At a temperature of $\pu{750 ^\circ C}$, $\ce{K2CO3}$ and $\ce{SiO2}$ react with each other in substantial proportions, forming $\ce{K2SiO3}$ and $\ce{CO2}$ in the process$\ce{^{[1]}}$.

We all know as well that potassium hydroxide reacts with glass. So, I was wondering if $\ce{K2CO3}$ was reacting with pyrex and/or normal glass in minute amounts, either in solution at say >$\pu{150 ^\circ C}$ or when being dried under vacuum in an oven.

One putative indirect mechanism in my wild imagination would be if $\ce{CO2}$ could leave the solution (maybe only at higher temperatures) and if potassium hydroxide at all existed in solution, upon the dissolution of $\ce{K2CO3}$ in water?


  1. Shikha Agarwal, In Engineering Chemistry: Fundamentals and Applications; Cambridge University Press: Delhi, India, 2016 (ISBN: 978-1-107-47641-7).
  • 3
    $\begingroup$ No clue at all about possible mechanisms, but some minor level of reaction seems plausible. After all, Gorilla glass involves reactions of glass with molten potassium salts. $\endgroup$ – Ed V Jun 21 '20 at 1:40
  • 3
    $\begingroup$ Even at lower temperatures (300 to 373 K) and dissolved in water, alkali metal carbonates attack glass slowly and will etch it. Ground glass fittings are difficult to take apart in equipment that has held the carbonates for a long while. Of course, some glasses are more resistant than others; see ISO rating: iso.org/standard/4895.html $\endgroup$ – DrMoishe Pippik Jun 21 '20 at 4:26

In agriculture, potassium silicate ($\ce{K2SiO3}$) is used as foliar fertilizer ($\ce{K2SiO3}$ is water soluble). The major synthetic method of the preparation of $\ce{K2SiO3}$ is heating $\ce{SiO2}$ and $\ce{K2CO3}$ at high temperatures $(\pu{600-850 ^\circ C})$ in various mole ratios (e.g., Ref.1):

$$\ce{SiO2 + K2CO3 ->[\Delta] K2SiO3 + CO2} \tag1$$

According to Ref.1, alkali metal carbonates such as $\ce{Na2CO3, K2SiO3, Rb2CO3,}$ and $\ce{Cs2CO3}$, all give corresponding metal silicates in this reaction (e.g., silica to metal carbonate mole ratios used in these preparations are $1:2$ to $3:1$).

Keep in mind that the temperatures used in these preparations are below the melting points of both reactants, $\ce{SiO2}$ ($\pu{1713 ^\circ C}$) and $\ce{K2CO3}$ ($\pu{891 ^\circ C}$), thus, the reaction is essentially a solid state one.

Potassium silicate fertilizer grades were successfully produced by direct fusion of silica ($\ce{SiO2}$) and potassium salt ($\ce{KOH}$ and $\ce{K2CO3}$) in furnaces at temperatures up to melting point of the mixtures (melt temperatures are as high as $\pu{1350 ^\circ C}$). The range of the weight ratio of silicon dioxide/potasium solid used in this method is $1:1$ to $5:1$. The reaction between $\ce{SiO2}$ and $\ce{KOH}$ is being:

$$\ce{SiO2 + 2KOH ->[\Delta] K2SiO3 + H2O} \tag2$$

According to the Ref.1, the other salts (e.g., halides and sulfates) of the metal carbonate catalyze the reaction. For example, in preparing sodium silicate, 6.12 parts by weight sodium carbonate, 3.47 parts by weight silica of minus 325 mesh particle size, and 0.41 part by weight sodium sulfate were ground together (which is $\ce{SiO2 + Na2CO3 + 0.05Na2SO4}$). The ground mixture was heated at a temperature of approximately $\pu{700 ^\circ C}$. for $\pu{4.5 h}$, after which time, the reaction was found to be 51% complete (c.f., without catalyst under same condition, completion was ~$23\%$).

Based on these date, it is safe to say that when silica made contact with an alkali carbonate or hydroxide at elevated temperatures, we can expect a reaction between two chemicals, regardless of they are been in solid form or melt. Once washed, the formed metal silicate would be washed away by water, leaving silica with damaged surface. According to Wikipedia:

The most familiar, and historically the oldest, types of manufactured glass are "silicate glasses" based on the chemical compound silica (silicon dioxide, or quartz), the primary constituent of sand. Soda-lime glass, containing around 70% silica, account for around 90% of manufactured glass.

Therefore, it is no wonder when you applied alkali hydroxide solution (e.g., $\ce{NaOH}$ or $\ce{KOH}$) solution to glass surface for a long time and found the surface become roughed. The alkali carbonated would do the same in higher temperatures.

Note: $\ce{Li2SiO3}$ is only partially soluble in water. However, its $\ce{Li4SiO4}$ version has an advantage of being $\ce{CO2}$ absorber (Ref.3) according to the reaction:

$$\ce{Li4SiO4 + CO2 ->[\Delta] Li2SiO3 + Li2CO3} \tag3$$

Unlike other carbonates, two $\ce{Li2CO3}$ molecules react with one $\ce{SiO32}$ molecule to give $\ce{Li4SiO4}$:

$$\ce{2Li2CO3 + SiO2 ->[\Delta] Li4SiO4 + 2CO2} \tag4$$


  1. Isadore Mockrin, “Production of Silicates,” US Patent 1958, 2,823,098 (PDF).
  2. Srie Muljani, Bambang Wahyudi, Ketut Sumada, Suprihatin, “Potassium silicate foliar fertilizer grade from geothermal sludge and pyrophylite,” MATEC Web of Conferences 2016, 58, 01021 (DOI: 10.1051/matecconf/20165801021) (Proceedings of the 3rd Bali International Seminar on Science & Technology)(PDF).
  3. Xianyao Yan, Yingjie Li, Xiaotong Ma, Jianli Zhao, Zeyan Wang, “Performance of $\ce{Li4SiO4}$ Material for CO2 Capture: A Review,” Int. J. Mol. Sci. 2019, 20, 928 (22 pages) (DOI: 10.3390/ijms20040928).
  • $\begingroup$ Thank you mathew. Do you know anything about what happens in the sub 200C in aqueous medium? $\endgroup$ – Hans Jun 21 '20 at 18:05
  • 1
    $\begingroup$ I'm sorry I dont know. But, we all know that $\ce{NaOH}$ in any concentration during titration experiments, we keep the solution in plastic bottles because $\ce{NaOH}$ reacts with glass and concentrations get reduced giving us wrong determinations. $\endgroup$ – Mathew Mahindaratne Jun 21 '20 at 19:14

Yes it does.

Some years ago I have made an experiment to extract potash K2CO3 from oak ashes.

I have evaporated the filtered leach liquid in an Pyrex dish while having the liquid close to the boiling point.

This whole operation made the Pyrex dish milky looking – respectively the K2CO3 did attack the glass.

  • $\begingroup$ What was the concentration and temperature? Did you measure how much loss? I expect it to be milder than for the hydroxides but I'd love to know how much less (if at all). $\endgroup$ – Hans Jun 22 '20 at 11:58
  • 1
    $\begingroup$ I have evaporated to dryness. So the concentration was at a maximum at some point. I dont know about the loss... I can only guess about it. If I have to guess, its just a few mg, but enough to make the glass milky. $\endgroup$ – Andreas Jun 22 '20 at 12:05
  • $\begingroup$ About what concentration do you think you were at when the solution started to be milky? $\endgroup$ – Hans Jun 22 '20 at 13:19
  • 2
    $\begingroup$ the solution did not turn milky - the glass turned milky as it got etched. $\endgroup$ – Andreas Jun 22 '20 at 13:59
  • 1
    $\begingroup$ at some point crystals of K2CO3 formed in the solution, thats where the maximum concentration was reached. $\endgroup$ – Andreas Jun 22 '20 at 14:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.