Mobility of Dissolved Organic Matter from the Suwannee River (Georgia, USA) in Sand-Packed Columns

Abstract

Transport of dissolved organic matter (DOM) from the Suwannee River and of synthetic polystyrene sulfonates (PSSs) was investigated in columns packed with naturally Fe/Al-oxide-coated sands from Oyster, Virginia. Surface-water samples were collected in May 2012 and processed by XAD-8 (humic substances; HPOA), XAD-4 (transphilic acids [TPIAs]), and reverse osmosis (broad range of components; NOM). Median transport time ( ) of PSSs increased with molecular weight (MW) from 1,000 to 8,000 Da but decreased for the largest PSS (18,000 Da), which is consistent with previous observations of MW effects on DOM adsorption and transport. Breakthrough curves (BTCs) of HPOAs and NOM were similar whereas TPIA transport was distinct; although all DOM samples had similar values, BTC asymmetry and dispersivity were greater for TPIAs. All samples exhibited power-law tailing that is characteristic of heterogeneous sorbent/sorbate interactions, potentially including kinetic effects. The one-dimensional advection-dispersion equation was unable to capture the tailing but it was captured well using a continuous-time random walk (CTRW) model. CTRW parameters were similar for the NOM and HPOA samples but distinctly different for TPIAs, which had more pronounced tailing. While retardation of organics generally tends to increase with MW, the lower average MW of TPIAs did not result in decreased overall retardation, which suggests the importance of compositional differences. Results suggest that while TPIAs tend to be a relatively minor component of DOM, their transport behavior differs from that of the predominant HPOA fraction, and they might thus have different impacts on pollutant transport.

Modelling the transport of environmental DNA through a porous substrate using continuous flow-through column experiments

“…Natural organic matter (NOM), ubiquitous in natural environments, is well known to be polydisperse, being made up of molecules with broad ranges in size and molecular weights. While similar work has been done studying the effect of polydispersivity on transport and how best to model it in the context of NOM [45][46][47], it offers simpler interpretation as while it is polydisperse, it can typically be treated as homogeneous; that is, in a well stirred bucket samples of NOM will all yield very comparable concentrations [25]. Additionally, the size distribution in NOM is such that most molecular weights fall into the range where they can be well described as a solute [47].…”

FracFit: A robust parameter estimation tool for fractional calculus models

“…Generally, the MAR is comparable to those obtained by the CTRW in McInnis et al . []. Our goal is not to compare CTRW and FADE model fits but rather demonstrates FracFit's ability to interpret a continuous injection anomalous transport breakthrough curve, which is clearly shown here.…”

“…As a second example, we fit continuous injection breakthrough curves (CBTCs) from laboratory experiments. These experiments studied transport of organic matter through porous media columns and displayed strong anomalous transport characteristics [ Dietrich et al ., ; McInnis et al ., ]. The data were originally fit with a CTRW model using the CTRW toolbox [ Cortis and Berkowitz , ].…”

“…Two data sets are considered: (1) synthetic polystyrene sulfonates (PSSs) in columns packed with naturally Fe/Al‐oxide‐coated sands from Oyster, Virginia [ McInnis et al ., ] and (2) dissolved organic matter (DOM) from Nelson's Creek, MI, in a column of porous medium (oxide‐coated quartz sand) [ McInnis et al ., ]. Both are continuously injected through sands via a gravity feed system with concentration measured at the outlet.…”

“…The fitted parameter for the sFADE and for the TFDE are shown in Tables and , respectively, along with the mean absolute residual (MAR), allowing comparison with the CTRW model fits from McInnis et al . [].…”