOH. Adulterants added were seaweed (Neptune’s Harvest Organic Seaweed Plant Food), approximately 8 g/L; NPK fertilizer (Miracle Gro All Purpose Plant Food), approximately 2 g/L; coal (bituminous), approximately 2 g/L, and molasses, at approximately 2 g/L. These mixtures with and without adulterant were analyzed in duplicate.

(b)

Lignosulfonate.—The effect of adding lignosulfonate to liquid humic products was evaluated by preparing a liquid solution containing 1.8 g IHSS Gascoyne Leonardite HA and

0.25 g IHSS Pahokee Peat FA in 1 L 0.1 M NaOH with about 5 g lignosulfonate added. The resulting mixture was analyzed

Table 2. Method precision in determination of HA and FA contents of IHSS Gascoyne Leonardite standard and Hangzhou FA standard^a

Extraction conditions were 1 g sample/L 0.1 M NaOH.

for HA and FA contents using the NSM protocol for liquid materials.

Method Detection and Quantification Limits

Guidance for determining the method detection limit (MDL) for HA and FA was based on communications with the humic products industry and the fact that we had consistently extracted levels as low as 0.025 g/L HA and 0.020 g/L FA from several liquid products prior to this study.

The MDL is the lowest concentration of HA and FA that can be distinguished from a blank sample but cannot necessarily be quantified. It was calculated by multiplying the sample SD by the Student’s t-value. The MDL for HA was experimentally determined by preparing seven replicates of about 0.025 g/L of IHSS Leonardite standard HA (1S104H) in 0.1 M NaOH and subjecting them to the NSM. This freeze-dried HA standard had been extracted and purified from the IHSS Gascoyne Leonardite ore by the IHSS. The extractions were conducted by the same analyst on various days over a 1 month period. These

Table 3. Method precision in recovery of HA and FA from humic ore materials^a

Extraction conditions were 1 g sample/L 0.1 M NaOH.

1% Hangzhou FA 2.5% IHSS Gascoyne Leonardite, g Expected recovery, g^a Actual recovery, g Recovery, % IHSS Gascoyne Leonardite^b HA plus FA Hangzhou^b HA FA HA FAHAFAHA

Rep 1 4.367 1.753 1.432 3.262 1.397 3.342 97.6 102.5 Rep 2 4.427 1.758 1.439 3.330 1.432 3.471 99.5 104.2 Rep 3 4.378 1.724 1.414 3.293 1.392 3.340 98.5 101.4 Mean NA^c NA NA NA 1.407 3.384 98.5 102.7 SD NA NA NA NA 0.022 0.075 1.0 1.4

Expected recovery based on previous analysis of IHSS Gascoyne HA = 75.22% and FA = 7.5%, Hangzhou FA = 63%.

Similarly, seven additional 1 L 0.1 M NaOH solutions were spiked with about 0.020 g/L IHSS Standard Pahokee Peat FA (2S103F), and FA was determined by the same analyst on various days over a 1 month period. These FA-amended blanks in 0.1 M NaOH were used to determine the recovery of FA at low concentrations.

The RSD for HA content of the commercial liquid humic products concentration ranged from 1.39% for L17 to 4.91% for L2 (Table 1).The RSD for FA ranged from 3.8% for L17 to 10.4% for L2. As expected, the relative error was higher for FA overall, and the highest was for sample L2, the material with the lowest concentration. Using L2 as an example, at 0.38% FA, the mass of FA measured was only 4 mg. For the dry materials, FA and HA concentrations for the IHSS Gascoyne Leonardite were 7.6 and 75.2%, respectively (Table 2). The mean FA concentration for the Hangzhou FA was 63.0%, and no HA was detected as expected for this purified FA (Table 2). The RSD of the IHSS Gascoyne Leonardite FA and HA were 5.8 and 0.8%, respectively (Table 2). The RSD of the Hangzhou FA analyses was 0.6%.The RSD for FA in the mined ores ranged from 3.7 to 10.3%, and the RSD for HA ranged from 1.4 to 9.0% (Table 3).

Table 5. Ruggedness testing variable—initial sample mass^a

1 4.59 55.18 7.39 70.02 200 2.37 15.42

2.5 4.6 62.28 12.66 78.19 600 3.62 15.98

5 4.48 62.77 12.87 76.66

Table 6. Ruggedness testing variable—extraction time^a

a

Extraction conditions were 1 g sample/L of 0.1 M NaOH.

no effect was observed for FA recovery from D1. The results obtained from the addition of 2.5 or 5 g of these ores suggests varying quantities of HA + FA added according to the quantity in these samples will not affect the results. The least HA + FA added was 1.7 g for 2.5 g of the D1 ore sample and the largest addition of HA + FA was 4.5 g for the Gascoyne Leonardite. At the concentrations tested for the liquid L16 material, initial sample masses of 0.2 g/L 0.1 M NaOH and 0.6 g/L 0.1 M NaOH combined concentrations of HA and FA did not appear to affect recovery of HA or FA (Table 5). Therefore, the initial solids concentration for dry samples should be changed to a range of 1.7 to 4.5 g of HA + FA/L 0.1 M NaOH to determine HA content in dry samples. No change in the initial sample content for liquid samples is necessary, and it can remain in the range of 500 mg/L 0.1 M NaOH HA and FA combined. Liquid products have already been subjected to extensive extraction, most likely in KOH. Therefore, HA in liquid products is already in solution.

(b)

Extraction time.—There was little difference in the HA and FA concentrations observed after 6 or 24 h of extraction for the dry mined material D1 sample, suggesting that extraction times greater than 6 h had little or no further benefit for extraction of the dry mined material (Table 6).

The effect of extraction time was greater for the IHSS Gascoyne Leonardite with somewhat more HA and FA extracted after 24 h than after 6 h , but somewhat less than that reported for determination of extraction precision using a 6 h extraction time using an ore/NaOH ratio of 2.5 g/L 0.1 M NaOH (Table 5). These data support the recommendation of extracting solid materials for a minimum of 6 h.

(c)

Effect of pH on the acidified alkaline extract.—Of all the steps in the procedure, the final pH to which the alkaline extract was acidified had the greatest effect on HA and FA concentrations (Table 7). With a decrease in final pH from 2.0 to 1.0, the amounts of extracted HA increased while the amounts of extracted FA decreased for all three samples. Hence, lowering the pH of the alkaline extract from 2.0 to 1.0 resulted in more humic substances precipitating out of the alkaline solution as HA, which is consistent with the findings of Kipton et al. (14).

Table 7. Ruggedness testing variable—pH at which HA is

Conclusions

A standardized method, the NSM, for measuring the concentrations of HA and FA in commercial humic products and their source ores was evaluated for both liquid and solid materials. It features gravimetric determination of the HA and FA on a dry, ash-free basis; moderated concentrations of HA, FA, and salts during extraction to control the partitioning between HA and FA; and a definition of FA based on adsorption to DAX-8 resin. This definition, which corresponds to that used by the IHSS, is stricter than the classical definition of solubility in both acid and base, and it allows the distinction of FA from hydrophilic components present in the fulvic extract as well as potential adulterants. Through testing of reference standards and humic test materials provided by HPTA members, this procedure was found to be precise, reliable, and accurate. It was relatively rugged, as only one of four selected steps that were systematically varied clearly altered the amounts of HA and FA determined, namely the pH at which the HA was flocculated. The procedure proved to be highly selective with quantitative recovery of amended amounts of added humic products when added to blank solutions and complete distinction of humic products from four of five amended adulterants. The evaluation of MDL and MQL demonstrated that the method is sensitive enough for determination of low concentrations of FA and HA. This method is intended for routine regulatory and industrial use to establish and verify contents of commercial products. To compensate for its inability to distinguish FA from lignosulfonates, an alternative approach for detecting lignosulfonates based on elemental S concentration and FTIR spectrometry was proposed.

Acknowledgments

This study was entirely funded by the Humic Products Trade Association. We gratefully appreciate the assistance and reference materials donated by the IHSS, and gratefully acknowledge the valuable assistance provided by James Bartos, Office of Indiana State Chemist, the AAPFCO, and the SLV study monitors: Elaine Wong (CDFA), John Peterson (Wilber Ellis), Nancy Thiex (South Dakota State University), and Debra Wong (Oregon Department of Agriculture).

References

(1)

International Humic Substances Society (2013) http://www. humicsubstances.org

(2)

Swift, R.S. (1996) in Methods of Soil Analysis, Part 3. Chemical Methods, D.L. Sparks (Ed.), Soil Science Society of America, Madison, WI, pp 1018–1021

(3)

Stevenson, F.J. (1994) Humus Chemistry: Genesis, Composition, Reactions, 2nd Ed., John Wiley and Sons, Inc., New York, NY

(4)

Waksman, S.A. (1936) Humus: Origin, Chemical Composition, and Importance in Nature, Williams and Wilkins, Baltimore, MD

(5)

Oden, S. (1919) Kolloidchem Beih. 11, 75–260

(6)

Thurman, E.M., & Malcolm, R.L. (1981) Environ. Sci. Technol. 15, 463–466. http://dx.doi.org/10.1021/es00086a012

(7)

Leenheer, J.A., & Croúe, J.P. (2003) Environ. Sci. Technol. 37, 19A–26A. http://dx.doi.org/10.1021/es0264089

(8)

Hayes, M.H.B., & Graham, C.L. (2000) Humic Substances: Versatile Components of Plants, Soil and Water, E.A. Ghabbour & G. Davies (Eds), The Royal Society of Chemistry, Cambridge, UK, pp 91–109

(9)

Lamar, R.T., & Talbot, K.H. (2009) Comm. Soil Sci. Plant Anal. 40, 2309–2322. http://dx.doi.org/10.1080/00103620903111251

(10)

California Department of Food and Agriculture (1999) Humic Acid Method, Sacramento, CA

(11)

Mehlich, A. (1984) Comm. Soil Sci. Plant Anal. 15, 1417–1422

(12)

Van Zomeren, A., & Comans, R.N.J. (2007) Environ. Sci. Technol. 41, 6755–6761. http://dx.doi.org/10.1021/es0709223

(13)

Analytical Detection Limit Guidance and Laboratory Guide for Determining Method Detection Limits (1996) Wisconsin Department of Natural Resources, Laboratory Certification Program, PUBL -TS-056-96, Madison, WI

(14)

Kipton, H., Powell, J., & Raewyn, M.T. (1992) Anal.Chim. Acta 267, 47–54. http://dx.doi.org/10.1016/0003-2670(92)85005-Q

(15)

Tipping, E. (2002) Cation Binding by Humic Substances. Cambridge Environmental Chemistry Series 12, Cambridge University Press, Cambridge, UK. http://dx.doi.org/10.1017/ CBO9780511535598

(16)

Tanford, C. (1980) The Hydrophobic Effect: Formation of Micelles and Biological Membranes, 2nd Ed., John Wiley and Sons, New York, NY, pp 2–3

(17)

Prado, A.G.S., Pertusatti, J., & Nunes, A.R. (2011) J. Braz. Chem. Soc. 22, 1478–1483. http://dx.doi.org/10.1590/S010350532011000800011

(18)

FAO Chemical and Technical Assessment (2008) Food and Agriculture Organization of the United Nations, pp 40–65

(19)

Ekeberg, D., Gretland, K.S., Gustafsson, J., Braten, S.M., & Fredheim, G.E. (2006) Anal. Chim. Acta 565, 121–128. http:// dx.doi.org/10.1016/j.aca.2006.02.008