2002 Denver Annual Meeting (October 27-30, 2002)

Paper No. 5
Presentation Time: 2:30 PM


SPERAZZA, Michael1, MOORE, Johnnie N.2 and HENDRIX, Marc S.2, (1)Department of Geology, University of Montana, Missoula, MT 59812, (2)Department of Geology, Univ of Montana, Missoula, MT 59812, sperazza@selway.umt.edu

Traditional grain size measurement of sediments via settling tube and pipette methods are subject to significant precision error (ca. 25-40%), can quantify only a limited number of grain size classes and is time consuming. Until recently, quantifying grain size distributions in the sub-micron size fraction has required the use of SEM techniques, which is expensive and tedious. Advances in laser diffractometry over the past 20 years now allows researchers to rapidly measure grain size distributions of synthetic, compositionally homogeneous materials down to 0.02 µm with precision approaching ± 1%. However, applying this technology to natural sediments is complicated by the mixed compositional nature of the samples. Constituent minerals have different optical properties and variable surface defects that influence the output of the laser diffractometer, which utilizes Mie Theory to compile grain distribution from light energy.

We conducted a set of controlled experiments on a Malvern Mastersizer 2000 laser diffractometer to better understand the effects on grain size determination by variations in sampling techniques, preparation and optical parameters. These variables included subsampling technique, sonicification, dispersant concentration, refractive index, obscuration, and absorption. For these experiments, we used fine-grained (clay to silt) sediment from Flathead Lake, Montana. We found that pipette aliquots from a bulk sample were more variable and underestimated the median grain size vs. discrete direct (wet) or dry sampling. Varying the sample refractive index (RI) within the range of constituent minerals identified by independent QXRD analysis changed the observed median grain size by only 0.75%. Similarly, variations in obscuration resulted in <2% error for median grain size over a range of 1 to 20%, and <0.75% between 10 and 20%. In contrast, variations in mineral absorption values between 0 and 1 not only produced shifts in median grain size of up to 300%, but also created a false bimodal distribution in some cases. The precision and time-efficiency advantages of laser diffractometry can enhance grain size determination as an analytical tool. However, identifying the proper optical settings and preparation techniques is critical to produce representative data from lacustrine sediment.