New advance in clean energy.
December 7, 2019 4:27 PM   Subscribe

There's been an advance in generating power from the natural mix of fresh and salt water. The idea has been around for quite a while. In 2013 a French Team made a membrane that could do that. It's a completely passive effect. They use Boron Nitride Nanotubes embedded in a Silicon Nitride membrane. The charge on the Nanotubes preferentially sorts the positive and negative charges of the salt water and only lets one polarity through to the fresh water side. This creates a Voltage difference across the membrane. The French Team's estimate was:
"...researchers estimated a single square meter of the membrane—packed with millions of pores per square centimeter—could generate about 30 megawatt hours per year."

Great idea but the struggle has been to get the Boron Nanotubes all lined up across the Membrane. One way of doing that has just been found.

Their technique is showing better than the French Team's on the first prototype:

***************
When the researchers placed their membrane in a small vessel separating
salt- and freshwater, it produced four times more power per area than
the previous French team’s BNNT experiment. That power boost, Shan says,
is likely because the BNNTs they used are narrower, and thus do a better
job of excluding negatively charged chloride ions.

And they suspect they can do even better. “We’re not exploiting the full
potential of the membranes,” Cetindag says. That’s because only 2% of
the BNNTS were actually open on both sides of the membrane after the
plasma treatment. Now, the researchers are trying to increase number of
open pores in their films...
posted by aleph (53 comments total) 33 users marked this as a favorite
 
Very exciting.
posted by No Robots at 4:40 PM on December 7, 2019


Is this anything substantive beyond typical popular science press release?
posted by odinsdream at 5:00 PM on December 7, 2019 [6 favorites]


Something really bugged me in the formulation "30 megawatt hours per year." It was the hours per year part. That's a pretty nutty unit. If you work it through they are claiming (assuming my math is right!) 3435 watts per square meter.

And you know what? I don't believe that at all.
posted by sjswitzer at 5:04 PM on December 7, 2019 [11 favorites]


I don't really think in "megawatt hours per year per square meter" so I did the math and it's 3.4 kilowatts / m². To match a typical 600MW coal fired power station would take a square sheet 420 meters (1/4 mile) on a side. Of a 6.5 micrometer thick film. That means lots of framing structure and concrete pylons and stuff, in some of the most valuable ecosystems in the world.

Every bit counts, for sure, but I kinda want all that estuary to go back to being interstitial wetlands and spawning beds and stuff like that?

Sunlight is only 1KW / m², but we don't need any new science, we could start building in the desert today.
posted by Horkus at 5:06 PM on December 7, 2019 [27 favorites]


Psych, Horkus
posted by sjswitzer at 5:11 PM on December 7, 2019 [1 favorite]


But you know, if you're willing to deal with all the difficulties of saltwater and extreme weather and all--and they are plenty!--harvesting wave energy and tidal energy seems really promising. Except that it's very very hard or we'd be doing (more of) it already.
posted by sjswitzer at 5:17 PM on December 7, 2019 [2 favorites]


I wonder how much the energy this generates (at scale) would change the temperature and salinity of ocean water. Warming oceans get more acidic as they absorb more CO2, and filtering salinity along a gradient tends to concentrate (toxic) heavy metals, as desalination plants have shown. Both of these issues have consequences for life, real estate concerns aside.
posted by They sucked his brains out! at 5:26 PM on December 7, 2019


Renewable energy technologies that are 10 years in the future get a lot of love in the press, and have since the 60s. Whereas technologies available now are disparaged in novel ways. I wonder if there's something sinister behind that (hahaha of course there is).
posted by Horkus at 5:30 PM on December 7, 2019 [23 favorites]


1. So, according to the article we can get about 2000 nuclear power plants worth of electricity by... diverting all of the fresh water to salt water rivers and estuaries? Everywhere? That sounds like a really bad, disastrous bargain.

2. The sea eventually eats everything. Everything.

3. This sounds like a great way to further pollute the sea with microplastics and, now, untested nanoengineered materials. I bet you these things are great at messing with sea life which tends to rely on stuff like, oh, ion transport as part of different biological processes.

4. At some point in using this technology the use of fresh water is going to economically run up against potable drinking water. What happens when municipalities ban rain gardens or rain collecting barrels? Some already do for no good reason. What happens when ground aquifers are seen as a profit center and power source? I mean, even more than now?

Commodity and community solar power now, please. It's now a mature technology. We can actually meet our energy needs.
posted by loquacious at 5:42 PM on December 7, 2019 [15 favorites]


Energy is weird. None of it is being made;it only changes form. Every source of energy changes the state of the world in such a way that it takes MORE energy to undo. This cannot be different. What about this changes the state of the world? Are we sure that, whatever that is, it is doing no harm?
posted by JHarris at 5:42 PM on December 7, 2019 [4 favorites]


This will work out. Or not. I think the technique to straighten out all these nanotubes and line them up across the membrane to form this structure will have more uses than this.
posted by aleph at 5:43 PM on December 7, 2019 [1 favorite]


I think the technique to straighten out all these nanotubes and line them up across the membrane to form this structure will have more uses than this

I don't know what you're alluding to here, but if you write this SF story I will definitely read it.
posted by sjswitzer at 6:04 PM on December 7, 2019


"I don't know what you're alluding to here..."

From the article:
******************
But creating even postage stamp–size films has proved impossible, because no one has figured out how to make all of the long, thin BNNTs line up perpendicular to the membrane. Until now.

At the semiannual meeting of the Materials Research Society here yesterday, Semih Cetindag, a Ph.D. student in the lab of mechanical engineer Jerry Wei-Jen Shan at Rutgers University in Piscataway, New Jersey, reported that their team has now cracked the code. The nanotubes were easy. Cetindag says the lab just buys them from a chemical supply company. The scientists then add these to a polymer precursor that’s spread into a 6.5-micrometer-thick film. To orient the randomly aligned tubes, the researchers wanted to use a magnetic field. The problem: BNNTs aren’t magnetic.

So Cetindag painted the negatively charged tubes with a positively charged coating; the molecules that made it up were too large to fit inside the BNNTs and thus left their channels open. Cetindag then added negatively charged magnetic iron oxide particles to the mix, which affixed to the positively charged coatings.

That gave the Rutgers team the lever it was looking for. When the researchers applied a magnetic field, they could maneuver the tubes so that most aligned across the polymer film. They then applied ultraviolet light to cure the polymer, locking everything in place. Finally, the team used a plasma beam to etch away some of the material on the top and bottom surfaces of the membrane, ensuring the tubes were open to either side. The final membrane contained some 10 million BNNTs per cubic centimeter.
posted by aleph at 6:13 PM on December 7, 2019 [2 favorites]


> Rivers dump some 37,000 cubic kilometers of freshwater into the oceans every year. This intersection between fresh- and saltwater creates the potential to generate lots of electricity—2.6 terawatts, according to one recent estimate

I don't have any intuition how much 2.6 terawatts is: for reference, current global primary power production is allegedly roughly 14 terawatts, with world electricity generation comprising about 3 of those 14 terawatts. So producing 2.6 terawatts of energy is a fairly bold claim, and it'd be good to see a bit of the working behind that recent estimate! cf. wikipedia / world energy consumption

If you want to break down energy use into person-sized daily consumption estimates, without hot air gives a good yardstick: Sustainable Energy – without the hot air: Chapter 18 Can we live on renewables?
posted by are-coral-made at 6:17 PM on December 7, 2019 [2 favorites]


This is fantasy at this point. Hopeful fantasy, but still. Nailing numbers down is way too early. (But it's fun to play around)
posted by aleph at 6:20 PM on December 7, 2019 [1 favorite]


It does seem like we should take this with a grain of salt.
posted by mubba at 6:24 PM on December 7, 2019 [17 favorites]


Hypertension salt levels at this point. The trick with the nanotubes is as importantly to me as this application. Which I believe *is* more important for the future. We'll see.
posted by aleph at 6:33 PM on December 7, 2019


Something really bugged me in the formulation "30 megawatt hours per year." It was the hours per year part. That's a pretty nutty unit.

Megawatt hours is units of energy generated / converted / stored / consumed. "30 megawatt hours per year" means you collect that much energy total from a square meter of the stuff over the course of a year. 10,000 hundred watt light bulbs would convert 30 megawatt hours (a year's worth of energy, if it were stored) into light and heat in 30 hours.
posted by ZenMasterThis at 6:51 PM on December 7, 2019 [1 favorite]


It still comes out to 3.4 kilowatts per square meter, which is implausible.
posted by sjswitzer at 7:05 PM on December 7, 2019


Something really bugged me in the formulation "30 megawatt hours per year." It was the hours per year part. That's a pretty nutty unit.

It is nutty. It's watts multiplied by one time unit and divided by a different time unit. As a couple of people have pointed out above, the time units cancel out and it reduces simply to watts. Why the weird formulation?

As usual, you wonder if this is directly from a published paper or the creation of a poorly informed science writer.
posted by JackFlash at 7:11 PM on December 7, 2019 [2 favorites]


This would be wizard in the exact right site circumstances. If one lived on seashore and had a source of fresh water hitting the sea on your property you could take a sheet of this stuff and bend it into a tube. Mostly submerge the tube in the ocean and then use a pipe to direct your stream into the tube. Your limit on power generation is length/radius of the tube and how much fresh/salt water it requires.

The article doesn't talk about what happens to the ions after electricity is generated (do they continue to float around or do they precipitate out?) and it doesn't talk about what happens when the the water from the two sides mix afterwards.

However I'd imagine keeping the membrane free of plant growth is going to be difficult.
posted by Mitheral at 7:20 PM on December 7, 2019


Much older article but just love the intro sentence: Boron nitride nanotubes (BNNTs) are the divas of the nanoworld.

These were previously suggested to be useful as a super raincoat. they shed water like a duck’s back, a quality known as the lotus effect. “Water just slides away,”

That was 10 years ago, and they'd been working on the problem for at least a decade. Have not seen the perfect waterproof umbrella on the market yet. I guess nano is hard.
posted by sammyo at 7:30 PM on December 7, 2019


Being able to create aligned nanotubes seems like it should enable a bunch of other important unforeseen technologies. Some cellular processes work by pushing stuff across a membrane in the right direction, or allowing stuff that REALLY wants to cross do so... for a price. For example, mitochondria build up a pH difference across a membrane, and then harvest energy from letting the H+ ions cross back.
posted by Jpfed at 7:37 PM on December 7, 2019 [3 favorites]


The abstract is on page 979 of this rather massive PDF, so I take the liberty of posting it in its entirety below:
Recent nanofluidic experiments with single or few nanopores in graphene, molybdenum disulfide and hexagonal boron nitride have shown unique fluidic transport properties and the potential for electrokinetic energy conversion with unprecedented power densities. In such nanopores, the high-surface charge makes possible a diffusio-osmotic mechanism for ion-selective transport, distinct from the Donnan exclusion that is typical of conventional membranes. In particular, experiments with single boron nitride nanotubes (BNNTs) have reported large surface charge in aqueous solution, and osmotic power densities up to several kW/m2 when extrapolated to macroscopic membranes. However, no such macroscopic BNNT membranes have ever been fabricated, and their performance at large scales and pore densities is unknown. Thus, we seek to devise scalable means to manufacture such large-area nanotube membranes, and to investigate their fundamental mechanisms and performance for ion selectivity and electrokinetic energy conversion.

Here, we describe the fabrication of the first-ever macroscopic vertically aligned- (VA-) BNNT membranes, and our study of their ion-selectivity mechanism and osmotic-power-generation performance. The membranes are fabricated with a unique solution-based technique in which BNNTs are aligned and concentrated in a liquid oligomer with an external field, and then locked in place by in-situ polymerization. With this scalable fabrication technique, we have made VA-BNNT membranes of size 1-10 cm2 and having tube densities of 108-109 BNNTs/cm2 and open pore densities on the order of 107 pores/cm2. We show that, due to the high surface charge in their pores, the BNNT membranes are highly cation-selective even when the Debye length is smaller than the inner radius of the nanotubes. Moreover, the membranes exhibit energy-conversion efficiencies of 30%, and have osmotic power densities (based on open pore area) comparable to and even exceeding that of single BNNTs, up to 7,500 W/m2 at pH 11 and 1 M:1 mM KCl molarity difference. To elucidate the mechanisms for ion selectivity and osmotic-power generation, we compare the membrane performance for different salts and for few-wall and multi-wall BNNTs of different diameters. Notably, these scalably fabricated VA-BNNT membranes, with relatively large yet highly selective pores, have enhanced electrokinetic-energy-conversion performance, and may enable efficient energy harvesting from salinity gradients, as well as efficient desalination and other separations via a different mechanism than conventional membranes.
posted by Not A Thing at 7:40 PM on December 7, 2019 [4 favorites]


As a couple of people have pointed out above, the time units cancel out and it reduces simply to watts. Why the weird formulation?
TFA is not written for EE people, but for the layman whose only frame of reference might be the kWh on their electric bill.
posted by Horkus at 7:44 PM on December 7, 2019 [9 favorites]


But I don't pay my electric bill annually so I would still need to divide by twelve.
posted by RobotHero at 7:52 PM on December 7, 2019 [2 favorites]


re: taking up space meant for valuable ecosystems

Could one not build a canal or channel that creates an artificial meeting point for salt and fresh water?
posted by constantinescharity at 9:03 PM on December 7, 2019


I hear the echo of the Grand Coulee Dam. Interesting technology but I don't think the principle of the free lunch and it's non-existence has been repealed.

I vote for massive solar stations orbiting the earth and beaming power down with microwaves. Or deep sub mantle endless geothermal. Or smart everything that throttles up and down and harvests energy from shoe soles.
posted by Pembquist at 9:27 PM on December 7, 2019


TFA is not written for EE people, but for the layman whose only frame of reference might be the kWh on their electric bill.

Science is supposed to be targeted at AAAS members and should be aiming higher than that.
posted by mark k at 10:03 PM on December 7, 2019 [2 favorites]


MWh/yr is a totally normal way to measure the annual output of a power plant. Usually this will take into account things like scheduled maintenance shutdowns, etc.

It's cool if you're not familiar with the everyday jargon of power generation, but maybe don't go acting like you uncovered some great fraud because you noticed the units cancel out.
posted by ryanrs at 10:09 PM on December 7, 2019 [20 favorites]


I worked briefly in a power station, and there they said "megawatt hour".

Think of your electric bill, it is in kilowatt hours. If you have a 1 kW electric heater and want to figure how much it will cost to run it for 12 hours, how many joules is that? How many kilowatt hours is it? It is similarly a convenient unit on the generation side.
posted by save alive nothing that breatheth at 10:11 PM on December 7, 2019 [1 favorite]


This will work out. Or not.

See also: everything.
posted by Halloween Jack at 10:35 PM on December 7, 2019


Oh no. Everything won't. Eventually.
posted by aleph at 10:50 PM on December 7, 2019 [3 favorites]


I don't have any intuition how much 2.6 terawatts is

It's enough to simultaneously activate the flux capacitors on a fleet of 2,148 Time-Traveling DeLoreans, with about 920 megawatts of spare change left over.
posted by radwolf76 at 11:19 PM on December 7, 2019 [7 favorites]


I worked briefly in a power station, and there they said "megawatt hour".

megawatt hours isn't the confusing part - it's the "per year" since then you're back to a quantity that could be more simply expressed as an output in watts

but I'm not really that shocked that MWh/y is a unit people might actually use sometimes if they are used to dealing with energy at the scale of megawatts and time at the scale of years
posted by atoxyl at 11:59 PM on December 7, 2019


On the megawatt hour issue, both the estimates given above "3.4 kilowatts / m²" and "...up to 7,500 W/m2" are probably correct in their own way. However they may not be overly useful in estimating power generation.
The membrane experiment was conducted at " 7,500 W/m2 at pH 11 and 1 M:1 mM KCl molarity difference "

1. Seawater is generally around pH 8.1. That's a long way from pH 11

2. The molarity difference across the membrane is 1,000 to 1 - so there is a significant energy contribution from osmotic pressure. I would guess the difference between river water and seawater would be around 70 to 1 and the difference between estuarine water and seawater would be smaller yet. So osmotic pressure not a big player.

3. My only experience with membranes involved millivolts. If that is the case here converting millivolts DC to 240 volts AC would involve quite a bit of energy loss.

To be clear I'm not bad mouthing the experiment or the conclusions - just trying to clarify the power estimates.
posted by speug at 12:14 AM on December 8, 2019


There are a number of reasons why units of energy generated over time is useful. All generators have a capacity factor, which is the amount of time that they are generating over time. An onshore wind turbine might have a CF of 0.25, an offshore wind farm of 0.35 (though new ones are expected to be in the 0.45-0.50 range. A new nuclear or coal plant would be 0.95. So for each MW of those installed you would expect to get 1MW x 8760(hr in a yr) x CF. This tells you how much of your product you will make each year (or over its lifetime or the length of time you want to see a return). This is fundamental to knowing what your project economics look like whether it is worth investing, access to finance, etc
posted by biffa at 1:28 AM on December 8, 2019 [1 favorite]


Huh. I'm usually one of the first people to call out overly-breathlessly-framed reports of novel power generation technologies; in fact, that's one of the very first comments I made on Metafilter. But... I'm not seeing the reason for all the negativity here. There's no claims that this is going to solve the climate crisis, no claims that this replaces or is better than other renewable energy technologies, nothing that really strikes me as that outlandish. Transitioning to a totally-renewable energy economy means having a diverse portfolio of renewable production methods, because each individual method has significant limitations, but together they complement each others' weaknesses. (E.g., solar panels don't work when it's night or cloudy, hydro requires suitable waterways, etc.) This seems like an interesting potential component of that portfolio, that may be extremely valuable in certain cases even if it's not a large part of the overall system.
  1. Megawatt-hours per year seems like a perfectly cromulent unit to me, I'm not sure why this raises any eyebrows. Yes, you can express it in SI units, 1 MWh/yr = 114 W, but since power generation and consumption are often expressed in MWh, using MWh/yr seems natural here.
  2. I haven't seen the calculations myself or anything, but intuitively 3.4 kW/m^2 doesn't seem all that unreasonable to me. The thermodynamics of osmotic gradients can result in really huge energy densities, so the kilojoules are probably there. A semipermeable membrane can move a lot of particles relatively quickly, so while this number is a lot higher than I probably would have guessed, I also see no reason to doubt it.
  3. The basic mechanism of using osmotic gradients to power electrochemical reactions across a semipermeable membrane is pretty well established. Maybe not within the context of electrical power generation, but this is basically the mechanism that powers you and all other life on earth. In life it's a remarkably efficient way to convert energy into work, and can liberate a lot of energy quite quickly. The 30% efficiency cited in the article is roughly comparable to the energy conversion efficiency of human muscle. And slightly better than typical solar cell performance, considerably worse than typical hydro dam performance.
  4. Regarding the suggestion above that this would require massive infrastructure to support the large sheets of membrane required, that's almost certainly not how you'd do it. Again, taking a page from life's strategy book, you want to fold your membrane up as much as possible. Unlike sunlight, which has to travel in a straight line, salt water is happy to flow through complex surfaces. Just as a mitochondrion maximizes its power output by packing a huge amount of folded-up membrane into a very small volume, you'd fold the large sheet of membrane into relatively small volumes within your generation plant.
  5. The 2.6 terawatts number is described as a calculation of the total energy budget from river-to-ocean osmotic gradients on Earth. It's a rough measure of the total amount of energy that's out there, not what's at all reasonable to achieve, similar to calculations that are sometimes done of the total amount of solar energy arriving on Earth.
  6. Ultimately, like most other renewable energy sources this is just another way of capturing some of the energy that arrives on Earth from the Sun. Solar panels capture this energy directly. Wind turbines capture energy from air masses differentially heated by the Sun. Hydroelectric dams capture gravitational energy from water that's heated by the Sun, evaporated, and precipitated at a higher elevation than it started. This technique captures osmotic potential energy from water that's purified by evaporation by the Sun.
So I dunno. I think this is pretty interesting. Not a magic bullet for solving our energy needs, but nothing will be.

More interesting in my opinion is what can be done with it at smaller scales. How about powering implanted medical devices?
posted by biogeo at 2:16 AM on December 8, 2019 [32 favorites]


I recently got into a Youtube comments argument about the nuttiness of megawatt-hours per year. What finally got me to grudgingly accept that it might be useful was an analogy (that I had to make up myself, sigh) with traveling for a year. If I travel 10,000 miles per year I can calculate that my average (mean) speed is 1.1 miles per hour, but that's not a very useful number and probably has nothing to do with any speed I actually traveled at any given time. So, in the case of national electricity generation (which is what the Youtube comments argument was about), what's actually wanted is joules (as a lump sum) per year, not the joules-per-second average of watts... but nobody seems to know what a joule is, so we have to settle for megawatt-hours.
posted by clawsoon at 5:05 AM on December 8, 2019


My biggest question about this would be the practical one of how you keep it clean. Wouldn't nanopores get clogged up pretty quickly with all the stuff that's in river water? Would there be an easy way to flush the membrane? How quickly would power drop from contamination?
posted by clawsoon at 5:09 AM on December 8, 2019 [1 favorite]


Next Ant-Man movie: Divas of the Nanoworld.
posted by Kirth Gerson at 5:15 AM on December 8, 2019 [2 favorites]


Just too many practical details trying to be worked out here for something that only seems "proof of concept" at this time. Hopefully, there's more here to be worked out.
posted by aleph at 5:58 AM on December 8, 2019


If nothing else, it could be useful for offsetting some of the energy usage of desalination plants, even if it isn't appropriate for wider uses. There are probably a lot of industrial processes that also involve similar gradients that could be exploited. Making better use of energy means less must be generated, and given sufficiently good economics, could drastically reduce the cost of some energy intensive materials.
posted by wierdo at 6:00 AM on December 8, 2019 [5 favorites]


My biggest question about this would be the practical one of how you keep it clean. Wouldn't nanopores get clogged up pretty quickly with all the stuff that's in river water? Would there be an easy way to flush the membrane? How quickly would power drop from contamination?

Extrapolating from reverse osmosis and town water supplies, there would be a settling tank where coagulants are added to the water in order to clear it up and remove bacteria. The water might also be treated with UV light, biocidic nanoparticles like titanium dioxide, alumina, silver, copper. Plus there would be valves to backwash the filters periodically to keep performance high and there'd be a rotating schedule of removing filters to maintain or replace them once they reach end of life.
posted by Your Childhood Pet Rock at 7:38 AM on December 8, 2019 [1 favorite]


For some context on the number of 3.5 kW/m^2 (or perhaps twice that much when operating, but anticipating 50% downtime): That's kind of nutty. My kitchen stovetop is about a square meter and has four cooking elements. The two big ones are 2.5 kW, and the two small ones are 1.5 kW. So the claim that this membrane can, when operating, generate 7 kW/m^2 is claiming it has the same power density as if I covered my stovetop with a big piece of sheet metal and put all the burners on high. If you like the 3.5 kW/m^2 number better, put the burners on medium.

That seems like an enormous power density for a passive chemical process. (Though I see a comment above suggesting it's not totally crazy; I'm not a chemist.) If it's true, it's very interesting. I'm going to enjoy reading more about it.
posted by fantabulous timewaster at 1:19 PM on December 8, 2019 [2 favorites]


It's not a chemical process if I understand it. It uses the charge on the Boron Nanotubes to preferentially suppress the travel of that kind of charge from the salt water through the tube and allow the other kind with no problems. It disturbing reminds me of Maxwell's Demon but not really. This is a kind of weird Solar Engine thing.

Sun => evaporates water => rains down as fresh => this technique lets the resulting diffusion of one kind of charged ion through the channel "run down hill" while at the same time it drives the voltage across the membrane "uphill".

If I understand it right.
posted by aleph at 4:57 PM on December 8, 2019 [1 favorite]


Here's a little back-of-the-envelope calculation regarding the amount of energy available in the osmotic gradient between seawater and fresh water.

Seawater is about 1 osmole/liter, that is, about 1 mole of dissolved salts per liter of seawater. The osmotic pressure at (roughly) standard temperature therefore is (1 mol / L) * (8.3 J / mol / K) * (300 K) = 2.5 MPa, or about 24 atmospheres of pressure, relative to fresh water. Pressure is the same thing as volumetric energy density, so we can represent the rate at which energy is extracted from an osmotic gradient through a surface as a flow rate.

The claimed power production rate is 3.4 kW / m^2. To find how much water needs to be processed through the membrane to produce this level of output, just divide this by the energy density of fresh water relative to salt water, multiplied by the efficiency. 3.4 kW / m^2 / (2.5 MPa * 30%) = 4.5 mm/s.

In other words, to produce the claimed power production at the claimed efficiency, their system would need to process water at a flow rate of just about 5 millimeters per second through the membrane. This seems pretty sane to me.
posted by biogeo at 8:24 PM on December 8, 2019 [2 favorites]


It seems roughly doable at this point depending on what these things typically depend on. Early days but I'm hopeful for more than a flash in the pan. And then there's always the technique for straightening out nanotubes.
posted by aleph at 8:47 PM on December 8, 2019


So at 5mm/sec you'd need 1 cubic metre of water every 200 seconds for a 1 square metre membrane or if I didn't drop a decimal 5l/sec? About a bathtub of water every 30 seconds.
posted by Mitheral at 9:32 PM on December 8, 2019


Do you get more power if you were to use desalination brine discharge instead of plain seawater for the salty side? Or can you put brine on one side and sea water on the other?
posted by Mitheral at 9:35 PM on December 8, 2019


Ahhh this is up my alley - or it would have been, when I was in grad school and was closer to all of this stuff. As it stands, I think a bunch of people have made good points in this thread. I agree with aleph's assessment that the technology to line up the nanotubes is the more significant development here. It's not clear at all, to me, that any electricity has been extracted using the membrane which is the focus of this proof-of-concept talk.

I was confused by some of the references in the article, so I did a bit more digging, and found the 2013 paper alluded to in the article: Siria et al., Nature 2013, 494, 455. (This 2012 review (Logan and Elimelech, Nature 2012, 488, 313), referenced in that article, was also helpful.) I've collected some thoughts below.

How are the BNNTs "negatively charged"?
At high pH, it's suggested that water can (chemically) dissociate within the BNNT, leaving negatively charged hydroxide groups on the surface. This happens less at lower pH.
It's worth noting, as speug has, that pH 11 is nowhere close to the pH of most bodies of water.

What kind of current is being drawn/produced? The article states If they then dunk electrodes in the pools and connect them with a wire, electrons will flow from the negatively charged to the positively charged side, generating electricity.
You still need a redox couple to release/accept electrons for them to flow. The negative charge built up in this proposed device isn't an excess of electrons, it's a collection of anions. So.... where do the electrons come from?
Based on the 2012 review I assume the BNNT membrane would be used in a reverse electrodialysis set-up, where a ferrous/ferric pair is suggested for electrode materials.

But the 2013 paper specifically says it's not like reverse electrodialysis!
It keeps talking about a diffusio-osmotic flow at the surface of the nanotube, and critically, says that the nanotube is not ion-selective, so any current generated comes from a salinity gradient, rather than ion separation. The article states the opposite.

This is now firmly outside my wheelhouse, so I'll wait for someone to point out the source of my confusion.
posted by invokeuse at 10:06 PM on December 8, 2019 [3 favorites]


As the first person to complain about the units here, I retract that complain: the defenses posted here make sense to me in terms of applications to large-scale power generation.

I've also used some funny units to make power calculations more intuitive. For instance about ten years ago, I was wondering how much electricity cost in practical terms. I found that electricity cost me about a dollar per watt per year. Which mean that, for instance, leaving a 100 watt bulb running all year (a porch lamp, say) would cost $100. I recently redid the calculation and IIRC it's now closer to $1.30.

(I also calculated the cost to wash and dry a load of laundry and I think it was a couple of bucks.)
posted by sjswitzer at 11:02 AM on December 9, 2019 [1 favorite]


I've been thinking about this a lot.

This paper on Electroosmotic Flow appears to have some insight.
The EOF [electroosmotic flow] is implemented through the surface charges dominant in the small scales. The surfaces of most channel materials (e.g., glass and polymer) are negatively charged in an electrolyte solution. This causes a surplus of positively charged anions in the double layer close to the channel walls. Under an electric potential along the channel, the excess charges in the double layer are attracted by electrostatic forces, and thus, move toward the negative electrode.
So if you were to do the reverse, powered by a salinty gradient pushing the ions through, rather than using current to move them, you would get the same effect happening but getting current back out. It seems to be electrostatic phenomena, which is why I think they say it differs from reverse electrodialysis where prussian blue is continually oxidized and redoxed by selective ion transfer.

So, if I had to take a stab in the dark, the lower concentration side has excess positive cations through both the osmotic pressure and the negatively charged nanotube interface attracting positively charged ions. You're guaranteed to have a ring of positive ions along with a smaller typical fluid flow causing a slight positive charge at the outlet of the nanotube. If you had electrodes on both sides, the slight charge imbalance will cause an electron to flow from the negative higher concentration side to the positive higher concentration side trying to balance the charge. That would mean an anion becomes positively charged. Which then gets attracted to the negatively charged nanotube interface. Which then gets dragged through via osmotic pressure, which then meets its electron on the other side to get back to its typical negative ionic state because the Na ion that's been given the extra electron to hold in the mean time to balance the charges really wants to give it up.
posted by Your Childhood Pet Rock at 7:42 PM on December 9, 2019 [3 favorites]


« Older Star Wars: The Rise of Skywalker: The Metafilter...   |   The Morning The Music Died Newer »


This thread has been archived and is closed to new comments