Published online 20 September 2006
Published in Soil Sci Soc Am J 70:1932-1935 (2006)
DOI: 10.2136/sssaj2006.0043N
© 2006 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
Pedology Note
Refinement of the Differential Water Loss Method for Gypsum Determination in Soils
O. Artiedaa,
J. Herrerob,* and
P. J. Drohanc
a Univ. of Extremadura, 10600 Plasencia, Cáceres, Spain
b Soils and Irrigation Dep., Agri-Research Center of Aragon, P.O. Box 727, 50080 Zaragoza, Spain
c Pine Lake Institute for Environmental and Sustainability Studies, Hartwick College, Oneonta, NY 13820-4020
* Corresponding author (jhi{at}aragon.es)
 |
ABSTRACT
|
|---|
Determining the gypsum content in soil is lengthy and cumbersome using methods based on SO4 determination. Moreover, as these methods do not strictly titrate gypsum, inaccuracies can be produced by the presence of sulfate minerals other than gypsum. The thermogravimetric properties of gypsum, however, allow determination of its content in a rapid and easy way sufficiently accurate for many pedologic purposes. Our objective was to expand the lower limit of gypsum detection to make the test useful for soil classification and management. We have refined the differential water loss method by estimating the gypsum percentage from the loss of water in the soil sample between 70 and 90°C. Our results, compared with gravimetric determinations of precipitated BaSO4, found coefficients of determination of 0.98 for gypsum contents ranging from 2 to 50%, and 0.99 for contents >50%. The method is valid for gypsum contents >2% and thus improves previous estimation procedures.
|
INTRODUCTION
|
|---|
STANDARD METHODS for determining the gypsum percentage in soil rely largely on the total SO4 content of the soil. Examples of such methods can be found in Bower and Huss (1948), U.S. Salinity Laboratory Staff (1954), Lagerwerff et al. (1965), Hesse (1971), Loveday (1974), Skarie et al. (1987), Method 6F1 of National Soil Survey Center (1996), and Method 4E2a1a1 of Natural Resources Conservation Service (2004). Paralleling the terminology used for carbonates, Herrero (1991) and Herrero et al. (1992) proposed the expression equivalent gypsum content to describe the results of such determinations. Methods based on the loss of crystal water of gypsum, however, as is the case of Nelson et al. (1978), do not rely on SO4 determinations and thus estimate the gypsum content sensu stricto, without interference by other sulfate minerals contained in the soil sample.
The thermogravimetric properties of gypsum have been widely used for many years to determine gypsum content in samples of soil and other materials (Weiser et al., 1936; Vieillefon, 1979; Elprince and Turjoman, 1983); thermobalances are often used for accurate determinations. For soil taxonomic or management purposes and pedology research, however, more accessible and more affordable equipment can also yield acceptable results.
To determine gypsum content via thermogravimetric properties, methods rely on the measure of the crystal water content of gypsum (Ca SO4·2H2O). Typically, the value 20.91% (w/w) is used; however, because not all the water is recovered at the heating temperatures used in the laboratories, the concept of a recovery factor (Burns et al., 2002) is useful when describing the analytical methods based on the loss of crystallization water. Nelson et al. (1978) proposed a differential water loss method for gypsum content determination by measuring the water loss of a sample air dried in a silica gel box for 48 h or more, and then oven dried at 105°C for 24 h. They used a recovery factor of 19.42% and attained a correlation of 0.999 with the standard chemical method 6L1b of the Soil Conservation Service (1972). Nelson et al. (1978) realized the limit of the method in low gypsum content soils, however, and suggested reporting the content as <4% when the measured gypsum content of the soil is 1 to 4%.
Artieda (1993) used a recovery factor for gypsum of 13.93% between 40 and 105°C (standard error of 1.6%), improving on the estimate of Nelson et al. (1978) and producing a standard error of 1.8%; however, Artieda (1993) recommended that, due to the measurement's low reliability in soils with <8% gypsum, such soils should be reported as <8% rather than a specific gypsum percentage. Limiting gypsum content values only to samples with content >8% minimizes the usefulness of the measure for taxonomic purposes, considering that the threshold for the gypsic horizon stated by Soil Survey Staff (1999) is 5% gypsum content. Therefore, to find a more accurate measure suitable for soils with gypsum content as low as 5%, we refined the method of Artieda (1993) and make possible thermogravimetric soil gypsum content measures <5%.
|
MATERIALS AND METHODS
|
|---|
Soils
Soils were sampled from the Ebro Valley (Spain)-33 from pedons classified by Herrero (1991) as Typic Xerorthent and Gypsic Xerochrept, following Soil Survey Staff (1990) and the other four samples from lutites of Tertiary sediments-and do not contain gypsum as indicated by x-ray diffraction patterns (Artieda, 2004). As determined by gravimetric titration of SO4, by addition of concentrated HCl and precipitation as BaSO4 (Porta et al., 1986), the gypsum contents of the 37 samples range from 0 to 86%.
Procedure
Ten to 20 g of <2-mm air-dry soil was transferred to a tared Pyrex crystallizing dish and weighed to the nearest 0.001 g. (While 10 to 20 g will produce sufficient results, standard weights should be used and we suggest 20 g.)
The crystallizing dish containing the sample was placed in an oven at 70°C until constant weight was achieved, then the dish containing the sample was placed in an oven at 90°C until constant weight was again achieved. Constant weight at 70°C was reached in 
3 d, but the time depends mainly on the oven volume and the number of samples put in the oven. In our laboratory, a 100-L-capacity oven loaded with 100 samples reaches constant weight at 90°C in 48 h. The oven must have ventilation. After removal from the oven and before weighing, the sample was cooled completely in a desiccator.
The gypsum percentage in the sample was calculated by the following expression:
 | [1] |
where ws = weight of the sample dried at 70°C plus Pyrex crystallizing dish, wf = weight of the sample dried at 90°C plus Pyrex crystallizing dish, wt = weight of the Pyrex crystallizing dish, and 14.95 is the recovery factor of gypsum between 70 and 90°C.
A previous step in establishing the above procedure was to determine the recovery factors for the tested temperature intervals by analyzing the water loss of reagent-grade Merck gypsum at different temperature intervals. The recovery factors obtained were 14.95% for the interval 70 to 90°C, 19.10% between 70 and 105°C, and 19.66% between 70 and 150°C.
|
RESULTS AND DISCUSSION
|
|---|
Results from the crystal water loss method were compared for 37 samples whose equivalent gypsum content was also determined by gravimetric titration of SO4 (Porta et al., 1986). Equivalent gypsum contents of the samples ranged from 0 to 86%. Results from soil sample analysis using the temperature intervals 70 to 90, 70 to 105, and 70 to 150°C were plotted against the corresponding gravimetric titration data and regression lines were calculated to discover the best interval to determine gypsum content (Fig. 1
). The three regressions have r2 = 0.99 with a standard error of the estimate of 1.5, 1.4, and 1.5%, respectively.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 1. Gypsum content determined from gravimetry of precipitated BaSO4 as related to gypsum content from loss of its crystal water between three temperature intervals: 70 to 90°C, 70 to 105°C, and 70 to 150°C. Each point is the mean of three determinations per sample.
|
|
If we consider only the gypsum contents <8% after the gravimetry of precipitated BaSO4, the regression analysis for the three temperature intervals (Fig. 2
) shows the interval 70 to 90°C as the one with the highest r2. The standard error of the estimate for these regressions is 0.67% for the interval 70 to 90°C, 0.70% for the interval 70 to 105°C, and 1.04% for the interval 70 to 150°C.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 2. Regression of the gypsum percentage obtained by the proposed method on that obtained from gravimetry of precipitated BaSO4 method for contents <8%.
|
|
For these gypsum contents (<8%), a relative error was calculated using the expression:
 | [2] |
where GE = gypsum estimated with the proposed method, and GD = gypsum content determined from gravimetry of precipitated BaSO4. Figure 3
shows the relationship between the relative error and the gypsum content estimated from gravimetry of precipitated BaSO4. The lower relative errors occur in the 70 to 90°C interval, and the relative error increases as the gypsum content determined from gravimetry of precipitated BaSO4 decreases. We judged that the achieved error becomes unacceptable below 2% gypsum content.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 3. Relative error in the gypsum determination for gypsum contents <8%. The samples with gypsum contents = 0% are not included.
|
|
Figure 4
plots the gypsum contents by the proposed method against the gravimetry of precipitated BaSO4 for gypsum contents ranging from 2 to 8%. Again, the gypsum percentage obtained from water loss between 70 and 90°C yields the highest r2. The standard error of the estimate is 0.61% for the interval 70 to 90°C, 0.68% for the interval 70 to 105°C, and 0.99% for the interval 70 to 150°C.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 4. Regressions of the gypsum percentage determined by crystallization water loss in three temperature intervals on the gypsum determined from gravimetry of precipitated BaSO4 between 2 and 8%.
|
|
Based on these results, the temperature interval 70 to 90°C yields the most accurate estimation of gypsum by crystallization water loss. This suggests that between 70 and 90°C an important loss of gypsum's crystal water occurs simultaneously with a decrease of soil moisture.
To check this observation, we studied the evolution of weight loss against temperature in four samples (Fig. 5
): the first without gypsum, i.e., not detected by x-ray; the second and third with 0.34 and 82% gypsum, respectively (both determined by the gravimetry of precipitated BaSO4 method); and the fourth reagent-grade Merck gypsum, i.e., 100% CaSO4·2H2O. In Fig. 5, the drying curves of the sample without gypsum and with 0.38% gypsum have an inflection point at approximately 70°C; further increases in temperature resulted in negligible weight loss. Both the sample with 82% gypsum and the reagent-grade gypsum underwent a sudden weight loss at approximately 70°C, with the loss stabilizing at 90°C. We therefore conclude that this temperature interval is acceptable for determining a soil's gypsum content while minimizing interference due to soil moisture loss.
By maintaining the samples at 70°C until constant weight, and later at 90°C until constant weight, we derived an estimate of the percentage of crystal water weight loss for this temperature interval, referred to as the weight of the dry sample at 70°C. The recovery percentage for this temperature interval was 14.95. Gypsum contents (Fig. 6
) calculated in this manner were found to be highly correlated (r2 = 0.99) with those determined by gravimetry of precipitated BaSO4.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 6. Comparisons between the gypsum content (percentage by weight) estimated from gravimetry of precipitated BaSO4 and four intervals of gypsum contents estimated from loss of its crystal water between 70 and 90°C: (a) >2%, (b) <2%, (c) between 2 and 50%, and (d) >50%.
|
|
In the four lutite samples, which had a high clay content (36%)-mainly illite and chlorite-and no gypsum, we determined the remaining water at 70°C and between 70 and 90°C for comparison with the soil moisture content at 105°C. At 70°C, the water contents of the studied samples ranged from 0.0 to 0.3% (mean 0.2%), whereas within the interval 70 to 90°C, the water content estimates ranged from 0% to 0.2% (mean 0.1%). This means that our proposed method for determining gypsum content should yield a detectable gypsum content range of 0.0 to 1.4% (mean 0.5%), in spite of the true value of 0.0% gypsum. This result is due to the weight loss computed as crystallization water, which is in fact soil moisture. For samples with some smectite, there is a risk that the release of water from clay could affect the computed weight loss and overestimate the amount of gypsum.
With the proposed method, estimated gypsum contents <2% can be erroneous due to the weight loss probably being attributed to the remaining soil moisture; in these cases, a qualitative test is needed. Tests based on the precipitation of dissolved SO4 when adding acetone (Bower and Huss, 1948; Kunze, 1965) produce, in our experience, variable results because of sample flocculation and turbidity. An alternative test, used by Dultz and Kühn (2005), is the precipitation of BaSO4 after the addition of BaCl2. If some of these tests yield a positive result, the gypsum content should be reported as "<g%," where g is the value obtained by our proposed method.
Comparisons between the gypsum content (percentage by weight) estimated from gravimetry of precipitated BaSO4 and four intervals of gypsum contents estimated from loss of its crystal water between 70 and 90°C (Fig. 6) present some important implications of our method. For soils with <2% gypsum, the method cannot accurately estimate gypsum content. In Fig. 6b, a high amount of variability is seen for contents <1%. For soils with >2% (Fig. 6a, 6c, and 6d) gypsum, however, our method appears suitable.
|
CONCLUSIONS
|
|---|
Our procedure for determining the water loss of a soil sample between 70 and 90°C, referred to as the weight of the sample dried at 70°C, allows an accurate estimate of the gypsum content of a soil using a recovery factor of 14.95%. For soils determined by our method to have a gypsum content <2%, the presence of gypsum must be confirmed by a qualitative test for gypsum. If the qualitative test for gypsum is positive, we suggest that the gypsum content be reported as <g%, where g is the gypsum content obtained by our proposed method.
In a laboratory equipped with a 100-L capacity oven, the gypsum content of 100 soil samples can be determined in 5 d, with about 5 h of work.
Received for publication January 27, 2006.
|
REFERENCES
|
|---|
- Artieda, O. 1993. Factores geológicos que inciden en el desarrollo de los suelos en un medio semiárido. El caso de Quinto, Zaragoza. (In Spanish.) M.S. Thesis. Univ. of Zaragoza, Spain.
- Artieda, O. 2004. Materiales parentales y geomorfología en la génesis de Aridisoles en un sector del centro del Valle del Ebro. (In Spanish.) Ph.D. Diss. Univ. of Zaragoza, Spain.
- Bower, C.A., and R.B. Huss. 1948. Rapid conductometric method for estimating gypsum in soils. Soil Sci. 166:199–204.
- Burns, D.T., K. Danzer, and A. Townshend. 2002. Use of the terms "recovery" and "apparent recovery" in analytical procedures. Pure Appl. Chem. 74:2201–2205.
- Dultz, S., and P. Kühn. 2005. Occurrence, formation, and micromorphology of gypsum in soils from the central-German Chernozem region. Geoderma 129:230–250.[CrossRef][ISI]
- Elprince, A.M., and A.M. Turjoman. 1983. Infrared dehydration method for determining gypsum content of soils. Soil Sci. Soc. Am. J. 47:1089–1091.[ISI]
- Herrero, J. 1991. Morfología y génesis de suelos sobre yesos. (In Spanish.) Inst. Nacional de Investigaciones Agrarias, Madrid.
- Herrero, J., J. Porta, and N. Fédoroff. 1992. Hypergypsic soil micromorphology and landscape relationships in northeastern Spain. Soil Sci. Soc. Am. J. 56:1188–1194.[Abstract/Free Full Text]
- Hesse, P.R. 1971. A textbook of soil chemical analysis. Chemical Publ. Co., New York.
- Kunze, G.W. 1965. Pretreatment for mineralogical analysis. p. 568–577. In C.A. Black (ed.) Methods of soil analysis. Part 1. Agron. Monogr. 9. ASA, CSSA, and SSSA, Madison, WI.
- Lagerwerff, J.V., G.W. Akin, and S.W. Moses. 1965. Detection and determination of gypsum in soils. Soil Sci. Soc. Am. Proc. 29:535–540.
- Loveday, J. 1974. Methods for analysis of irrigated soils. Tech. Commun. no. 54. Commonw. Agric. Bur., Farnham Royal, UK.
- National Soil Survey Center. 1996. Soil Survey laboratory methods manual. Soil Surv. Invest. Rep. 42. Version 3.0. U.S. Gov. Print. Office, Washington, DC.
- Natural Resources Conservation Service. 2004. Soil survey laboratory methods manual. Soil Surv. Invest. Rep. 42. Version 4.0. U.S. Gov. Print. Office, Washington, DC.
- Nelson, R.E., L.C. Klameth, and W.D. Nettleton. 1978. Determining soil gypsum content and expressing properties of gypsiferous soils. Soil Sci. Soc. Am. J. 42:659–661.[Abstract/Free Full Text]
- Porta, J., M. López-Acevedo, and R. Rodríguez-Ochoa. 1986. Técnicas y experimentos en edafología (In Spanish.) Colegio Oficial Ingenieros Agrónomos, Barcelona, Spain.
- Skarie, R.L., J.L. Arndt, and J.L. Richardson. 1987. Sulfate and gypsum determination in saline soils. Soil Sci. Soc. Am. J. 51:901–905.[Abstract/Free Full Text]
- Soil Conservation Service. 1972. Soil survey laboratory methods and procedures for collecting soil samples. Soil Surv. Invest. Rep. 1. U.S. Gov. Print. Office, Washington, DC.
- Soil Survey Staff. 1990. Keys to Soil Taxonomy. 4th ed. SMSS Tech. Monogr. 19. Virg. Polytech. and State Univ., Blacksburg.
- Soil Survey Staff. 1999. Soil Taxonomy. A basic system of soil classification for making and interpreting soil surveys. 2nd ed. USDA-NRCS Handb. 436. U.S. Gov. Print. Office, Washington, DC.
- U.S. Salinity Laboratory Staff. 1954. Diagnosis and improvement of saline and alkali soils. Agric. Handb. 60. U.S. Gov. Print. Office, Washington, DC.
- Vieillefon, J. 1979. Contribution à l'amélioration de l'étude analytique des sols gypseux. (In French). Cahiers ORSTOM, sér. Pedologie 17(3):195–223.
- Weiser, H.B., W.O. Milligan, and W.C. Ekholm. 1936. The mechanism of the dehydration of calcium sulfate hemihydrate. J. Am. Chem. Soc. 58:1261–1265.[CrossRef]
TOP
ABSTRACT
INTRODUCTION
