THEORY


Introduction Safety Ingredients & Materials Formulations & Techniques Pictures & Video Contact & Links


Unlike many other pyrotechnic devices, the reaction processes inside of a sparkling Senko Hanabi remain largely unknown.


The following information was extracted from two articles that are freely available on the internet. They're from the hand of Hideaki Ito, a Japanese chemist. The articles are available in Japanese PDF only. What I did was to run a character recognition program over the pdf's and feed this raw japanese text into three online translators (Google, Livedoor and OCN). The resulting english text is sometimes totally nonsensical, but with a little bit of background, it's possible to derive the most important information from the documents. However, errors may and will be present in the following paragraphs. If you know someone who could translate the articles in person, please let me know, so that I can point out which paragraphs are worth investing in.




One article mainly discusses the fabrication of a strand of 'Senko Hanabi', both in the normal way with KNO3 as in an alternative fashion using K2Co3 as potassium-donor...

Simple Preparation of Japanese Toy Firework.pdf


The other article then goes on to hypothesize on the reaction mechanism...

The mechanism of Spark Generation from Japanese Toy Firework.pdf

Very roughly, it comes down to the following:

  1. first the KNO3 reacts with the charcoal, emitting heat
    4KNO3 + 2C > 2K2CO3 + 2N2 + O2
  2. The resulting K2CO3 then goes on to react with the sulphur
    8K2CO3 + (3n+1)S2 > 6K2Sn + 2K2SO4 + 8CO2
  3. K2SO4 and C react
    4K2SO4 + 7C > 2K2CO3 + 2K2S2 + 5CO2
  4. The polysulfide reaction: K2Sn, carbon and oxygen
    The hexagonal molecular polymer of carbon-layers can be taken apart by the potassium ion and sulphur-tail to form fragments of carbon-hexagons that react with atmospheric oxygen (available through convection at the side of the fire ball). It seems that this are the fragments shooting out of the fire ball, accellerated by the production of carbon dioxide. This would mean that the reaction can be stimulated by providing more oxygen or stopped by removing oxygen from the atmosphere. I decided to test both hypotheses myself:

    Here is a video of a strand of Senko Hanabi burning in pure oxygen
    (produced from hydrogen peroxide and MnO2)

    Although it may seem that the reaction is more vigorous, close inspection shows that this is largely due to the burning of the paper. The droplet doesn't seem be too influenced.


    Here is a video of a strand of Senko Hanabi suffocating in pure CO2
    (produced from hydrogen chloride and CaCO3)

    In this case, it's clear that the reaction is influenced.

As can be derived from the experiment without NO3- (a mixture of K2CO3, sulphur, charcoal and soot heated in al alcohol flame), the first reaction is only necessary to bring the temperature of the slag to 850°C. No more nitrate seems to be present in the reaction slag. The overall reaction seems also strongly influenced by the viscosity and density of the fire ball. The amount of soot present in the mixture has the function of controlling (lowering) the burning temperature as well as of lowering the density of the fire ball. Percentages of 16% and up steeply diminish heat output (by blocking the space between layers of carbon present in the charcoal?). The addition of more soot may prevent the fire ball from falling. The falling of the fire ball is also attributed to the amount of ash present in the paper and charcoal. In the article, it is hypothesized that charcoal with lower ash content may be beneficial.


Perhaps that in the future, visitors will share what they know on the mysteries of the Senko Hanabi mechanism. If they would, I would be most thankful.