Brendan Smith1 Methely Sharma1 2 Jeffrey Grossman1

1, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States
2, University of Waterloo, Waterloo, Ontario, Canada

Over 250 billion gallons of fracking produced water (PW) are brought to the surface each year in the US alone. This water is generally contaminated with total dissolved solids (TDS) less than 2 ┬Ám in diameter at levels as high as 170,000 mg/L, over 300 times the recommended TDS limit for potable water. 60% is permanently lost to disposal well injection while much of the remainder is stored or released at the surface, often adversely impacting surrounding ecosystems and posing a major health hazard. Recently, treatment of PW by Nanofiltration (NF) membranes at the site of production has received attention as a potential solution to facilitate water recycling or release into the environment, offering high throughput, lack of requirement for chemical additives, and small footprint compared to alternative treatment approaches. Unfortunately, current state-of-the-art polymer NF membranes lack the thermal and chemical stability, as well as anti-fouling capabilities necessary to interface with fracking PW, while commercial ceramic NF membranes are prohibitively costly and not well suited to reject contaminants with sizes below 20 nm.

In order to tackle this challenge, we have developed a large-area fouling-resistant nanoporous silicon (NPSi) membrane capable of removing nanoscale contaminants, while offering permeability similar to current NF membranes. The NPSi membrane can be manufactured in a simple and scalable manner using a solution-based catalyzed etching process, which is capable of producing nanopores with diameters less than 5 nm and aspect ratios greater than 2000:1 through crystalline silicon foil with micron-scale thickness and diameter of 3 cm. Subsequently, the use of atomic layer deposition (ALD) is demonstrated as an effective approach both for tuning membrane pore size with sub-nm resolution, and also for modifying the membrane surface and pore wall to maximize permeability, contaminant rejection, and resistance to fouling via hydrophilic character. The morphological and chemical properties of the resulting membranes are characterized via electron microscopy, x-ray, BET, and contact angle analysis. Filtration performance is evaluated using a dead-end filtration setup with simulated cross-flow via stirring. Test solutes used to evaluate rejection capability include gold nanoparticles and molecular dyes. Chemical and thermal robustness are demonstrated by ex-situ incubation in elevated temperatures and high concentrations of chlorine and sulphuric acid, while anti-fouling capability is studied via prolonged filtration testing in the presence of common foulants. Finally, a preliminary cost model for implementation of the developed membranes is presented to elucidate the potential for penetration into the PW treatment industry.