j ourna l homepage: www.e lsev ie r .com/ locate /c lay
Note
The influence of disaggregation procedures on soilgravitational separation
Enzo Salemi a, Umberto Tessari a,⁎, Nicolò Colombani b, Micol Mastrocicco a
a University of Ferrara, Department of Physics and Earth Sciences, Via Saragat 1, 44100 Ferrara, Italyb University “Sapienza” of Roma, Department of Earth Sciences, P.le A. Moro 5, 00185 Roma, Italy
Article history:Received 17 June 2013Received in revised form 10 June 2014Accepted 11 June 2014Available online 26 June 2014
Keywords:SoilParticlesSeparationDispersing agent
The use of dispersants in particle size analysis is a common practice, but this could cause bias on the gravitationalseparation of the different particle fractions in natural soil. The study highlights the results obtained in gravita-tional separation of silty and clayey fractions byusinghydrogenperoxide (H2O2) and sodiumhexametaphosphate((NaPO3)6) in different combinations. The efficiency of the different treatments was verified by comparisonagainst the results obtained on the same sediments without any treatment. The separation method is based onStokes law to calculate the settling time of particles in deionizedwater under controlled temperature. This meth-od was applied to three different agricultural soils of the Po River Plain. The sample treated with H2O2 and(NaPO3)6 at 4% showed the best results in terms of particle size degree of purity, ranging from 95% to 97% forclay and from 91% to 95% for silt. The degree of purity indicates the percentage of sediment with the particlesof the provided grain sizes.
Aggregates of soil particles, which are mainly formed as a result ofprecipitation processes of insoluble and soluble salts due to physicaland chemical action of organisms and to electrochemical effects onclay particles, can significantly affect the results of the distribution par-ticle size analysis (Matthews, 1991). For this reason using dispersants ordisembedding agents is a common practice during the preparation ofsediments for textural analysis (Evans et al., 2001; Psuty and Mureira,2000). The most used chemical agents during soil or sediment samplespretreatment are H2O2 and (NaPO3)6 solution. H2O2 is mainly used toremove organic matter and stimulate deflocculation of particles, while(NaPO3)6 is used to facilitate the dispersion of electronegative colloids.The effect of particle disaggregation and deflocculation due to the addi-tion of these chemical agents is useful both to improve the analyticalresponse of grain size distribution analysis and to minimize the errorsin the separation of silty and clayey fractions. This procedure can bevery useful in studies regarding absorption/desorption of organic andinorganic contaminants or in studies of microbial degradation on sepa-rate particle fractions (Allen andWalker, 1987;Wang and Keller, 2009).
Awide range of techniques of separation arewell known in the liter-ature (ASTM, 2007; Day, 1965; de Jonge et al., 2000; Gee and Bauder,1986; Huang et al., 1984). In addition, the separation of considerableamount of sediment (7% of dispersing phase) is possible and a high de-gree of purity of silty and clayey fractions can be achieved, by applying a
gravitational separation method based on the Stokes law with a strictcontrol of temperature (Salemi et al., 2010).
The aim of this study is to quantify the changes in particle size in-duced by sample pretreatment of the silty and clay fraction of agricul-tural top soils following the Salemi et al. (2010) procedure.
2. Materials and methods
The study was conducted on three different soils of the Po Riverlowland located in the Ferrara province (from here ITA, CCR and APO).These soil textures are very common in this area, overall they occupyan area greater than 50% of the province (Mastrocicco et al., 2010).According to the Wentworth classification (Wentworth, 1922) ITA iscomposed of 83% sand, 12% silt and 5% clay; CCR is 12% sand, 65% siltand 23% clay; and APO is 9% sand, 63% silt and 28% clay. The soil sampleswere collected in the first 10 cm of agricultural fields and their organicmatter content, determined by applying LOI method (Heiri et al.,2001), is 2.7% for ITA soil, 4.5% for CCR, and 3.2% for APO. In order toproceed with the different treatments, the soil samples, after mixingand quartation, were divided into four samples (named A, B, C and D).For each sample triplicate analyses were performed. The A samplesdid not undergo any treatment; B samples were subjected only to oxy-genation (H2O2 at 16 volumes) at room temperature until the completedisappearance of the effervescence; C samples were subjected to oxy-genation and addition of (NaPO3)6 solution at a low concentration(0.5% in volume) to limit the cation exchange with the soil and mini-mize the chemical contamination; and D samples were subjected to ox-ygenation and addition of (NaPO3)6 at 4%, as suggested byASTM(2007).
242 E. Salemi et al. / Applied Clay Science 97-98 (2014) 241–245
The sand fraction (N63 μm)was separated from the rest of the sedimentthroughwet sieving. Thefiner fraction, after the separation,was dried inan oven at 60 °C for at least 48 h in order to avoid any dimensionalchanges due to the extraction of water contained in the lattices of clayminerals. Then the fine fraction of samples A and B was rehydratedwith 0.2 L of deionized Milli-Q water, while the fine fraction of samplesC andDwas rehydratedwith deionizedwater andmixedwith (NaPO3)6at different concentrations. The complete dispersion of the sedimentwas ensured by stirring for 10 min using a magnetic stirrer. The disper-sion was introduced into a 30 cm high borosilicate glass cylinder withan internal diameter of 8 cm, filled with 0.8 L of deionized water.The procedure was repeated three times, introducing the dispersionon the water surface using a borosilicate glass funnel. At each cycle,only the dispersion containing the clay particles (b4 μm) was siphonedrespecting the settling time calculated by the Stokes law. The settlingtime was calculated taking into account the distance between thewater surface and the bottom of the cylinder. The fine fraction density,fundamental in the settling time calculation, was determined by aMicromeritics AccuPyc 1330 pycnometer (2.62 ± 0.03 g/cm3 for thesoil ITA, 2.68 ± 0.02 g/cm3 for CCR, and 2.66 ± 0.03 g/cm3 for APO).The water viscosity was kept constant using a thermostatic bath madeof a PVC tank (1× 0.6 × 0.6m)filledwithwatermaintained at a constanttemperature (20 ± 0.1 °C) by a heat exchanger (Resun CL450) and anelectric pump (Eden 140). The particle size distribution and size limitsof the separate fractions were carried out using a Micromeritics 5100X-ray absorption Sedigraph (Artigas et al., 2005).
3. Results and discussion
The size fractions separated by gravitation were analyzed to deter-mine the degree of grain size distribution purity. The results of thetests conducted on the treated samples (Fig. 1) show variable degreesof purity ranging from 85% to 97% in the clayey fraction and from 81%to 95% in the silty fraction. The untreated samples (A) after separationshowed purities ranging from 87% to 95% for clay and 85% for silt. De-spite the untreated samples (A) show high levels of grain size distribu-tion purity, as discussed below, significant differences in texturalfrequency distributions appear evident when compared with the treat-ed samples.
In particular, for the silty fraction, the adopted treatments show tobe efficient only if the oxygenation is associated with the action of(NaPO3)6. For APO and ITA soils, treatment with only H2O2 leads topoorer results. This can be attributed to the refractory organic matter,which resists to treatment with H2O2. This is typical of soils rich inclay minerals and under long-term agricultural use (Leifeld and Kogel-Knabner, 2001). However, in all soils, it is evident that the use of
Fig. 1. Degree of purity (%) of the different fractions was analyzed. Error bars indi
(NaPO3)6 at a low concentration (0.5%) does not induce any significanteffect. The highest values of dimensional purity are obtained only with(NaPO3)6 at a higher concentration added after oxygenation. Thesevalues do not significantly differ in the three soils analyzed. Thismeans that the action of the dispersant is not influenced by the originaltextural nature of the soil. Similar observations can bemade consideringthe degree of purity achieved in the clay fraction, for which, however,values are high also in the untreated samples A for APO and ITA soils.Also in this fraction the only oxygenation does not lead to significant im-provements, while once again, the highest purity is achieved with theconcomitant use of H2O2 and (NaPO3)6 at 4%. The effectiveness of thistreatment is also supported by the low variability observed in the ana-lyzed replicates. To better understand the action of the treatments in re-lation to the particle dispersion, the frequency grain-size distributionsof the different samples separate fractions, obtained via X-ray absorp-tion Sedigraph, can be considered.
Regarding the silty fraction (Fig. 2) the grain-size distribution ap-pears to be very similar in the three analyzed soils. The untreated sam-ples (A), the oxygenated one (B), and the oxygenated and treated with(NaPO3)6 at a low concentration (C) show very similar grain-size distri-butions, with an evident presence of particles with dimensions close tothe dimensional limit of the lower class considered, or even lower.
For this reason, their purity percentage is lower compared to sampleD. The latter shows a considerable decrease in frequency of particleswith dimensions close to the lower limit and the almost total disappear-ance of smaller particles are observed. This can be explained by betterdispersion of the particles and considerable decrease in electrochemicalaggregates. Accordingly, the grain-size classeswith higher frequency in-crease their relative importance. Similarly, analyzing the frequency dis-tribution of the clayey fractions, a significant change is observed only insamples D, for which the decrease in coarser particles is associatedwithan increase in finer particles (Fig. 3).
Comparing size distribution changes of the two fractions can be thusassumed that the electrochemical aggregates, with an average sizeranging from 7.5 to 9.5 phi and found both in the silty and clayey frac-tions, are realistically composed of clay particles smaller than 10.5 phi.The dispersion of these aggregates can be effectively achieved byusing a dispersantwith an adequate concentration. The dispersive effectof agents was also evaluated considering the textural parameters (Folkand Ward, 1957) of the analyzed fractions (Table 1). Regarding thesilt, the samples B and C do not significantly differ from the referenceuntreated sample A, while sample D shows a considerable variation,with a shift of themean diameter and themedian toward coarser valuesand a slight improvement of the sorting. Similarly, in the clay fraction,the consequences of the treatments are evident only in samples D. Inthis case, they show a significant decrease in the mean diameters and
cate the mean and ± the standard deviation of the triplicate measurements.
Fig. 2. Frequency distribution of silty fractions. Error bars indicate the mean and ±the standard deviation of the triplicate measurements.
in the medians. This is due to the greater dispersion of the smaller sizeparticles. This dispersion is also observed in the general reduction inthe values of sorting.
Overall, the greatest changes of textural parameters are observed inthe clay fraction. This implies that the particles belonging to these di-mensional classes aremost affected by aggregation phenomena and col-loidal formation. Because of ineffective, or sometimes discordant, actionof simple oxygenation, it is feasible that the bonds between the finersoil's particles have essentially an electrochemical nature and that thesmall particle aggregates reach a size that, on average, can also influencethe distribution of classes in the silty fraction.
4. Conclusions
The results obtained by gravitational separation based on Stokes lawunder controlled temperaturewere analyzed for three different agricul-tural soils of the Po River Plain. The silty and clayey fractionswere treat-ed using two of themost common chemical agents: H2O2 and (NaPO3)6.The efficiencywas verified by comparisonwith the same untreated sed-iments. In terms of particle size degree of purity, this method showedresults ranging from 85% to 97% in clay and from 81% to 95% in silt sam-ples. The separate analysis of clay and silt grain size distributions andthe textural parameter comparison showed that: (i) the simple reaction
with H2O2 does not affect the particle dispersion, especially in the siltyfraction of ITA and APO soils. In this case the dispersion is worse in com-parison with the same untreated samples; (ii) the combined use of ox-ygenation and (NaPO3)6 at a low concentration (0.5%) does not lead to asignificant increase in aggregate disruption in both silty and clayey frac-tion of the three different soils; (iii) the addition of (NaPO3)6 at 4% givesthe best results in terms of dimensional purity and this is evident in theanalysis of grain size distribution, where dispersion of aggregates is ob-served in both fractions for all soils. In the silty fraction the decrease inclay particles is associated with a relative increase in the classes withhigher frequency. In the clayey fraction the decrease in coarser particlesis associated with an increase in frequency of the classes ranging from10.5 to 12 phi.
References
Allen, R., Walker, A., 1987. The influence of soil properties on the rates of degradation ofmetamitron, metazachlor and metribuzin in soil. Pestic. Sci. 18, 95–111.
Artigas, R., Rodas, M., Sanchez, C.J., Mas, R., Dondi, M., Arribas, J., 2005. Clayey materialsfrom the Sierra de la Demanda Range (Spain): their potential as raw materials forthe building ceramics industry. Clay Miner. 40 (1), 25–41.
ASTM, 2007. D 422-63 Standard Test Method for Particle Size Analysis of Soils. AmericanSociety for Testing and Materials, West Conshohocken, United States.
Day, P.R., 1965. Particle fraction and particle-size analysis. In: Black, C.A., et al. (Eds.),Methods of Soil Analysis, Part I, Agronomy, vol. 9. ASA, Madison, WI, pp. 545–567.
de Jonge, L.W., de Jonge, H., Moldrup, P., Jacobsen, O.H., Christensen, B.T., 2000. Sorption ofprochloraz on primary soil organomineral size separates. J. Environ. Qual. 29, 206–213.
Evans, J.R., Kirby, J.R., Long, A.J., 2001. The litho and biostratigraphy of a late Holocene tidalchannel in Romney Marsh, southern England. Proc. Geol. Assoc. 112, 111–130.
Folk, R.L., Ward, W.C., 1957. Brazos river bar: a study of significance of grain size param-eters. J. Sediment. Petrol. 27, 3–26.
Gee, G.W., Bauder, J.W., 1986. Particle size analysis, In: Klute, A. (Ed.), Methods ofSoil Analysis: Part I, 2nd ed. Agronomy Monograph, 9. ASA and SSSA, Madison, WI,pp. 383–411.
Heiri, O., Lotter, A.F., Lemke, G., 2001. Loss on ignition as a method for estimating organicand carbonate content in sediments: reproducibility and comparability of results. J.Paleolimnol. 25, 101–110.
Huang, P.M., Grover, R., McKercher, R.B., 1984. Components and particle size fractions in-volved in atrazine adsorption by soils. Soil Sci. 28, 20–24.
Leifeld, J., Kögel-Knabner, I., 2001. Organic carbon and nitrogen in fine soil fractions aftertreatment with hydrogen peroxide. Soil Biol. Biochem. 33, 2155–2158.
Mastrocicco, M., Colombani, N., Salemi, E., Castaldelli, G., 2010. Numerical assessment ofeffective evapotranspiration from maize plots to estimate groundwater recharge inlowlands. Agric. Water Manag. 97 (9), 1389–1398.
Matthews, M.D., 1991. The effect of pretreatment on size analysis. In: Syvitski, J.P.M. (Ed.),Principles, Methods, and Application of Particle Size Analysis. Cambridge Universitypress, pp. 34–42.
Psuty, N.P., Mureira, M.E.S.A., 2000. Holocene sedimentation and sea level rise in the SadoEstuary, Portugal. J. Coast. Res. 16, 125–138.
Salemi, E., Tessari, U., Colombani, N., Mastrocicco, M., 2010. Improved gravitational grainsize separation method. Appl. Clay Sci. 48, 612–614.
Wang, P., Keller, A.A., 2009. Sorption and desorption of atrazine and diuron onto waterdispersible soil primary size fractions. Water Res. 43, 1448–1456.
Wentworth, C.K., 1922. A scale of grade and class terms for clastic sediments. J. Geol. 30,377–392.
