An alternative to melting:
Recrystallized Potassium Nitrate and Sugar
Recrystallized Rocketry
Home Page
"Rcandy" is a technique for making a rocket propellant from potassium nitrate and sugar without the necessity of melting either of these components.  Instead, they are dissolved and re-crystalized, yielding a stable-burning propellant with interesting molding properties - it can be formed by hand into almost any shape.  I believe this to be safer than the traditional method of melting the components together in an oil bath, but that is yet to be proven.

Please note that I NO LONGER USE the oven method presented here.  The skillet method is better. 
If you do this start with small batches
Large batches are more difficult, more dangerous, and often wasteful as the first batch or two may not turn out right. 

Check out my latest launch using this propellant

The Propellant
Blessings of Recrystallized KNO3/Sucrose

I like this kind of propellant especially because it is non-toxic.  It consists of a food additive and a fertilizer.  Thus I don't have to worry a lot about handling it or dropping bits in the yard.  Potassium nitrate is used as a meat preserver in making sausage, and as a medicine as well.  In my adolescence I obtained KNO3 from a drugstore, where the bottles indicated that one should take 1/4 teaspoon dissolved in water, as a diuretic.  I also found some in a butcher shop where it was used in making sausage.  And I just noticed that my tube of toothpaste includes KNO3 as a desensitizing agent.  Thus one should not have to worry about moderate  contact, and even ingesting small amounts of it should not cause immediate harm to most individuals.  Looking at a bag of Peter's Special liquid-feed fertilizer, I note that Potassium Nitrate is the first item on the list.  Plants love it.

Cautions on KNO3:

  1. There is evidence that use of nitrates/nitrites in food can cause cancer, so while eating KNO3 is possible, it is not recommended
  2. Once ingested, potassium nitrate can be metabolized into potassium nitrite, which is toxic and especially dangerous to children.  Keep it out of their reach!
  3. Some people develop dermatitis upon contact with KNO3.  If you do, then gloves are recommended when handling it.
  4. Breathing large amounts of KNO3 dust can cause lung damage.  If you are exposed to much airborne KNO3, wear an appropriate respirator.
See the Safety page for more information

Other good things about Rcandy:

* No promises expressed or implied.  Every propellant has some degree of sensitivity, and will ignite when subjected to a certain level of violence.  This propellant seems to have a very low sensitivity as it has never ignited on me through shock or friction in thousands of opportunities.  But I cannot guarantee that it will never ignite from shock or friction.  Adequate precautions must be taken to ensure that if accidental ignition occurs, no serious injury or significant property damage will result.


1.  I do not like the fact that this propellant is hygroscopic, absorbing moisture from the air.  It must be sealed against air or it will be rendered useless in a short time, especially in Florida, where 90 percent humidity is considered low.  In a way, this is a safety factor, as little bits dropped here and there will not be flammable for long.  They will soon become fluid, revert to the earth, and the plants will eat it up.

2.  This propellant is somewhat brittle, thus it must be arranged in the rocket motor so that it does not break or crack during handling or upon firing.  Fears include that case-bonded grains may crack if the case expands, and that Bates-type grains may slam against each other.  There has been some concern expressed about shear forces caused by extreme acceleration breaking off pieces which clog the nozzle.  Conservative engine design can compensate for this limitation, but it may restrict the options for grain geometry.

3.  This propellant has lower ISP than some other propellants, notably composites.

4.  It produces very little visible flame when fired in an engine.  This is mostly an aesthetic problem - many would like to see a brilliant white or colored flame as the rocket streaks skyward.  I have yet to find an additive that creates a "tail" of visible flame or sparks.  Dennis Welch reports that titanium creates a nice tail, but I have yet to try this myself.

5.  Making propellant by this method requires the use of an oven, normally found in the kitchen.  Warning!  This may jeopardize domestic tranquility!

Not in the kitchen!
Please note that I do NOT recommend making this propellant in a home kitchen.  I hope one day to have a well-documented safety program such that moderate amounts can be safely made in a kitchen.  The risk of an accident may not be any greater in a kitchen than a workshop, but the consequences are more likely to be catastrophic.

This will pose a dilemma for many, as most ovens are in a kitchen and not easily relocated.  I am searching for a solution to this problem, and welcome any suggestions.  Currently I am exploring the possible use of a plug-in convection oven for the workshop, thermostatically-controlled gas grills for use outdoors.  None of these options have been tested so far.  My preference is to attach a conventional gas stove to a portable propane cylinder, such as this one on my back porch.

A little good news:  I have recently made half-batches of rcandy in my toaster-oven, using a suitably-sized Pyrex baking dish.  So if you are not ready to spring for a an oven in your workshop, this might provide a good way to get started.
Safety:  You MUST read the safety page before making this propellant! Click here. 
And don't come back until you have read it all!

The Recipe

Photos in this document are thumbnails - clicking on one will call up a larger version of that picture.
Hit your Back button to return to this page.

Click here to read a rambling narrative text-only version of this recipe.

The ingredients:
100g KNO3 (aka:  potassium nitrate, saltpeter)
  50g sucrose (aka:  white sugar, cane sugar, beet sugar also works)
  18g light corn syrup (Karo Syrup is the standard brand here.)
    5 tablespoons water = 80ml =1/3 cup
No Karo?   No Problem!
Use a little Dextrose and Fructose

Equipment includes:
- 2 Pyrex glass pie-pans, "9-inch" nominal diameter
- 2 stoneware or glass plates that fit nicely on top of the Pyrex pie-pans, thus serving as covers.
- Small saucepan - I prefer stainless or enameled steel, but see no reason why other pans should not work.
- Triple-beam or another scale accurate to a gram or better and capable of weighing 100 grams or more.
- Table knife and spoon, fork is optional.
- Measuring spoons or measuring cup
- Oven, preheated to 300 degrees F. (150 celsius)
- Pot holders or some other hot-pan-grabber.
- 35mm film cans, or other small containers that will seal air-tight and won't fragment if burst.  I have never had a spontaneous ignition with this propellant, but you never know...

Protective equipment is required.  I recommend using a full-face shield, a heavy apron, long-sleeved shirt, and flash-resistant gloves.  These must be worn at certain phases of production.

Fire-control equipment is also necessary.  Hopefully you won't ever need it, but just in case, you had better have it.

This is some of the saltpeter used in this batch.  I found it at an oriental-food store.  In several tests it acted just like KNO3, so that is what I assumed it to be.  It worked fine in this batch of propellant.  It is expensive.  At $.89 for 2 ounces, it comes to $7.12 per pound.  You can also purchase KNO3 at a good pharmacy, although it will be expensive and they may look at you funny.
Mostly, I use potassium nitrate purchased from pyrotechnic chemical suppliers, such as Firefox and Skylighter.  Both of these outfits have treated me very well.  They sell KNO3 for $2.50 to $4.00 per pound, depending upon the grade and the grind.  Finer powders tend to cost more.  For this process the coarsest grade will do, as dissolving the saltpeter reduces it to a very fine particle size..
I normally weigh a small pan, then calculate the added weight for each component and add it.  The pan weighs 251 grams.
So I set the scale at 351 grams and add KNO3 until it balances 
Now I set the scale at 406.0g and add sugar. 
More sugar...finally it balances. 
Now a little corn syrup.  I have never used this brand (Roddenberry's) before but it worked just fine.  Tastes pretty good too.
Honey works OK as a substitute for corn syrup but makes the resulting propellant more hygroscopic, so I don't recommend it.  Recently I treid "Honey Girl" pancake syrup, which is mostly corn syrup with a little flavoring, and it worked OK. Maple syrup did not work - it made a propellant that burned well but could not be easily formed into shape.
A finger is a very useful thing.  I use mine to determine when the scale is about to come into balance so I don't pour in too much sticky syrup.
Adding 5 tablespoons of water, which is just about 1/3 cup.  I use tap water.  If you live where there is a lot of stuff in your water you might want to used distilled.  We are most fortunate in North Florida to live atop one of the world's great aquifers.  The water from my well tastes better than most of what is sold in fancy bottles. In fact, you might buy some of it:  Dannon has set up a water-bottling plant near High Springs, FL, just a few miles north of my house.  I believe that they are drawing from the same aquifer.

Pan goes on the fire.  This is a moderate danger point.  It's hard to see it the photos, but there is a medium-heat flame under the pan.  Note the two different cookers.  Both work.  Sometimes I have used a plug-in hotplate, which works great.  I am sure it could be done over a campfire in a pinch.   It is not critical how you heat the pot as long as the KNO3 and sugar get dissolved.  It IS critical that you observe and do not allow this to boil over or to cook dry!

Now I watch and stir.  This seems like a hazardous thing, but in 25 years of making this stuff, it has never ignited at this point.  But I have never left it cooking unattended either!

Finally, all of the solids are dissolved, the mixture turns clear.  Five minutes can seem so long, but this IS a watched pot.

Roughly half goes into each pan.  Seems like I should have to measure here to make sure each pan gets exactly half, but small discrepancies don't make much difference.  And I like to pour it out quickly, for fear that the skin which forms so readily would stick in whatever measuring vessel I might use, unbalancing the mix.

Pans go into the oven, preheated to 300 degrees Farenheit (150 degrees celsuis).  Please note that the color of the rear pan is an illusion:  it is really clear white, not amber.  This is due to the rear pan being lit more by the oven-light, the front pan lit more by the camera flash.  I assure you that in real life they are the same color.  The camera does lie, at least this camera.

Pans after 10 minutes.  Just beginning to bubble up.

Pans after 25 minutes.  Bubbles are well-formed.  There is still some liquid in the middle of each pan.

Pans after 35 minutes.  No liquid is apparent, but the flakes and bubbles are all still bluish-white.

Pans at 40 minutes, and about done.  The color of the mixture has subtly changed from bluish-white to light ivory.  This does not show in the photos.  Again note that the amber color of the rear-left pan is an illusion - they are both the same color, more like the front-right pan..

Note that the mixture is NOT STIRRED until it is completely cooked.  Breaking up the bubbles prematurely will inhibit evaporation of the water, making for a very slow process from that point on.  So don't stir this mixture until it has passed the snap-test described below.

At 45 minutes, a pan is removed from the oven, and a small sample is taken.

It is rolled into a pea-sized ball, pressed flat on a cool, dry surface and allowed to harden until cooled, perhaps 20 seconds.  The pan is returned to the oven while the sample cools.

The cooled sample is bent in half.  If the sample bends over without breaking, the mixture is too moist and should be cooked a bit longer.  Sample every 3 minutes or so until it snaps cleanly when bent.

(Time-warp:  more recently I have found that a good texture can be achieved by leaving a bit of moisture in the mix.  If the cooled sample is stiffly flexible, it may be OK.  This might reduce the brittleness of the resulting propellant, but that has yet to be determined.)

In this case, I took the pans out of the oven just in time.  The cooled sample snapped cleanly, indicating proper texture.  In fact it was a little grainy, almost too dry.  Five more minutes and I would have needed to rescue it. *

* To rescue, add 1 tsp of water to each pan, cover with a stoneware plate, and return to the oven for 15 minutes, then test.  Contunue cooking and repeating the testing procedure until proper texture is achieved.

Once the mix passes the snap-test it is done and should be processed immediately.  Time to scrape it up.  This is a danger-point at which you should put on all your protective equipment. I have never had an accidental ignition with this propellant, but keep thinking that if these loose flakes were to happen to catch fire, they would fly all over, causing lots of excitement.

The flakes in each pan are scraped up, and both combined in one pan.

I use a strong tablespoon to press, stir, and knead the flakes until they begin to consolidate.  They may resist sticking together at first.  Persist.  Unless you have cooked it too dry, it will stick eventually.

Another new trick:  Mound up the flakes and tap on the pile using a wooden or plastic mallet.  This will encourage them to consolidate with much less effort than simply stirring and pressing.

After a minute or two of cooling, the propellant is rather hot but it can be handled lightly.  I am tossing it from one hand to the other to maintain good relations with my nerve cells.

Here I am breaking off a small sample to roll into a stick.  Note the board covered in denim.  Smooth cloth makes an excellent rolling surface for this propellant.

You can roll this into nice sticks, just like clay.  It will harden as you roll, allowing precise hand-forming.

Patting the lump down so that I can cover it with a plate.  This lets me keep it warm and soft a bit longer.  The oven temp is reduced to 200 degrees Farenheit (93 degrees celsuis.)  The propellant can be kept covered and soft at this temperature for a long time with no degradation.

An alternative to patting it down by hand:  Put it in a food-processor!  This must be done outside, or in a very safe place.  I plug the food-processor into an unplugged drop-cord, turn it on, then plug the cord into an outlet from a safe distance.  I have yet to experience an accidental ignition, but there is always a chance.

After about 1 minute, the propellant makes a ball that goes 'round and 'round.  The wetter batches will make a slurry that never balls-up, but whenever I am satisfied that it is well-blended, I unplug the machine and take out the propellant.  It is usually rolled into smaller balls which are allowed to cool and placed in a tight container.

I always save out a piece to make into sticks for testing burn-rate.

A chunk of propellant is rolled by hand  into a rod 1/4 inch in diameter.   A piece is cut exactly one inch long, and one end ignited..  This configuration is called a "strand" of propellant.  Ignition does not really require a propane torch, it's just a good, stable flame source that leaves one hand free to hold the strand, the other free to hold the stopwatch..

I start my stop-watch the instant the strand ignites, and stop the watch the instant it burns out.  Since the rod will burn from one end to the other, this provides a measure of its burn-rate in open air.  This sample burned one inch in 11 seconds, which is about average for a light batch.  Some batches burn 8 or 9 seconds per inch.  It will burn much faster under pressure, as in a rocket motor.

The stick has cooled, so I break it into pieces and put them in a film-can.  This propellant will keep for years if tightly sealed.  I use these 1/4-inch diameter sticks for whizzers and micro-rocket engines.

Be sure to label each container clearly, and store in a safe place.

Perhaps the most remarkable feature of this propellant is that can be re-heated and re-formed into most any shape.  Placed in a 200 degree oven until heated through, it regains its putty-like texture and can be formed by hand like clay or pressed into a mold.  Gentle hand-pressure is adequate to make it conform to the shape of a mold.

For example, I will make a propellant grain for a model-rocket sized engine.  This is a simple, cylindrical grain 5/8 inch in diameter with a 1/8 inch perforation going all the way through the long way. It is 1-1/8 inches long and weighs 10 grams.  It is primed with a touch of commercial FFFG black powder.

Cut off a chunk of about 10 grams or a little more.

Roll it into a cylinder about the size of your rocket-engine tube.  In this case, the tube is 5/8 inch inside diameter.

Sprinkle a pinch of 3F black powder on the rolling-board, then roll the propellant pellet over it until the grains stick.  If they won't stick, put the propellant back in the oven at 200 degrees Farenheit (93 degrees celsuis) for a few minutes until it is soft again.

Make a hole through the middle of the propellant-charge, going the long way.  Make it big enough for your fuse to pass through.  Here I am using a bamboo skewer such as is sold for making shish-kebabs or tempura.  I use these sticks for so many things....

Now I try fitting the propellant grain into a 5/8 inch tube.  It is a bit too fat.

So I roll it again.  It is a bit harder now, and so can be rolled more precisely.

Didn't show you, but I weighed this pellet and found it to be over 12 grams.  That will almost certainly burst the case, so I cut off a bit with a sharp knife.  It weighed 10.2 grams afterward.  Close enough.

Now it fits quite well, just slipping into the 5/8 inch tube.

Since I won't be using this grain right away, I will put in a film-can.  Along with the rest of this batch.  Just make wads and slugs of propellant that will fit into your containers, put them in warm, and seal them up.  Lay the cans on their sides while they cool.  If you stand them on end and the propellant runs down to the bottom and cools there, you will have quite a time getting it out.

Jimmy Yawn
rev 2/16/08