Skip to Content

Chemical News

Down At the Small Surfaces

Mechanochemistry – getting chemical reactions to occur by pressing, pulling, and grinding solid substances – continues to produce weird and interesting results. Here are a couple of recent ones from the same issue of Angewandte Chemie, both from a group at McGill.

This paper is about making soluble compounds of the noble metals (such as palladium, platinum, and gold). That’s generally a rather painful procedure – they got the nickname “noble” because they’re quite chemically resistant. People have been willing to use ridiculous amounts of cyanide just to get gold into solution for mining and extraction purposes, for example, because the metal laughs at normal solubilizing conditions (short of say, freshly prepared aqua regia, which I don’t count as normal).

The group shows that if you take one of these metals (powder, wire, what have you) and mechanically mill it with either potassium chloride or ammonium chloride along with the weak oxidant Oxone (potassium peroxymonosulfate), all as powders, you get a powder at the end that completely dissolves in water. That’s a startling thought to a chemist, grinding up palladium powder or gold wire and having it form just a clear colored solution at the end. That solution contains ammonium tetrachloroaurate (or the corresponding other salts), which is already a known soluble compound, of course – it’s just that in the past, if you wanted your gold in that form, you had to sort of go all the way around the barn to get it. If you add other ligands to the grinding mixture, you can form triphenylphosphine complexes and many others directly.

So what’s going on here? You’d have to think surface area and fresh metal exposure, right off. The mechanical stress is constantly roughening the surface of the metal samples and removing any chemically passivated layers (which makes me think that ball-milling a thermite mixture might be a spectacularly bad idea). When you really zoom in, metal atoms that are on the surface are certainly different from the ones down in the bulk phase, and there are different kinds of surface atoms, too (imagine one sitting up on the top of a microscopic peak in the material – it’s surely more reactive, since it’s interacting with far fewer partners).

As an aside, I remember when I was about ten years old, and an elementary school teacher asked me if I thought that when a person touched a piece of metal, if some of the atoms from either our fingers or the metal ended up getting moved from one to the other. I have held that thought for over 45 years, coming back to it as I learned more and more chemistry, and here it is again).

The second paper looks at the digestion of cellulose into glucose, which (as you’d imagine) is a process that’s been investigated a great deal over the years. Cellulose is normally pretty sturdy, which is probably why plants make so much use of it, but there are several species of fungus that produce various cellulase enzymes that will break it down. Stirring cellulose powder in water with the (commercially available) powders derived from these species will indeed convert the polymer to glucose, albeit slowly. Actually, as this work shows, just mixing the dry powders together with a small amount of water (1 microliter/mg) will do the same thing. But taking the same ingredients and ball-milling them gives much better conversions, and further experimentation showed that a combination of five minutes of milling, followed by 55 minutes of “aging” was exceptionally effective, giving up to 50% conversion to soluble sugars after 12 hours.

The technique was still effective on going from microcrystalline cellulose powder (which is rather well-processed beforehand) to plain old cedar-wood sawdust (which is not!) I am no expert on cellulose degradation, but the group does seem to have substantially improved the reported yield/time figures, and with a rather simple technique. In this case as well, surface area effects must be a big part of the picture. It’s hard to picture what the local microenvironment around the enzyme molecules is like under these conditions – are they still clumped together into microparticles and able to assist each other somehow in a way that’s less likely in solution? (Remember, there’s more than one kind of cellulose-degrading enzyme present in these fungal powders). At any rate, it’s clear (once again) that there’s a whole world of surface chemistry and nanoscale solid phase chemistry that we don’t really understand and can be taking more advantage of.

43 comments on “Down At the Small Surfaces”

  1. BK says:

    I seem to recall a synthesis in grad school, that was the mono, di, and maybe tri-nitration of phenol and toluene with neat nitric acid and grinding with a mortar and pestle. Not so sure how it worked with toluene, but a labmate was telling me about it. I told him to not talk any more about it, let alone pursue it.

  2. Mad Chemist says:

    Ball-milling thermite! That sounds like fun, thanks Derek!

    1. Derek Lowe says:

      No warranty is expressed or implied.

  3. Jim Hartley says:

    Anyone want to comment on the scaleability of a ball-mill process?

    1. db says:

      I don’t know about these specific applications but in the industries I work in, we use ball mills that process hundreds of tons per hour.

  4. Isidore says:

    I wonder how well ball-milling might work if one of the reactants were a gas or volatile liquid, if the ball-mill apparatus could be made air tight. Perhaps this has been done.

    1. db says:

      Why not use a jet mill for an application like this?

    2. John Wayne says:

      You know some Russians tried this at one point trying to make fluorinated Telfon or something like it.

      1. wildfyr says:

        I believe Teflon is already rather well fluorinated.

  5. Anon says:

    “The second paper looks at the digestion of cellulose into glucose … five minutes of milling, followed by 55 minutes of “aging” was exceptionally effective, giving up to 50% conversion to soluble sugars after 12 hours.”

    Is this why domestic ruminants are given vegetable feed containing grit?

    1. Ted says:

      I don’t know about ‘grit’ (which is added to poultry feed to aid in gizzard processing), but clays (bentonite or montmorillonite) are frequently added to ruminant feed. In the US, clays are only sold as ‘anti-caking’ agents, whereas the same additives are sold in the rest of the world as ‘anti-caking/mycotoxin binding’ agents. Of course, cattlemen in the US aren’t dumb, so many of them add clay specifically because of mycotoxin exposure concerns.


  6. anon says:

    Interesting result with the cellulose. I would be more convinced if they used isolated enzymes as most lyo cell powders contain a boatload of inorganic salts. Some good control experiments would be to characterize the cell powders, check the apparent pH prior and after grinding, and maybe try the same experiment with just the inorganic buffer salts as well. The grinding causes localized extreme pressures and temperatures, so that in combo with a buffer salt might also rapidly degrade cellulose.

    1. MBP says:

      You could test if the enzymes are necessary for the process (vs. pressure/temperature/buffer salt) by using heat-denatured lyo cell powder and look at the effect on yield.

      1. anon says:

        good idea…Frankly, I’d be surprised if the grinding isn’t rapidly denaturing the enzymes and what they are seeing isn’t really enzymatic catalysis, rather an artifact of the other cellular components. A simple control like this would be good to have (have to test that you’ve first really denatured all activity prior).

        1. Algirdas says:

          I was wondering about denaturation as well. Most proteins despise being mechanically harassed (enzymes even more so). Once, I precipitated out an extremely stable and well-behaved protein, by shaking a dilute solution in a tube ~ 100 rpm overnight.

          On the other hand, sugars are well know for stabilizing proteins, so if you release some glucose in the first hour or so, that, together with other cellular crud, will chaperone your cellulases for subsequent milling cycles.

          I agree that a control expt with a heat denatured cell extract is a good idea.

          1. anon says:

            Sugars will stabilize proteins only so much. Having “evolved” many enzymes for industrial use over the years, it’s fairly expected to get thermo and solvent stability engineered into them, but shear sensitivity is much tougher/impossible to engineer for. Even taking something rock solid like CalB, if your tip speed is too high during stirring, it’ll cause cavitation and some fairly extreme sheer forces that mechanical denature it fairly quickly (air/liquid interfaces cause the most problems, like during vortexing). It’s analogous to making meringue, and that’s almost exactly what you see in the vessel…a big jelly blob. I really can’t imagine how some wild type cellulases are going to survive mechanical sheer forces in a ball mill. It’s a pretty big claim to say they are actually seeing active enzymes under these conditions.

      2. Derek Lowe says:

        That is an excellent idea for a control experiment, I have to say.

  7. DrWest says:

    Hmm, this gets me thinking:

    Is it possible to seperate the various noble metal complexes from each other in a way that is easily facilitated? Or, at least seperate one specific desired one from the rest of the solution?

  8. Chris Phoenix says:

    Ball milling thermite has been studied.!Self-sustaining_Takacs2002.pdf
    Self-sustaining ignition happens under surprisingly deterministic conditions (time spent milling, ball size and number, charge weight and stoichiometry, etc.), and often after a surprisingly long time. They were even able to replicate ignition times from a different study.

    1. tangent says:

      Aw yeah. There is still so much information in the world that can’t be found just by searching [some dude ball-milling mixed thermite] on YouTube.

  9. Barry says:

    The mixture of KCl/Oxone is not dissimilar to aqua regia. Oxone is not a weak oxidant. Persulfuric acid is quite as effective as the old hexavalent Chromium “Chromerge” for chewing organic tars off glassware. And of course the (tightly bound) Cl- ion is there to stabilize the the Auric as it is formed.
    As to thermite, I have it on good authority that you want to mill your aluminum and your ferric oxide separately, and only then mix the two.

    1. Jochen Brandt says:

      Absolutely! If wikipedia is correct on peroxymonosulfate’s standard electrode potential (+2.5 V), it’s a stronger oxidant than Xe(VIII) (+2.4 V)and not far off fluorine gas (+2.8 V).

      Oxidations with oxone are often selective because they are kinetically hindered until you add Ag+ or a similar catalyst.

    2. tangent says:

      Tangentially, I had never realized until now that peroxymonosulfate is retailed in big plastic tubs as “non-chlorine pool shock”. If I had realized that as a dumb pyro teen I wonder what I would have ended up at.

      1. NLL says:

        +/-10,000 feet, ballistic. (Or at least parts of you.)

  10. Bob says:

    For metals processing, it sounds interesting. However, for cellulose one wonders whether the mechanical energy input is out of proportion to the value of the transformation. Twelve hours in a ball-mill sounds like a lot of energy to me.

    1. Jim Hartley says:

      Probably right. There go my visions of trucks full of waste cellulose rolling up to giant ball mills on the plains of Nebraska. Oh well.

      1. barry says:

        The best thing we can do with that waste cellulose (“stoker”/”bagasse”/”straw”/”newsprint”…) is to protect it from cellulases* and rebury the gigatons of carbon we have pumped from fossil pools into our atmosphere and oceans. Whether you call it “anthropogenic peat” or “biochar”, we need to take it back out of the cycle

        *in ball mills, in ungulate guts, in termite guts, in soil fungi

  11. Chris Phoenix says:

    It seems you can mill thermite without it ever going runaway, if you simply cool it to -50 C. The authors state that if you mill it too long, you’ll get a pre-reacted material that won’t ignite – but it won’t ever have gone into runaway.

  12. anon says:

    I never thought about solids reacting until one day I was about to run a Corey-Fuchs dibromoolefination reaction. The aldehyde was in the flask and I went to weigh our the CBr4 and the PPh3. Being lazy, I put them, consecutively, next to one another on the same piece of weighing paper. When I was walking the paper to the hood, the piles of reagents slid into one another on the weighing paper. The mix starting popping and sending off brown fumes. Pretty surprising to me at the time.

  13. Graham Clark says:

    Flann O’Brien’s “The Third Policeman” has a discussion of atoms being transferred to and from people. Apparently people who spend time cycling exchange atoms with their cycles, to the extent of being more than half bicycle, and their bicycles are more than half people. Indeed, a local postman is said to be seventy-one percent bicycle.

    1. Derek Lowe says:

      I’ve read “At Swim-Two-Birds”, and I can well believe that this is Flann O’Brien.

  14. Uncle Al says:

    Fully acetylating small carbohydrates is not a happy day. Ball milling the anhydrous reactant and then acetylating is a good day – especially when marching to left of the decimal place.

    1. Mark Thorson says:

      Acetylation of wood at large scale is
      a commercial process.

      1. Uncle Al says:

        Cellulose acetate tow – cigarette filters – is rather immortal. Even so, acetylated wood hints hydrolysis from ambient humidity and condensing moisture will micro-leak acetic acid. Adjacent metals will corrode. I would perform accelerated aging tests.
        … Chinese cheap sheetrock micro-leaked hydrogen sulfide in humid areas (Florida), gutting home wiring and all else containing copper.

        1. Mark Thorson says:

          Agreed. I can smell a whiff of acetic acid when I open my bottle of aspirin (since discarded). Apparently, acetylation is not as stable as one would wish.

          1. Barry says:

            aspirin (an acetylated electron-deficient phenol) is a poor model for the stability of an acetylated polysaccaride.
            I expect you’d have to cleave quite a few acetates off before it was a substrate for cellulase, which might take several enzymes. That’s not an impossible barrier for Darwin (a few million years after the advent of woody plants, something did come up with the first cellulase, which seems a harder problem). But it could take quite a long time, by human standards.
            And since this material is hydrophobic, it offers a lousy environment for a bacterium or fungus in the process of evolving those esterases.

      2. Barry says:

        thinking about per-acetylated cellulose…Surely part of the strength of crystalline cellulose is in the H-bonds, rather than just the covalent polymer chain? Per-acetylation must disrupt that. But maybe in wood, cellulose is not crystalline at all?

  15. Lauri Hauru says:

    You didn’t have a link for the second paper, I assume you’re talking about this:

    I was more surprised by the cellulose hydrolysis paper than the metal paper, because it’s enzymes we’re talking about. But I read the paper and I’m far less impressed. They use “microcrystalline cellulose” (MCC), which is manufactured by hydrolysis in sulfuric acid. A lot of people unfortunately use MCC as a model for cellulose even in technological experiments. MCC is not natural cellulose and cannot be used to represent real pulp or wood, either scientifically or technologically, because it’s highly processed and its morphology, particle size distribution, composition or molar mass distribution are very different.

    Yields are also unacceptably low. Cellulose is a polydisperse material. It’s not unexpected to get the lowest-molar mass fraction to react; the goal should by 100% hydrolysis. You can always say they’ll optimize it later, but if the limiting factor is molar mass, no, they don’t.

    Also, energy consumption of ball milling is extreme; such processes that exclusively rely on it are very unlikely to ever leave the lab. You can’t have a 20 day milling process in real life. They consider it a good result if they get 50% hydrolysis in 12 hours, but that logic only holds water if the reaction was linear, which it isn’t. Also, 24 hours of ball milling is a lot of energy per ton of pulp.

    Besides, this is of questionable novelty. Guerra et al. 2006 – 10.1021/jf060722v – and King et al. 2009 – doi/org/10.1021/jf901095w – determined that 96 hours of milling is required to get all hydroxyls in wood accessible to solvent and cellulose hydrolyzed. I can concede that the main point in the paper, trying to get a higher glucose concentration, is valid, and should be pursued. But alone, you don’t get very far with this paper.

  16. milkshaken says:

    Oxone might be easy to handle as the potassium tripple salt, but KHSO5 is not a mild oxidation agent = it takes ketones to dioxiranes, Mn(II) directly to permanganate, and the like.
    It also has a sweet tooth for oxidizing phosphines to phosphine oxides (even plain H2O2 can do that instantly) so let me express here my doubt about the research quality of this paper.

    Anything that replaces NaCN in gold mining is good (the only viable alternative so far is thiourea, which has its own problems, and a rather creepy chronic toxicity too) but the alternative has to be very cheap, to compete with NaCN leaching. Also I would imagine sulfidic ore components like pyrite would gobble up huge quantities of Oxone, and iron would catalyse its decomposition- just a thought.

    1. Nick says:

      “Oxone might be easy to handle as the potassium tripple salt…”

      Isn’t Oxone a trademarked product name referring specifically to the triple salt? The paper seems to say that’s what they used.

      I also don’t see any claim in the paper that this would be a useful technique for isolating elemental gold or palladium from ore, just the claim that this is a more facile synthesis of soluble noble metal complexes compared to current techniques.

      1. milkshaken says:

        I take aqua regia over milling followed by extraction of the resulting stuff contaminated with lotsa salts and possibly metals from the milling machine. With aqua regia you just evaporate and have pure HAuCl4 or H2PtCl6

  17. Richard says:

    Potassium peroxymonosulfate is used as a shock in hot tubs. That is all.

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.