Study on the Parameters Affecting Acetone Peroxide Synthetic Yield

[megalomania]

Over the last several years I have pondered on how the yield of acetone peroxide could be optimized through a series of controlled experiments testing different parameters encountered during the reaction. I don’t suppose I will ever find the time to do the study in the manner I would prefer since it seems to be quite involved. Therefore the least I can do is share my thoughts and ideas on how I would conduct such a study were I to sit down and do it. Rbick’s thread on comparative yields of acetone peroxide using different OTC sources of acetone (Comparative study of OTC Acetone in AP Synthesis - http://www.roguesci.org/theforum/showthread.php?t=7068) has prompted me to share my knowledge in case some Forumites were interested in optimizing acetone peroxide syntheses themselves, but perhaps were not sure how to begin testing.

I have never actually sat down and wrote out the full experimental conditions on which I would test acetone peroxide synthesis; this information is just my mental musings. Defining a detailed experimental methodology would require some library research I have not yet invested in this project since I don’t consider this a priority experiment on my list of things to do. I only kept notes on this experiment in my head, which is in dire need of defraging…

The big question I want to answer with this study is to dispel some of the myths and misconceptions I hear about acetone peroxide. There is a significant difference between the procedures used in professional literature, and the procedures used by amateur scientists. The amateur scientists provide copious information on the success of the reaction, but their information is only anecdotal. Professional chemists have, to my knowledge, never published anything other than a procedure, with no mention as to how or why they derived their methods. There exists a gap of information on this compound caused by professional chemists basically writing off acetone peroxide as a viable energetic material.

My study can actually be broken down into many subparts, a feature that may benefit some would be researchers. Some may have only the resources to test for specific subset of the study, others may find the number of experiments required to be exhaustively thorough to be daunting (part of the reason I have not undertaken this task), and so may choose to limit their efforts to a specific subtask. All reactions should ideally be conducted only with laboratory grade reagents using analytical instruments as precise as possible. All experiments should be replicated a minimum of three trials to minimize experimental error, and to provide enough data for statistical analysis.

The conditions that affect acetone peroxide yield, and the reasons why I would test these factors, are as follows:

Determination of the effectiveness of acid catalyst on acetone peroxide yield: According to published literature sources (citation needed, may have been an anecdotal reference) only mineral acids are suitable for preparing acetone peroxide. There is also the question of why published procedures use sulfuric acid, but amateur procedures use hydrochloric acid. This part of the study would compare acetone peroxide yields of a variety of common acidic substances including sulfuric acid; nitric acid; hydrochloric acid; phosphoric acid; citric acid; and acetic acid. This experiment should take into account both the molarity and pH of the acid catalysts. I would initially focus on volumes of acids adjusted to have equal molarity. All other conditions remaining equal, the reactions would be done simultaneously in multiple test tubes to keep the temperature of the reaction, length of reaction, etc. constant. The results of the replicate trials should be averaged and compared via graph plots to see if any one acid is favored over another.

Acidic factors affecting the yield of acetone peroxide: The preliminary data of what acid is more effective from the last experiment only tests for yield based on a single arbitrarily set molarity. Each acid tested should be varied over a range of concentrations. Starting from a common batch of acetone and hydrogen peroxide, serial addition of an increasing molarity of acid in each test tube would test for the effects of acid concentration. There seems to be a wide variety of concentrations of acids in both published and amateur procedures for acetone peroxide synthesis. This experiment will shed some light on what effect a range of concentrations and a range of pH levels will have on acetone peroxide yield. Each acid should be tested separately, for example testing only sulfuric acid with each test tube having a slightly higher molarity of acid, all test tubes then immersed in the same temperature bath, addition of reactant controlled to keep conditions as similar as possible. The concentration experiment is separate from the pH experiment, although both data should be recorded for every reaction. The pH experiment would be independent of concentration by varying the pH over a range of values below 7. What then is more important, molarity or pH? This experiment should determine if organic acids are in fact inferior to mineral acids when pH factors are equal by comparing pka’s to hydronium concentration.

Temperature factors affecting the yield of acetone peroxide: Proper temperature control is apparently important on the yield of acetone peroxide, but where is the data supporting this? Published literature gives vague warnings about not letting the temperature get to high, but I have not seen any studies that explicitly test for temperature differences. We take for granted the unsubstantiated claim “higher temperatures cause more diacetone peroxide to form.” Is this in fact a truthful statement (citation needed, I think this is a claim made by a published source)? First, does varying the temperature of the reaction have an impact on overall yield? Second, does varying the temperature lead to different ratios of triacetone and diacetone peroxide yield? Testing for the presence of diacetone peroxide is rather difficult since it would require sophisticated equipment. Unless there is some sort of purification technique that can separate the products. Determining overall yield is not a suitable yardstick if a significant portion of “crystallized product” is in fact not triacetone triperoxide. What percentage of diacetone diperoxide is formed by varying reaction temperatures? Vary the temperatures; do yields depend on initial low temperatures followed by warming, are yields improved by sustaining low temperatures for hours, days, weeks?

Determination of the rate and order of addition reagents, mixing speed, and length of time beforehand that reagents are combined affecting the yield of acetone peroxide: Does the rate of addition, and the order of addition of reactants make a difference in the yield of acetone peroxide? Should the acid, hydrogen peroxide, or acetone be added all at once, slowly, with stirring, without stirring? How long beforehand can you combine reagents? Will there be adverse side reactions if certain combinations of reactants are kept over long periods of time? This experiment would test for acetone/hydrogen peroxide mix, acid added; acetone/acid mix, peroxide added; peroxide/acid mix, acetone added. Addition conducted very slowly, slow, rapid, all at once. Reaction mixtures rapidly stirred, slowly stirred, not stirred at all (shaken but not stirred?). Reagents mixtures combined immediately before reaction, stored for some minutes, hours, days. Does the reaction benefit from brief stirring, how long should stirring be conducted?

Determination of length of time needed to optimize acetone peroxide yield. What affect does time have on the yield of acetone peroxide? Is a reaction done after minutes, hours, days? How do the time differences come into play depending on reagent concentration, temperature, reaction size?

Determination of hydrogen peroxide concentration on acetone peroxide yield. Another unsubstantiated claim is that using weaker hydrogen peroxide lowers the yield of acetone peroxide even when equal molarities of hydrogen peroxide are added. This may have something to do with the concentration or pH of the acid catalyst rather than concentration of hydrogen peroxide. This experiment should test for acetone peroxide yields over a wide range of hydrogen peroxide concentrations, both with adjusted pH levels, and without adjusted pH levels. Maintaining similar molarities of acid should also be tested for. Adding additional acid might be all that is needed to sustain high yields, or perhaps water does have an adverse role in this reaction.

The hard part of doing these reactions will be evident if you consider you would have to conduct hundreds, perhaps a thousand or more separate reactions to test for all of these parameters. Each experiment should vary the parameters from the earlier experiment, although I didn’t put these in any particular order that need be followed. You have 6 acids in the first experiment, 6 different reactions for molarity, 6 reactions for pH, 12 reactions overall. In the second experiment you test each of the six acids over a range of molarities, say 10 different points, giving 60 reactions, then test for 10 different pH levels for another 60 reactions, giving 120 reactions for this experiment. The third experiment could test over 5 different temperature points, with six acids is 30 reactions. Those same 30 reactions tested over the earlier 120 different reactions of acid concentration and pH would be 3600 different reactions to be conducted!

To test exhaustively you would have to repeat thousands of experiments, but that sounds like a masters thesis to me. Instead, you do a series of experiments on the list and if an acid does not give useful yields you don’t use that acid anymore. If there is no difference between sulfuric and nitric acid, don’t use nitric acid. If yields drop off over 10 degrees, testing the 5 temp points at -5, 0, 5, 10, and 15 degrees, don’t test all of the varying acid concentrations at greater than 10 degrees. If yields are the same at -5 to 0 degrees, just do the reactions at 0 degrees.

The purpose of doing the study in a series of separate experiments is to focus on a specific condition, hold everything constant, and optimize that condition. Sure, you may never know that 9M citric acid at 13 degrees gives 99.9% acetone peroxide yield in 15 minutes if you don’t test for every parameter, but if every other experiment with citric acid gives terrible yields regardless of concentration, pH, or temperature, chances are every trial with I will be terrible. That’s why you have to plot graphs, calculate slopes, and determine trends in the data.

There are probably a few other conditions I forget to mention. Who knows if there are funky conditions affecting yield like exposure to visible or UV light, presence of atmospheric oxygen, volume of reaction, etc. I doubt the ancient chemists ever thought to conduct an acetone peroxide synthesis in the dark with the exclusion of air. Why can say if there will be a difference? Maybe this reaction is activated by UV light, or having air bubbled into it. I didn’t even tough on crystals forms of acetone peroxide. It is more than likely that recrystallization from different solvents can have an impact on crystal structure, so this is more of a physical characteristic for explosive engineers to worry about rather than a synthetic characteristic for chemists to worry about.

The most important considerations I have are figuring out what the deal is with the different acids, hydrogen peroxide strength, and reaction temperature. These are the three conditions with conflicting data, and in al likelihood the only important affecting acetone peroxide yield.

[Ratman2050]

Very thorough analysis, and if someone would do apply this theoretical knowledge it would be greatly knowledgeable yet tedious.

I've been through some of these parameters in my experiments such as the temperature variation. For example, when conducted outside in a cold atmosphere (0 C, not outside in solution), precipitate seem to show instantly, and is clearly visible and solid.

When done in an ice bath inside at room temperature (less than 10 C) precipitate is shown once acid is dropped, however, this precipitate quickly dissipates into smaller crystals which are less visible. Although, when taken to a freezer (exactly 0 C for mine) for 30-60 min precipitate seems to become more visible and solid similar to the outside experiment.

As for different acids I only have H2SO4, nothing else since I am a novice yet learning and haven't conducted nitration's. And for H2O2 strength, I have done over 5 experiments before I had 27% Hydrogen Peroxide and all yields were less than 1 gram, I tried many quantities for variations (some were left in fridge for a week with no results). When I used 27% H2O2 it didn't matter what quantities were used as long as there was excess acetone I produced over 15g yields of AP.

I hope that helps.

[megalomania]

With an analytical balance these experiments can be conducted in test tube batches at the 100 mg scale allowing materials to me minimized, space to be conserved, and such things as temperature conditions to be shared across as many vials as will fit in the ice bath. For example I would test all of the different acids in the same batch at the same temperature and length of reaction. The effective concentration of peroxide could be determined by serially diluting 30% peroxide with more and more water, then running all the reactions using all of the acids in the same temperature bath.

Using a test tube rack with the tubes crammed in would speed up doing the reactions, but you would still have to weigh all of the products individually, which is where the tedious part comes in. With the right equipment once could devise a means of filtering and aspirating dry dozens of test tubes at once, then they just need to be weighed )before and after) to calc the yield. If one had the best resources, a combinatorial reactor could do nearly all of the work automatically.

[Ratman2050]

Although I lack many of those resources, wouldn't it be more proficient to test these analysis's on a different explosive, something more readily used such as a secondary, since after most of us make AP we don't make more than 500g+ batches because it would be easier to get a secondary that could require 1/8 of that quantity to make the same impact.

[Rbick]

Well I think the ideas suggested by Mega can be used in almost any explosive synthesis. You just have to change the wording around a bit. I think AP is a good explosive to discuss, as so many people use it, newbies and experts alike. When making a product as sensitive as AP, I think it is also a good idea to know what procedure will yield you the most trimeric forms and the most stable and contaminant free.

You certainly are not wrong though, this could be used on secondary explosives, although at a much higher price I'm afraid.

Hell, we could break this whole experiment down and assign different parts to different people. Considering most of us don't have the materials, funds, or time, we could just split it up a bit to those willing to participate. it could be its own section, well maybe. :D

[megalomania]

Oh I could spend my time curing cancer, ending world hunger, and putting a man on Mars, but I don't care about that stuff. I am curious about AP, specifically the lack of verifiable data on its yields. What I choose to work on is my decision.

[Barnacles]

Made a booboo. sorry could a mod please delete this. Thanks in advance

[Silentnite]

I have a project for the summer. I enjoy AP and the creation of. Once I finish building my lab I'll respond back.

I do believe however that HCl is used more by amateurs simply because it's more widely available. I can never seem to find H2SO4 but have obtained a source for HCl.

While batching I've noticed that the colder the ingredients prior/when mixing the higher the yield. Letting the mixture sit prior to adding acid for a day or three has also increased yield substantially.

[Rbick]

I have a great amount of access to both acids, and I will always go with HCl for organic peroxide synthesis, including HMTD.

The reason for this is Sulfuric is too bothersome and a hassle for a simple oxidation reaction. I also believe it is too strong of an acid to use in this particular reaction, which means it should be diluted, which means more work than necessary. Plus, regardless of where you buy your acids, HCl will always be cheaper. I bought 3.7L of 31.45% HCl for about 2 dollars as compared to even the crappiest Sulfuric drain cleaner being 7 dollars per liter.

I would much rather save my Sulfuric Acid for my NG/ETN/NC/ect. labs. I have never done a comparative yield with SA and HCl reactions, but I am perfectly happy with my HCl, as the entire reaction vessel is stuffed with fine crystals at the end of the lab.

[Azido]

While this is far from an analytical response, nor is it backed by a particular amount of theory, I have made some interesting observations that could be used to further research into this topic.

I decided for to make AP for the first time the other day. I went out and bought seven 473mL (around 2.6 cup) bottles of 3% topical stabilized solution, and a quart of pure acetone. I also picked up a bottle of 32% HCl.

I did a quick run through of some threads and decided the Acetone to Peroxide(3%) ratio should be around 1:10.4. i mixed my first batch of the chilled precursors, one bottle of Peroxide and 50mL of Acetone. I slowly added drops (40) of the HCl, but got only a fraction of a rise in temperature, and four days later, no yield, not even a trace of precipitate.

My next batch I thought, perhaps if the peroxide isn't being utilized fully, the yield has been dissolved in the remaining acetone. This batch I did the same thing, albeit with half the acetone. No results. Nothing, even still.

I even brought samples of the two out into the room temperature environment to test if the reaction would progress in a warmer environment. Still nothing.

Frustrated with my lack of success and apparent waste of time and chemicals, I read more into the threads and picked up vague hints that the three percent solution might be unsatisfactory, and that a higher concentration could be attained through boiling. Using common sense, I brought two bottles, (five cups, 946 mL) to a heat just before boil, so bubbles were just forming and subsiding. I continued this for a few hours until I had just over a cup and a half (approx 315 mL) of peroxide remaining. Assuming I lost very litttle peroxide to decomposition, I was left with two cups of around 10% solution.

I adjusted my ratios accordingly and chilled the precursors. This time almost immediately upon addition of the catalyst, small crystals began to form on the surface. The mixture was placed back in the freezer and continued to form a fine, amost milky precipitate that eventually (6 hours) formed a conglomerate mass of cloudy precipitate.

Finding hope in this method of concentration, I boiled three bottles (8 cups, 1419 mL) down into two cups of peroxide (368 mL). Assuming 12% concentration, I adjusted my ratios and proceeded as usual. There was almost instant precipitation, and the reaction seemed very healthy.

Curious as to why neither seemed to be producing much heat, I took samples of both (less that 10mL to avoid affecting the yield too much) and exposed them to room temperature. It was really interesting to note the change. Rather than forming the powdery, cloudy precipitate, the samples began to form large flaky clumps of precipitate, which rose to the top in a bubbly foam of precipitate, and larger clumps sunk. Perhaps I'd begun to form the dimer? Or perhaps just a result of the temperature change.

After 24 hours, filtration, washing, and neutralization, the yield of the 10% batch, with around 50mL of Acetone, was approximately 24 grams of Acetone Peroxide.

The yield of the 12% batch was an impressive 78 grams for 146ml of Acetone.

Now I've got a shitload of acetone peroxide and nowhere to put it.

[Enkidu]

May I remind you NOT to put in a container with a screw on lid. The crystals are known to sublime and pervade the threads between the cap and the container.

[Azido]

I used porcelain bowls. Bet grandma never imagined what her bowls would be used for. :)

[-=HeX=-]

Azido: Nice tests, any chance of putting all the ratios in millilitres because us europeans do not use cups, whatever they are.

As for the excess, do a few storage and sensitivity tests on it and post results. Or maybe make acetone peroxide and rubber cement putty. I am sure you can find a use.

Come the summer I will do a shit load of tests on TATP, HDN and ETN. Maybe I will do the tests mega suggests when my new labware arrives.

[Azido]

At the risk of straying from the thread topic, I've stored my ap in damp filter paper, in an altoids container, and buried it. I'm thinking about buying some starch and pressing the damp ap into lock pick sticks (as mentioned in another thread) or some other shapes.. but for now it stays buried. I will perform some friction, crater, drop and flame tests as soon as I am afforded some alone time. I'm also thinking about running my two rxns again at room temperature and doing friction and drop tests on the product to determine whether or not there is an increase in sensitivity. If there is, it might be an indicator that I've formed the dimer, like I predicted.