The overall objective of this research was to develop an integrative passive sampler (IPS) design and protocol that accurately integrates time-weighted concentrations of munitions constituents (MCs) within water from epibenthic environments. Although integrative passive sampling has been demonstrated as a promising technique for MCs, current sampler designs and technology do not account for variations in sampling rate due to changes in flow and turbulences near the sampler. In complex environments at the sediment–water interface, this may limit the utility of passive sampling. Thus, the specific objectives were targeted at three different approaches to improve calibration of samplers including: 1) test options for optimization of passive sampler design to reduce boundary limitations and make sampling rates more uniform across environmental conditions; 2) modify and incorporate sensors for measurement of in-situ flow and temperature to correlate environmental conditions with lab calibrations; 3) test potential performance reference compounds to measure in-situ sampling kinetics allowing adjustments of sampling rates. Following improvement of measurements during varying flow, further testing could proceed in follow-on studies to test validity of the samplers in the epibenthic environment for measuring exposure to benthic and epibenthic organisms.

This study was divided into three tasks, each investigating an approach for improving the accuracy of sampling rates of IPS across changes in flow and turbulence.

In Task 1, researchers investigated whether modifications to IPS design could decrease flow dependency of the sampling rate. Researchers compared existing Polar Organic Chemical Integrative Sampler (POCIS) samplers, a design with demonstrated utility but susceptibility to flow variations, with two additional configurations they hypothesized would limit dependency of flow on sampling rate. Tested configurations included thickening the membranes of the sampler and placing a nylon screen 8mm from the membrane. Calibration studies of these design configurations were conducted using high and low flow conditions. Criteria for a successful sampling design was to have minimum differences in sampling rate between flows, yet high enough sampling rates to prevent loss of method sensitivity.

In Task 2, researchers investigated the use of performance reference compounds (PRCs) to adjust IPS calibration in situ allowing more accurate prediction of water concentrations during fluctuating flow conditions. POCIS samplers were manufactured containing PRCs including 13C-caffeine. Calibration studies were conducted under static and high flow conditions. The relationship of dissipation of the PRCs from the IPS and uptake of MCs were evaluated to determine if these compounds were acceptable as PRCs.

In Task 3, researchers investigated the potential for fitting an IPS with a microelectronic flow device. If flow is known, IPS calibration can be adjusted based on a series of calibration studies. A sensor was identified and tested for its potential in measuring flow on IPS both exposed directly to the external environment, and within deployment canisters typically used to protect IPS membranes from the harsh field environment. In addition, temperature sensitivity was determined and waterproofing techniques were investigated.

In Task 1, doubling the sampler membrane did not greatly reduce the impact of flow on sampling rate. However, the addition of the nylon screen did greatly reduce the impact of flow on sampling rate. Changes in sampling rate between static and high flow were between 200-500% in traditional POCIS depending on analyte. With the screen in place, changes due to flow were less than 40%. Sampling rates were reduced with the addition of the nylon screen, but not enough to greatly impact sensitivity. Placement of a nylon screen over the sampler requires limited technology improvement and it can be placed over commercially available samplers allowing rapid technology transfer.  

In Task 2, PRCs resulted in improved calibration across flows for most analytes; however, a few important analytes such as RDX were not improved. For TNT and other nitrotoluenes, corrected sampling rates were 73-122% between flow conditions. However, for RDX, sampling rates varied by 191-206%, resulting in a similar error potential that existed due to flow rate without using a PRC. Despite some promise, technology transfer for the PRC approach will not be easy. The commercial provider appears to be reluctant as there is not a consensus choice of PRC and the best PRC is likely analyte specific. Addition of compounds to the sampler would likely be a custom process. Moreover, the PRC approach will require additional analyses cost as most PRCs will not be on the target analyte list.     

In Task 3, an electronic sensor that was sensitive to water flow was found, successfully water-proofed, and calibrated. The influence of temperature was described and can be easily adjusted. The device was highly sensitive across low flow conditions, which matched the flow rates that were most impactful to POCIS sampling rates. The device was successfully deployed inside and outside of a POCIS deployment canister demonstrating that a small device could be deployed inside the canister. Moreover, there were flow differences between the inside and outside of the deployment canister demonstrating that an internal device was necessary. The device could be broadly applicable to numerous aquatic studies even beyond passive sampling. Thus, researchers expect interest in commercial transfer. 

In conclusion, the addition of a nylon screen over commercially-available samplers and placement of a flow sensor were found to be promising approaches to reduce the impact of flow on IPS sampling rates. With minimal further development, these approaches will be ready to test in mesocosm and field studies within the epibenthic environment. In these studies, researchers will be able to further optimize sampling approaches for unexploded ordnance (UXO) sites. 

The improved ability to accurately sample and quantify concentrations of MCs, and other moderately polar organic contaminants, in the epibenthic environment provides a valuable tool for monitoring potential low-level and/or episodic releases at UXO sites, enhancing existing capabilities with POCIS for the water column. It is anticipated that the possibility of employing the highly promising POCIS technology to multiple environments, including open water and epibenthic zones in fresh and salt water, will provide more options to DoD end users associated with the Military Munitions Response Program (MMRP). Thus, decision making will be improved with respect to potential concerns regarding the need for removal or other costly remedial actions. For example, if detection of buried and leaking UXO is possible through epibenthic measurements, sampling and analysis of sediment and pore-water with unknown UXO sources can be minimized.