We moved the oil off the pavement. We did not make it stop existing.
Percy Anderson works on oil sorbents — the pads, booms and loose material that soak up a spill. His question is the one nobody asks at the scene: the bag goes on a truck. Where does the truck go? A sorbent that can only be used once is not a cleanup. It is a transfer.
He was ten years old in Biloxi in the summer of 2010, which means he did not watch the Deepwater Horizon spill on television. He watched it approach.
On 4 June Mississippi closed Petit Bois and Horn Islands — the barrier islands that stand between the open Gulf and the Mississippi Sound, which is to say between the oil and his town. On 6 June the governor of Mississippi said on national television that the state had “virtually no oil.” On 27 June tar balls came ashore on Mississippi beaches.
He did not have the words for it then. He has them now: he watched a hedge come off in real time, about his own water. The oyster beds did not recover. Mississippi has not had an on-bottom oyster season since 2018 — the spill, then the Bonnet Carré Spillway openings in 2011 and 2019, one after another on the same beds.
Everyone on that coast learned the same lesson that summer: the cleanup is not the end of the material. It is the part where the material stops being your problem and becomes someone else’s.
The islands close. The governor says virtually no oil. The tar balls arrive anyway. His father is at work on the storm drains, because his father is the one person in that conversation who knows exactly where the rainwater goes.
The gift: purpose. Open water, not land hydrology. Everyone else on the coast was angry at BP. He was thinking about the bags.
Changing his own oil. The drain pan slips, oil goes into a puddle on wet pavement. He grabs the roll of paper towels from the garage and presses it down.
The towel drinks the water. The oil sheen slides out from under it and spreads.
The gift: the physics. Cellulose is hydrophilic. Paper wants water. It will take the water first and leave the oil sitting on top, every single time, and no amount of pressing changes that.
A delivery spills in the lot. The employees use what is behind the counter — paper towels, a bag of clay oil-dry. A response crew arrives with proper pads and a boom, lays them out, soaks it up, bags it, and puts the bags on a truck. It is competent, fast, and correct.
He stands there doing the only calculation nobody else in the lot is doing: the rain is moving the sheen toward the storm drain, and the bag is going somewhere.
The gift: the question. Why is the pad single-use? Where does the bag go? If the answer is a landfill, what exactly did we clean?
His father spent more than thirty years as a stormwater drainage specialist for the City of Biloxi. Not a contractor. Not a consultant. The man who knows which inlet feeds which line and which line lets out where, because he has had his arm in most of them.
Growing up, the family dinner conversation included things like which streets flood first and why, and what a first-flush actually carries off a parking lot. It is a strange inheritance and it is the whole reason he is useful: he is the only person standing in that auto parts lot who is watching the drain instead of the spill.
Recover and reuse. Wring the sponge out, get the oil back, use the sponge again. Argonne’s Oleo Sponge; the carbon aerogels made from waste paper and cotton.
Digest in place. Load the sponge with nutrients, let bacteria eat the oil where it sits, walk away. The 2026 Ben-Gurion paper.
Nobody has published the hybrid. Not found in his searches — which is not the same as not existing, and he says so.
Both camps build their sponges out of waste paper. Same feedstock. Opposite endgame. One side spent a decade learning to get the material back; the other side is designing it to be left behind.
The working notebook.
Bench entries, including the ones that did not work. He keeps the failures in because the failures are where the thesis lives — a sorbent that works once is not news; a sorbent that still works on cycle forty is.
Read Karsagi Byron et al. properly for the third time and finally saw the thing that has been sitting in front of me since undergrad. They had to carbonize the cellulose before it would pick oil over water. Pyrolysis at 270, 360, 450 °C. Contact angle on the 450 °C product: 139°.
My paper towel in the driveway is in their methods section. It is the control. It is the version that fails.
Sat with Table 4 for an hour. Crude oil degradation after 21 days, their numbers:
NSCA 270 °C 69 ± 8%
NSCA 360 °C 58 ± 7%
NSCA 450 °C 62 ± 11%
270 °C degrades the most oil. They recommend 450 °C. That is not a contradiction — it is a trade, and they are open about it. The 270 °C aerogel is made below a critical point, it is less stable, it leaches phosphorus (0.31% of original mass over five weeks in artificial seawater), and in their own runs it changed. The 450 °C version holds together.
So the material that eats the most is the material that falls apart, and the material that lasts is the one you would have to go back and get. That is my whole thesis in one table and it is their table.
Important and easy to miss. The GC–MS numbers are linear alkanes. The paper says so itself — calculations were made on linear alkanes “as indicators for technical purposes.”
Linear alkanes are the easy fraction. They are what bacteria eat first. The aromatics, the resins and the asphaltenes — the tar, the part that is still on a rock in Prince William Sound — are not in that number.
This is not a criticism of the paper. It is standard practice and they flagged it. But it means “62% degraded” and “62% gone” are different sentences, and only one of them is true.
Did not work. Squeeze-regeneration on my own N/P-loaded samples. Cycle 1–3 fine on oil uptake. By cycle 4 the uptake is holding but the nitrogen signal is down badly on XPS. I am recovering the sponge and losing the thing that made it a bio-sponge.
Two readings and I cannot separate them yet: either mechanical squeezing is stripping surface nitrogen, or I damaged the samples handling them — they are milligram-scale and I am not gentle. Re-run with six samples and a proper fixture before I say anything about this out loud.
Ge et al. 2014 names my problem in its own abstract: sorbent consumption, and slow oil recovery. Their answer is to stop treating the sponge as a bucket and start treating it as a filter — pump through it continuously, and capacity stops being limited by how much sponge you brought.
If that works with a nutrient-loaded sponge, the retrieval problem gets much smaller, because the sponge never has to leave the water to be emptied. Nobody has tried it with a loaded sponge that I can find. Flagging that as a search result, not a fact.
Viscous crude does not go into a sponge quickly; that is the entire practical problem with real spills as opposed to diesel in a lab beaker. Ge et al. 2017 wrap the sponge in graphene and run current through it — Joule heating cut sorption time by 94.6% and sped recovery.
Solar-heated versions followed in 2018 and 2019. Note for me: every one of these is a recovery paper. Heat the sponge, get the oil out, use it again. None of them has anything living in it.
Who he checks his work against.
One source is a guess. Two sources is a hypothesis. Three is engineering. The standing question is taped to the shelf over his bench, written on the back of a receipt, and it has not changed since his first week.
Took him the Table 4 problem — that 270 °C degrades best and 450 °C is the one they recommend. Expected him to be interested in the chemistry. He was not.
His line: “You have found a trade-off, which is not a finding. Every engineer alive has found a trade-off. Tell me what you would build that does not have to make that trade, and then tell me why nobody has built it, and if your answer to the second part is ‘I did not find it in a search’ then say exactly that and stop dressing it up.”
Brought the cycle-4 nitrogen loss. He would not let me report it. “You have two samples and a hypothesis you like. That is how people publish things that are not true. Go get six.”
Then, on the way out, the part that actually landed: “You are not a materials chemist who happens to care about the ocean. You are an infrastructure person who found out the infrastructure ends at a landfill gate. Do not let anyone talk you into being the first thing.”
Obodat also advises Seika Chiyon in this department, on the polymer left after diaper recycling. He did not introduce us. He did something more useful: he pointed out that her Stage 1 separates fluff pulp — cellulose — and that cellulose is the feedstock for the carbon aerogels on both sides of my field.
Her material loves water. Mine has to be taught to hate it. Same wall, opposite sides: what is the second life of the material?
To be clear about what this is: nobody has published diaper fluff pulp as an oil-sorbent feedstock that I can find. It is a thread, not a result.
The Oleo Sponge came out of the same spill I grew up inside — built for the subsurface plumes nobody had a tool for in 2010, funded by the Coast Guard and BSEE, tested at Ohmsett on diesel and crude. Wring it out, use it again, hundreds of cycles.
It has been eight years. I have never once seen a reusable sorbent on the back of a response truck in Biloxi. The pads on that truck are the same pads. That gap — between what exists in a paper and what is in the box on the truck — might be the real thesis, and it is not a chemistry problem.
Dr. Etadol Obodat is a fictional faculty member. The research on this page is real, published and cited — see the Research Map. The work he argues with belongs to Karsagi Byron, Barak, Halpern, Kramarsky-Winter, Shelly, Azaria, Hayun, Sivan and Kushmaro at Ben-Gurion University of the Negev; to Barry, Mane, Libera, Elam and Darling at Argonne National Laboratory; and to Ge, Bi, Han and their co-authors. None of them has anything to do with this campus, and nothing on this page is attributed to them that is not in their papers.
From spill to second life.
Five stages between oil on the water and a sponge you can use again. Four of them are solved by somebody. The fifth is the thesis, and it is the one that makes the other four worth connecting.
Pads, booms, loose sorbent, skimmers. This is what is on the back of the truck today and it works. He does not reinvent this. He starts where the bag gets loaded.
A carbonized cellulose aerogel carrying bound nitrogen and phosphorus. The sponge concentrates the oil in its fibres and feeds the bacteria at the same time — and the nutrients stay put, which is what the Exxon Valdez fertilizer work could never manage in open water.
Their numbers: adsorption 78 ± 16 g crude, 81 ± 10 g diesel, 98 ± 6 g kerosene per gram — the authors note these came from roughly 2 mg samples and should be read as conditional, not as equilibrium capacities. Degradation after 21 days: 62 ± 11% at 450 °C against 34 ± 11% with no nutrients.
Read the caveats: 10 mL of artificial seawater, 30 °C, dark, rotating, deliberately inoculated with Pseudomonas. And the degradation figures count linear alkanes — the paper says so — not the aromatics, resins or asphaltenes.
If the sponge is going to be reused, it has to come back. Booms, nets, tethers, something magnetic — he has not researched this yet and will not pretend otherwise. It is the least glamorous stage and probably the one that decides whether any of it is real.
There is a way around it: Ge et al. 2014 pump oil through a sorbent continuously, which turns the sponge into a filter and means it never has to be retrieved to be emptied. Whether that survives a sponge with bacteria living in it is unknown.
This is the reuse camp’s decade of work. Argonne’s Oleo Sponge wrings out and goes back in, hundreds of cycles. Waste-paper carbon aerogel takes up to 188× its weight in pump oil and regenerates by squeezing or distillation. Joule-heating a graphene-wrapped sponge cut sorption time by 94.6% on viscous crude.
Every one of these was done on a sponge with nothing living in it.
Deploy a loaded sponge. Let the bacteria take the easy fractions down in place. Retrieve it. Squeeze, pump or heat the heavy remainder out. Put it back in the water.
Four questions he cannot answer yet, and neither can anyone else he has found:
— does biofilm clog the pores and kill the capacity?
— does distillation or combustion regeneration destroy the nutrient chemistry?
— and the one the papers make worst: the version that degrades best is the version that falls apart.
Karsagi Byron, D.L., Barak, H., Halpern, B., Kramarsky-Winter, E., Shelly, L., Azaria, N., Hayun, S., Sivan, A. & Kushmaro, A. (2026). Nutrient-supplemented carbonized aerogels for integrated oil adsorption and biodegradation in water. Chemical Engineering Journal Advances 27, 101391. Open access, CC BY-NC. Held and read in full.
Barry, E., Mane, A.U., Libera, J.A., Elam, J.W. & Darling, S.B. (2017). Advanced oil sorbents using sequential infiltration synthesis. J. Mater. Chem. A 5, 2929. — the Oleo Sponge.
Bi, H. et al. (2014). Carbon Microbelt Aerogel Prepared by Waste Paper. Small 10, 3544–3550. · Bi, H. et al. (2013). Adv. Mater. 25, 5916–5921. · Han, S. et al. (2016). Carbohydr. Polym. 136, 95–100.
Ge, J. et al. (2014). Pumping through Porous Hydrophobic/Oleophilic Materials. Angew. Chem. Int. Ed. 53, 3612–3616. · Ge, J. et al. (2017). Joule-heated graphene-wrapped sponge… Nature Nanotech. 12, 434–440. · Ge, J. et al. (2016). Advanced sorbents for oil-spill clean-up. Adv. Mater. 28, 10459–10490.
Named but not held here: the 2010 Deepwater Horizon dates and the Mississippi governor’s remark come from contemporaneous reporting and public timelines, not from a document in this folder. The Mississippi oyster figures are from the Mississippi Department of Marine Resources by way of local reporting. The EPA used-oil rule on wrung-out sorbents (40 CFR 279.10(c)) has not been checked against the regulation and is not asserted anywhere on this page.