Silica Aerogel (TEOS, Base-Catalyzed)

Editor’s Note: This is an adaptation of the silica aerogel procedure from the Lawrence Berkeley National Laboratory site about aerogels, which for a long time was the only procedure for making aerogels publicly available. That procedure, we’re sorry to say, does not work. Maybe you’ve tried it. If you have, you’ll have noticed that the solution stays separated as two layers and a gel never forms. That’s because there’s not enough alcohol. Maybe it was a typo. So we modified that procedure and present the modified version that works for us below.  If for some reason you have trouble with the procedure below, please leave a comment!


  • Tetraethoxysilane (tetraethyl orthosilicate), Si(OC2H5)4
  • Absolute (200-proof) ethanol
  • Deionized water
  • Ammonium hydroxide, 28-30 wt % in water
  • Ammonium fluoride, NH4F


  • Acetone

Gel Preparation

An Excel calculator for determining amounts of chemicals required by target volume (mL) or mass (g) is available.

  1. Weigh 1.852 g NH4F and add it to 100 mL of water. Add 20.50 g (22.78 mL) ammonium hydroxide solution. Store this in a bottle so you can reuse it later. This is the “ammonium fluoride/ammonium hydroxide stock solution”. If you already have stock solution prepared you can skip down to step 2.
  2. Mix 4.7 g (5.0 mL) TEOS and 8.68 g (11.0 mL) ethanol in a beaker. This is the “alkoxide solution”.
  3. Mix 7.0 g (7.0 mL) water and 8.68 g (11.0 mL) ethanol in another beaker. Add 0.364 g (0.371 mL, ~8-10 drops from a disposable pipette) of ammonium fluoride/ammonium hydroxide stock solution. This mixture is the “catalyst solution”.
  4. Pour the catalyst solution into the alkoxide solution and stir. This is the “sol”.
  5. Pour the sol into molds and allow gel to form. Gel time is approximately 8-15 min.

What Everything Does

TEOS is the source of the silica. Water is what hydrolyzes the TEOS so that it can polymerize. Ethanol is a co-solvent that is miscible with both TEOS and water to get both into the same phase so they can react. Ammonium hydroxide is a basic (alkaline) catalyst that helps to make the reactions go faster. Fluoride ion is a catalyst that helps hydrolysis happen more quickly.

What Doesn’t Work

  • Not using ammonium fluoride. It actually makes a big difference with TEOS. Although fluoride also makes reactions with TMOS go faster, TMOS will work fine with just a basic catalyst without fluoride.
  • Using denatured alcohol that contains anything other than methanol or isopropanol as a denaturant instead of absolute ethanol. Some hardware store alcohol works, some doesn’t.
  • Using sodium hydroxide (NaOH) instead of ammonium hydroxide in equal molar concentration. NaOH is a strong base so if you use it you’ll need to use a lower molar concentration of it than for ammonium hydroxide.

Variables You Can Play With

  • Try adjusting the amount of solvent used to adjust the density of the resulting aerogel.
  • Try adjusting the amount of the catalysts in the stock solution or the amount of stock solution you add. This will change the gel time and possibly the clarity of the gel (more catalyst means faster gel time but possibly lower transparency).
  • You can substitute sodium hydroxide, sodium carbonate, or potassium carbonate for ammonium hydroxide but you will have to experiment with the amount.
  • You can substitute sodium fluoride for ammonium fluoride in equal molar concentration, although your gel time may be affected since you lose the buffering effect of the extra ammonium ions.

Gel Processing Conditions

  1. Once the gel has set, place it under ethanol and allow the gel to age for at least 24 h.
  2. Exchange into 200-proof ethanol or acetone at least four times over the course of several days to a week.
  3. Supercritically dry. A suggested procedure would be to heat the CO2 through its critical point (31.1°C and 72.9 bars) to ~45°C while maintaining a pressure of ~100 bars. Depressurize at a rate of ~7 bar h-1.

What You Should Get

A transparent silica aerogel with a blue cast from Rayleigh scattering that appears yellowish when viewed in front of a light source from Mie scattering.

  • Density 0.040 g cm-3
  • Surface area 700 m2 g-1

Useful Information

Tetraethoxysilane (tetraethyl orthosilicate):

  • Molecular weight 208.33 g mol-1
  • Density 0.933 g mL-1
  • Smells a little bit like spearmint
  • Sigma-Aldrich part number 131903


  • Molecular weight 46.07 g mol-1
  • Density 0.789 g mL-1
  • Sigma-Aldrich part number 459836 or 459844, or get Everclear from a liquor store

Ammonium fluoride:

  • Molecular weight 37.04 g mol-1
  • Form is a fluffy, lightweight solid
  • Sigma-Aldrich part number 216011

Ammonium hydroxide:

  • Concentration is 28-30 wt % in water typically
  • Molecular weight of NH4OH is 35.05 g mol-1, but this is not the molecular weight of the solution
  • Density 0.9 g mL-1
  • Form is a pungent liquid that smells like cleaning ammonia, use in a vent hood
  • Sigma-Aldrich part number 221228

51 Responses to “Silica Aerogel (TEOS, Base-Catalyzed)”

  1. Melaku says:

    Thank you guys, Big respect !!! I have got more information to start my project and your blog is one of the few which share knowledge for free. I’m gonna test it today and I will let you know the results. one thing I would like to say is plz add more drying methods other than super or subcritical methods, for example, drying method for aerogel (Biosensors).

  2. Fahim says:

    I’m trying to make a lightweight aerogel that can take a lot of stress. I’m not sure whether to make TMOS or TEOS aerogel. Any suggestions?

  3. Guests says:

    Please support us on the Pyrogel XT high temperature insulation blanket.

    Any ideas where I could get a contact for the services?

    The other thing is, with Aerogel Technologies, Buy and, are these three websites the same company or?

    Please provide us and let us know as soon as possible.

    We are looking forward to purchase the Pyrogel XT blankets as in need.

    Thank you to dear sir or madam for the supports.

  4. Rio says:

    Hi, Im working on lithium silicate synthesis and TEOS as the silica source while lithium nitrate as lithium source. What bother me is the gelation time is very rapid, almost around 30 minutes only the sol already transformed to clear white gel. But for the ethanol , I used ethylene glycol and acetic acid as catalyst.

    I wonder if the gel produced is really make a strong bond and got reaction inside or not. Please advise. Thank you.

  5. knowledgeforall says:

    What are the by-products that are removed during the alcohol replacement process and can you quantify it?

  6. Sanjiban Chakraborty says:

    I have successfully prepared the aerogels with TEOS and also prepolymerized TEOS (Silbond-H, USA). Could you please tell me or send me the data for the pore size distribution and what method did you use to calculate that?

  7. Laurie says:

    Could you share the procedure for making the insulating fabric/material/blanket?

  8. James McGoff says:


    I am doing a school project that requires me to build an insulated container, much a cooler. I need this to preserve temperatures for up to 18 hours despite ambient temperatures ranging from -4C to 37C. Obviously, I thought I’d use aerogel as the main component for the insulation. I ordered some and have been confused by the insulation properties, here’s what I mean:

    1. I will place about 700cc of granular aerogel into a plastic kitchen container.
    2. I will place some ice (4-5 cubes) in a small plastic bag, seal it, and bury it down within the aerogel.
    3. I will put a lid on this and let it sit at room temperature.

    Now, the reason that I have been disappointed is that the ice will only last perhaps 6-7 hours before melting. Surely there is a way to improve this. I was thinking that perhaps, since I am using granule aerogel, the air trapped between the particles is the convective heat transfer pathway.

    If I were to vacuum all of the air out of the bag, do you think the ice within would last any longer?

    Thank you for your time and help!


    • Stephen Steiner says:

      Hi James,

      Good questions and experimentation! First off, I would recommend you chill your container and its contents before your start the clock, so to speak–if you put a bag of ice cubes in a container packed full of granules such as Cabot’s Lumira aerogel, but the container is warm, any heat that can transfer will and melt the ice. Make sure the bag you’re putting the ice cubes into is cold (freezing temperature) when you put the ice in and place it into the container as quickly as possible. If the bag is warm, you’re basically forcing intimate contact between the warm bag and cold ice cubes with no where else for the heat to go but into the ice cubes with the surrounding aerogel particles.

      Second, regarding evacuating the container, yes I expect that you would see improved performance by removing the air, thereby reducing convective transport. Make sure your container is opaque or covered in aluminum foil to prevent radiative transport from the outside as well.

      Third, make sure you tap the container to settle the aerogel particles.

      Lastly, remember the thermal conductivity of aerogel isn’t zero, so perhaps using a larger quantity of aerogel particles would help–you want to make sure you have several centimeters of particles on all sides of the bag. If the bag is closer to one side of the container than another, for example, in contact with the bottom, then heat will work its way through there and melt the ice.

      Please post updates on your findings for others to learn from!

  9. chris says:

    how much aerogel can you get for the $100-$200?

    • Stephen Steiner says:

      Hi Chris,

      It really depends on what type of aerogel material you need. If you are looking for a monolithic (classic) silica aerogel material, you can get up to a 2″x3″ tile (which sell for ~$200 on If you are looking for insulating blankets such as Cabot Aerogel’s Thermal Wrap or Aspen Aerogel’s Spaceloft, you can get ~4 ft x 2 ft for under $200 (noting that if you were an architect and were using thousands of square feet it would be a lot cheaper). If you are looking for aerogel particles such as Cabot Aerogel’s Lumira Aerogel, you could get about a gallon. These prices are based on what’s available commercially on (which sponsors, but you may find different prices for these different materials with different vendors.

  10. several questions:

    o I’m interested in different formulas for making synthetic opal(ine materials)
    (ie: monodisperse silica spheres in the range of ~200 – ~350 nm)
    It seems that your process would produce that, instead of a gel, becahse the pH is so high.
    (& I have it in my head that a gel forms when the pH is lowered)
    How might I alter your formula to yield what I want?
    And how am I wrong in the above?
    o It’s very very confusing to have at the top of your Page:
    “That procedure, we’re sorry to say, does not work.”

    ..and then, in a reply from *** (who I took to be one of you (responsible for this Page)):
    “Yes, the method works (well it should anyway–we’ve tested it thoroughly). The note is just a reflection on our own difficulties with getting previous recipes found on the Internet to work.”

    : If that’s what the note was meant to be, it would be best to have it /say/ that.

    in another Reply (again, sounds like one of you, to me):
    “Well it works for us! We tried it out multiple times with multiple variations to make sure that it would work before we posted it. Also, I believe Jing (see Connect > Questions and Answers) has successfully used it and found it worked well!”

    So why not edit the Editor’s Note?

    • Stephen Steiner says:

      Hi Michael,

      The process here does work–what may be confusing is that we wanted to explain there is a process on the web from Lawrence Berkeley (upon this procedure was initially based) that does not work for us, and that our modified version, which is on this page, does work. I disagree that this process would result in monodisperse silica spheres. It should give a gel. If you observe a different result than a gel when actually implementing, please let us know!

      Regarding making monodisperse silica spheres, you probably won’t get a monodisperse product using an approach like this, as the particle formation is statistical in nature and you will get a polydisperse distribution of particles. You could use a capping agent such as 2-methoxyethanol or acetylacetonate to help form particles over a gel structure–the approach would be to passivate the silica nanoparticles that form (i.e., the particles in the sol) so that they cannot interconnect to form a gel.

      Hope this helps!

  11. Blackheart says:

    Hi, silly question:

    can i make aerogel from agar-agar? make water agar-agar gel and dryng by the method above described?


    • Stephen Steiner says:

      Hi Blackheart,

      Not a silly question at all! You can totally make aerogels with agar. You would need to exchange the water-based pore fluid in the agar gel with something like ethanol or acetonitrile by repeated soakings, and then like a silica gel place it in a supercritical dryer, exchange the organic solvent with liquid carbon dioxide, and then supercritically dry. Actually, Samuel Kistler made such aerogels when he first discovered aerogels! One issue with biologically-derived aerogels you need to be careful with is that proteins often denature in organic solvents and may cause the gel to turn opaque and distort. If this happens, try solvent exchanging into a different polar organic solvent.

  12. taylan nurlu says:

    hhiii actually i dont understand that where is superdrying process in that metod? ( are all process under room conditions ?)

  13. idee says:

    for the Ammonium hydroxide i have 500ml of six molar ammonium hydroxide. How do I convert that to make it 28-30 wt % in water? How much water do I need? How much ammonium hydroxide do I need?

  14. George says:

    I would like to know the cost of the equipment and chemicals required too. Seems very interesting would love if you have a video of what you did.

  15. Jenn says:

    This is quite possibly a dumb question, but I would just like to make sure I am clear. If you supercritically dry the gel with the suggested CO2 procedure, I would be under the impression that the air chambers in the gel would be filled with CO2 gas and not just air. Is this correct?

    My second question would be that we have an interest in actually containing a gas mixture in the finished aerogel. Assuming we can seal the outside in such a way as to actually keep the gas contained (and any suggestions on that front would be appreciated), do you know of a recommended way in which to force a gas into the aerogel?

    • Stephen Steiner says:

      Hi Jenn,

      Your question isn’t dumb. There are no dumb questions, just dumb decisions not to ask questions.

      You are correct that following CO2 drying the pores of the aerogel material are filled with CO2. However, because aerogels are open-porous architectures, when exposed to air the gas in the pores will immediately diffuse out and mix with the surrounding atmosphere and after a period of time (minutes to days depending on the size of the sample), the pores in the aerogel will be basically filled with air.

      The easiest way to force a gas in is to just place the aerogel in an atmosphere of the desired gas and let diffusion do its thing. You may wish to flow purified gas in occasionally to fully replace any pore vapor that diffuses out. You could also build an apparatus in which the aerogel acts like a plug with a pressure of the target gas on one side and a vacuum on the other to accelerate diffusion through the monolith, however this is shape-limited and may be tricky to do.

      Regarding sealing a gas in, good question. One way would be to laminate the aerogel using a polyethylene or Mylar film like the kind used for making vacuum insulating panels (VIPs). This works reasonably well. You could also encase the aerogel in a viscous polymer and set the polymer.

  16. Cman says:

    does anyone know where to buy some aerogel cheap?

  17. mojtaba says:

    how long we should take the gel in supercritical dryer at T=45c &p=100bar?

  18. mojtaba says:

    how long we should take the gel in supercritical dryer at T=45c &P=100 bar?

    • Chris says:

      Mojtaba – The directions state:
      A suggested procedure would be to heat the CO2 through its critical point (31.1°C and 72.9 bars) to ~45°C while maintaining a pressure of ~100 bars. Depressurize at a rate of ~7 bar h-1.

      Based on the de-pressurization rate, it should take about 14 hours. Please correct me if I’m wrong.

  19. soheil says:

    what is break coefficient of sillica aerogel

  20. Evan says:

    I’m confused on what the Editor’s Note means. Does this method actually work, or is the note saying it doesn’t? Thanks for the help.


    • Stephen Steiner says:

      Hi Evan,

      Yes, the method works (well it should anyway–we’ve tested it thoroughly). The note is just a reflection on our own difficulties with getting previous recipes found on the Internet to work. And if you find this recipe does not work, please let us know that is the case and any deviations from the stated procedure as soon as possible!

  21. Justin says:

    I am looking to make some Aerogels for fun and I did a search on the internet about the TEOS chemical. I have found that it is sold in gas form, is it also sold in solid form as well? Any information on where some could be purchased would also be great!


  22. Daniella says:

    I dont really understand the Gel Processing Conditions Part. Could someone explain it to me more precisely. And had someone tried this, because I am interested on proposing this in my class.
    Thank you a lot

    • Vinayak says:

      Gel processing period means fisrstly after hydrolysis and condensation of the metal alkoxide, the gel is formed but there shoud not strong network of silica. So, for that we have to age this gel at 50 degree celcius. After that you will do your drying procedure.

  23. Mike says:

    I have been searching for hours and haven’t found a single company that will ship Tetraethyl orthosilicate to a residential address. Where can you get this without ordering through an employer account?

  24. Mike says:

    Will I be able to add Europium Doped Alkaline Earth Metal Aluminate Glow Powder without an adverse reaction?

  25. Ed says:

    I’m interested in making some aerogel (TEOS) with my two high school age kids. The information on this site is invaluable! Still researching controlling fumes (subcritical drying). Meanwhile, I made a couple of modifications to the spreadsheet which you are welcome to distribute. The percentages where off a little bit (missing about 4 to 5 % when you run a totals column so I made the necessary adjustments. I found a third column interesting to view (now have vol. for desired vol., weight for desired vol. and weight for desired weight). Ran subtotals and totals on all columns as well. Finally I calculated the dried aerogel mass (2 different density estimates). Send me an email if you’d like me to send it to you.

  26. jim bell says:

    Does anybody know where the oxygens in the resulting silica come from? In other words, do they come from the water, the TEOS, or somewhere else?
    Please email to (MIT Chemistry Class of 1980).

  27. Blubba Hank says:

    How toxic are the ingredients and how toxic are the resulting mixtures? I am thinking of trying this recipe in my garage and driveway and would like advice on how dangerous or toxic this activity might be. What personal protection precautions and materials (eye goggles, N95 nose mask, gloves, etc.) should I use? Thanks, Blubba

  28. josh says:

    The clarity of this batch is important information for a base or standard for the other recipies. Aerogel seems to be a interesting option. In Make -> apreseating your work -> characterising aerogels -> Transparency

    I am interested in the inferred and wavelenths that add to heat production.
    Aerogels could be used for a insulating coating like evacuated tube vacume tube solar collectors. I wonder how a aerogel coated solar heat collector would compare to a vacume tube one?? So you can see how transparency is important.

    By comparing the brightness of the beam with and without the sample present, the the transparency of the sample as a function of wavelength can be measured. The results can then be plotted as a graph of % transmittance or % absorbance vs. wavelength.

  29. Stephen Steiner says:

    Hi Caroline,

    Well it works for us! We tried it out multiple times with multiple variations to make sure that it would work before we posted it. Also, I believe Jing (see Connect > Questions and Answers) has successfully used it and found it worked well!

  30. Caroline says:


    I want to make silica aerogels for a chemistry project at Sherbrooke’s University, does anyone know if this recipe works? I don’t really mind about the final quality of the aerogels, but I just want to know if I’ll get something or not.


  31. Stephen Steiner says:

    Hi MAtt,

    The amount of alcohol listed here is a sufficient amount. The Lawrence Berkeley recipe originally called for the equivalent of 7.5 mL of ethanol, where we suggest 22 mL.

  32. Stephen Steiner says:

    Hi Athmuri,

    Depending on how large of a volume of chemicals you buy, they chemicals will run you $100-$200 but can potentially make a large volume of materials. Regarding costs of equipment, see Make > Getting Started and Make > Build a Supercritical Dryer for more information.

  33. Stephen Steiner says:

    Hi JJJJJ,

    Sorry, perhaps the calculator is not clear. The calculator helps you to calculate the amount (either volume or mass) of liquid solution desired. So if you prepare 100 mL of solution, you will get about 100 mL of gel and, after supercritically drying, the resulting aerogel would weigh about 4 g.

    Hope this helps!

  34. JJJJJ says:

    There seems to be a glitch or problem with the Excel calculator. The site list the density of the substance as 0.04 grams/cm^3. Since cm^3 is equivalent to mL, then the density is 0.04 g/mL. On the calculator, there are spots for both end volume and end mass for determining chemical amounts. So, for example, an end sample of 4 grams should have a volume of 100 mL, if this density is correct. However, putting 100 in for the end volume and 4 in for the end mass yield different chemical amounts, with the mass-determined amounts being about 21 times less than the volume-determined amounts. Is this some glitch, or am I just not using the calculator correctly? Thank you in advance for your reply.

  35. Athmuri Kalyan Ram says:

    im working on developing a silica aerogel. Could you tell me what would be the approximate cost of the equipment and chemicals required for making it?