Silicon production from quartz

Silicon (Si) is a chemical element with a wide array of applications. It's used in high strength alloys such as ferrosilicon, but by far the biggest and most well known is its use as a semiconductor in circuit boards. With the rise of computer technology, silicon is playing an ever larger part in the modern era.

I investigated two methods for preparing elemental silicon, the first led to an extremely impure product so I chose not to present it. The second method is a thermite reaction between silicon dioxide and aluminium. Since this reaction isn't thermally self sustaining, a side reaction of sulphur and aluminium is introduced to provide the necessary heat. Many sources of silicon dioxide can be used for the reaction such as silica gel and sand, but I chose to use quartz.

To begin, I added 11.87g (0.1976 moles) of finely powdered quartz, 15.83 (0.4936 moles) of sulphur and 13.19g (0.4888 moles) of 325 mesh aluminium powder to a plastic container. I stirred this mixture intimately until a uniform gray powder was obtained, then poured the powder onto a brick. I placed a piece of magnesium ribbon in the mixture and ignited the ribbon with a torch. A very intense, but fairly slow burning thermite reaction began, giving off large amounts of heat and smoke. Once the thermite was complete, I scraped up the residue and added it to a 500ml beaker containing 200ml of water. After a few seconds, the mixture began giving off a huge amounts of hydrogen sulphide gas, this lasted about 30 minutes. I allowed the mixture to stand with occasional stirring for about 24 hours to insure the evolution of hydrogen sulphide gas was complete. The mixture became a clear liquid with a thick gray precipitate of aluminium hydroxide. Anyway, I stirred the mixture to form a suspension, waited a few seconds, then decanted off the aluminium hydroxide suspension.

With an additional 200ml water added each time, the decantation step was repeated 4-5 times until almost all the aluminium hydroxide had been removed. Left at the bottom of the beaker were a lots of silicon beads, which appeared metallic with a slight bluish luster. I added 50ml of 33% hydrochloric acid and left the silicon to soak in this for 15 minutes with occasional swirling of the beaker. I then decanted off the acid and added 200ml of water to the silicon. I filtered the mixture through a regular kitchen sieve to remove the fine grains of unwanted material and washed the silicon beads caught by the sieve with two 200ml portions of water. After drying, I was left with 3.19g of elemental silicon representing a 57% yield.

Left = silicon beads  /  Right = quartz and aluminium thermite with sulphur

3 SiO2 + 4 Al ==> 2 Al2O3 + 3 Si    /    2 Al + 3 S ==> Al2S3

Al2S3 + 6 H2O ==> 2 Al(OH)3 + 3 H2S   /   Al2O3 + 6 HCl ==> 2 AlCl3 + 3 H2O

Al(OH)3 + 3 HCl ==> AlCl3 + 3 H2O

Benzyl chloride via free-radical halogenation

Benzyl chloride, or a-chlorotoluene, is an extremely useful reagent used in organic synthesis. Commercially, it's mainly used to make benzyl esters, which find use in the perfume industry. Unfortunately, benzyl chloride is an extremely dangerous chemical and has been used historically in chemical warfare due being a strong lachrymator (severely burns eyes, like a tear gas) and very toxic. I plan to use benzyl chloride to make benzylamine.

I made some benzyl chloride using a procedure that I modified slightly (link). Toluene is halogenated directly with chlorine in the presence of a halogen lamp to encourage the equillibrium shift of diatomic chlorine molecules to free radical chlorine atoms. Benzal chloride and benzotrichloride are also produced, but the quanitity of these side products can be kept reletivly low by using a large excess of toluene.

To a 500ml round-bottom flask, I added 100ml (0.9442 moles) of technical grade toluene and 2.5ml of 33% hydrochloric acid. I attached a Liebig condenser to the flask and refluxed the mixture on a medium heat for 2 hours with a halogen lamp placed a couple of centimetres from the flask. During reflux, I slowly added in 36.56g (0.1573 moles) of trichloroisocyanuric acid (TCCA) in small portions through the condensor. Complete addition took 1 hour. Sometimes the small amounts of the TCCA become stuck in the condenser so a long rod was used to push it down into the flask. With the first few additions of TCCA, a considerable amount of heat was given off and the formation of yellow-green chlorine gas was observed. For the second hour of reflux, the flask was ocasionally swirled. Anyway, after the reflux, I turned off the heat and left the mixture to stand overnight. The next morning, I filtered off the precipitated white cyanuric acid, transferring the orange filtrate to a 500ml round-bottom flask. I attached a stillhead and condenser to the flask and began distilling, using a thermometer to track the fractions.

The first fraction to come over was excess toluene at 100-124 C. I collected around 35ml of this. The vapour temperature then rose and I began to collect benzyl chloride at 160-186 C. After all the benzyl chloride had come over, there was nothing left in the distilling flask and I ended the distillation. I then redistilled the benzyl chloride, this time collecting the fraction boiling at 160-182. The residue in the boiling flask should consist of mainly benzal chloride and benzotrichloride. Anyway, I dried the benzyl chloride over anhydrous calcium chloride, then transferred it to an amber vial for storage. I got 16ml (0.139 moles) of benzyl chloride which is a 29% yield. The product had a density of 0.921g/cm³ which isn't too far away from the theoretical 1.1 g/cm³. This shows the benzyl chloride I made is probably moderately pure.

The lachrymatory properties of benzyl chloride were actually nowhere near as bad as I was expecting. My eyes were attacked viscously when I was washing the glassware and briefly during the initial reflux but that was about it.

The pour yield could likely be dramatically improved by using a larger excess of toluene. Adding the TCCA more slowly probably would have also helped to increase the yield. I did notice there were a lot of higher chlorination products left over in the distilling flask after the second distillation.

C3N3O3Cl3 + 3 HCl ==> 3 Cl2 + C3N3O3H3

C6H5CH3 + Cl2 ==> C6H5CH2Cl + HCl

Iron(iii) oxide

Iron(iii) oxide, otherwise known as ferric oxide, is a brick red compound used in the product of iron and steel. It also finds use as a pigment. iron(iii) oxide happens to be the principle constituent of rust and this is probably how most people encounter the compound. Just for fun, I decided to try making some.

18.3g (0.0658 moles) of iron(ii) sulphate heptahydrate were dissolved in 75ml of water in a 250ml beaker. In a separate flask, I dissolved 11.06g (0.1316 moles) of sodium bicarbonate in 110ml of water. Since the dissolution of sodium bicarbonate is endothermic, slight heating was needed to dissolve everything. I then added the iron(ii) sulphate sulphate to the sodium bicarbonate solution in small portions. A white precipitate of iron(ii) carbonate formed and the mixture foamed as carbon dioxide was given off. After stirring for a few minutes, I filtered the mixture. By this time, the iron(ii) carbonate precipitate had turned green and in some places red.

Anyway, I dried the filtered off iron(ii) carbonate, which became completely red. Whats happening is the iron(ii) carbonate, which is a white solid, slowly reacts with oxygen in the air to form iron(iii) oxide which is red. Once the material was completely dry, I transferred it to a crucible and gassed it with a butane torch for 10 minutes to make sure the conversion to iron(iii) oxide was complete. During heating, the colour became a darker, richer red as everything was oxidized to iron(iii) oxide. Once the iron(iii) oxide had become uniformly red, I transferred it to filter, washed it on the filter with 50ml of boiling water then finally dried it. I got 3.23g (0.0202 moles) of iron(iii) oxide. The product is probably not pure and likely contains some Iron(ii,iii) oxide.

2 NaHCO3 + FeSO4 ==> Fe(HCO3)2 + Na2SO4   /   Fe(HCO3)2 ==> FeCO3 + H2O + CO2   /

2 FeCO3 + O2 ==> Fe2O3 + CO2   /   FeCO3 ==> FeO + CO2   /   2 FeO + O2 ==> Fe2O3

 4 FeO ==> Fe3O4 + Fe   /   4 Fe + 3 O2 ==> 2 Fe2O3   /  4 Fe3O4 + 2 O2 ==> 6 Fe2O3

Nitrostarch

Nitrostarch is not a specific compound, but rather a mixture of different esters of nitric acid and starch. It's a secondary explosive and was used as a filler in grenades in world war 1. The energetic properties of nitrostarch are very similar to nitrocellulose, both consisting of chains of glucose molecules with varying degrees of nitration.

Nitrostarch is easily prepared by the mixed acid nitration of starch. Normally a mixture of nitric and sulphuric acids is used, but I chose to use potassium nitrate and sulphuric acid which forms nitric acid in situ.

To begin, I poured 36ml of 98% sulphuric acid into a 250ml beaker in an ice bath. Once the temperature of the acid had dropped below 10 C, I began adding 11.2g of potassium nitrate in small portions with good stirring. During the additions, the temperature was never allowed to rise above 20 C. Once all the potassium nitrate had been added, I added 10g of corn starch to the mixture in one portion with good stirring. The mixture became quite viscous and turned slightly yellow. I stirred the mixture continuously for 15 minutes, then left it to stand for 2 hours with occasional stirring, maintaining the temperature at below 10 C. After this, I began cautiously adding 400ml of water to the reaction mixture. Once a small amount of water had been added, the mixture became fluid enough to be easily poured, so I transferred the mixture to a 1000ml flask before adding the rest of the water. I left the mixture to stand until all the nitrostarch had settled to the bottom of the flask, then decanted and discarded as much of the cloudy supernatant liquid as I could without losing any nitrostarch. I then added a further 400ml of water to the nitrostarch and repeat the above process (letting then nitrostarch settle then decanting as much of the supernatant liquid as possible).

Again, I added 400ml of water to the nitrostarch that was left but this time also added in 3ml of 100% undiluted nitric acid. I boiled the mixture for 10 minutes with good stirring, covering the top of the flask to help prevent loss of the volatile nitric acid. Still heating, I then neutralized the mixture with sodium bicarbonate until the liquid became neutral-slightly alkaline. By mistake I sometimes added a bit to much bicarbonate and the mixture bubbled over a couple off times. I think this may have led to fairly substantial loss of product. Anyway, after the neutralization, I filtered off the nitrostarch, washing it on the filter with 1000ml of water. After drying, I was left with 2.62g of nitrostarch as a white powder with a slight yellow colour.

Left = nitrostarch   /   Right = deflagration of nitrostarch

KNO3 + H2SO4 ==> KHSO4 + HNO3

HNO3 + 2 H2SO4 <==> 2 HSO4 (-) + NO2 (+) + H3O (+)