Previous   |    Next

doi:10.2204/iodp.proc.337.203.2017

Materials and methods

All measured samples were unconsolidated sands based on visual core description and required no crushing to loosen individual grains. We measured bulk, untreated samples containing all lithogenic, biogenic, and authigenic components. These 28 unconsolidated sand samples were selected out of a set of 144 samples originally collected for a separate rock magnetic and geochemical study (Phillips et al., 2017) that covers the full range of major lithologies. The samples measured here are representative of the unconsolidated sands recovered from Hole C0020A. Collection of these samples was conducted to avoid contamination of drilling mud (Figure F2).

Each sample was treated with 20 mL of a 5.0 g/L sodium hexametaphosphate dispersing agent. Each sample was then agitated using a vortex mixer, left for a minimum of 24 h, and then agitated again immediately before analysis.

All samples were measured using Malvern Mastersizer 2000 laser diffractometer particle size analyzer with a Hydro 2000 wet dispersion unit. Our method is based on the approach of Sperazza et al. (2004). Grain size measurements were made using a particle refractive index of 1.55, a dispersant refractive index of 1.33, and a particle absorption index of 0.01.

All samples were passed through a 2 mm (#10 mesh) sieve before entering the dispersion unit, and no material >2 mm was observed on the sieve in any sample. Each sample was completely dispersed in the dispersion unit to avoid subsampling errors, which generally have a higher impact on sample uncertainty than obscuration rate (Sperazza et al., 2004). Samples in sodium hexametaphosphate were added to the dispersion unit and diluted to obtain an obscuration between 8% and 22% (average = 14%). The pump speed was run at 2000 rpm, the stirrer was run at 770 rpm, and the sample was sonicated for 10 s prior to the sample analysis.

Grain size distributions are reported as frequencies in 89 diameter bins from 0.01 to 2187 µm. Median, tenth percentile (d(0.1)), and ninetieth percentile (d(0.9)) were calculated from this distribution using the accompanying Mastersizer software. Mean grain size was calculated as both a surface area moment mean (Sauter mean diameter or D[3,2]) and a volume moment mean (De Brouckere mean diameter or D[4,3]). We calculate sand (>63 µm), silt (4–63 µm), and clay (<4 µm) from each sample’s grain size distribution based on the classification of Wentworth (1922).

The Mastersizer 2000 was calibrated with a Malvern glass sphere standard with mean diameter (d(0.5)) of 61 µm, tenth percentile (d(0.1)) of 37 µm, and ninetieth percentile (d(0.9)) of 90 µm. A local, well-sorted sand from Wallis Sands Beach in Rye, New Hampshire (US), was run four times throughout the run of samples as a check standard for consistency. This check standard repeated to within <1 µm for all mean and percentile statistics: D[3,2] = 242 µm, D[4,3] = 262, d(0.1) = 175 µm, d(0.5) = 251 µm, and d(0.9) = 362 µm.

One sample (337-C0020A-5R-3, 36–39 cm) was run in duplicate as a measure of repeatability across the range of obscuration (9% and 17%).

Top of page   |    Previous   |    Next