# Why and how do aqueous silica systems precipitate and can its occurrence be mitigated by quantifying its speciation correctly?

Background/Context:

For context, I have been doing a lot of work on experimental systems in geopolymer chemistry lately with a focus on "activator" solutions which are essentially mixtures of sodium hydroxide, sodium silicate and water used to enhance the reactivity of aluminosilicate materials upon mixing all the components together (simplified explanation). When making these "activator" solutions I usually change the $$\frac{Si}{Na}$$ ratios around in the hope of optimising the reaction with the aluminosilicate material which I've been measuring through other means experimentally.

The Ingredients I use are the following:

• Commercial Sodium Hydroxide (Aqueous):

• 49.9 wt% $$NaOH$$;

• 50.1 wt% $$H_2O$$;

• Commercial Sodium Silicate (Aqueous):

• 30.5 wt% $$SiO_2$$;

• 16.7 wt% $$NaOH$$;

• 52.8 wt% $$H_2O$$;

• "Extra" (Distilled) Water (Liquid):

• 100% wt% $$H_2O$$;

Most of the time, the activator solution I make is clear and good to go. Due to the exothermicity of $$NaOH$$, I usually leave the solution to cool down for easier handling and so that I can be sure everything has mixed well and reacted together to a steady state. Here is a picture of what the solution looks like in this case with an electric overhead stirrer constant at 200rpm. Note also the beaker in the background which is another activator solution which has stirred a bit more and become clear and "good to go" made with the following feedstocks:

• 97.94 g Commercial Sodium Hydroxide
• 171.63 g Commercial Sodium Silicate
• 68.3 g Extra Water

This corresponds to the following mols:

• 0.87 mols $$Si$$
• 1.94 mols $$Na$$
• 11.55 mols $$H_2O$$

However, I've noticed that sometimes when I make it at specified $$\frac{Si}{Na}$$ ratios I seem to get precipitation occurring after I've made the solution and allowed it to cool down as seen in the picture below. This is NOT good as it becomes hard to physically handle (like a viscous gel) and limits reaction with the aluminosilicate source; it essentially becomes useless to me and my experiments. This precipitated mixture is made with the following feedstocks:

• 97.29 g Commercial Sodium Hydroxide
• 171.67 g Commercial Sodium Silicate
• 59.47 g Extra Water

This corresponds to the following mols:

• 0.87 mols $$Si$$
• 1.93 mols $$Na$$
• 11.04 mols $$H_2O$$

I've done some reading and consulted some colleagues and I still do not have a clear answer of what's going on and why it's precipitating. I've had a few ideas as to why this may be the case including the following which I explore in the rest of the question:

1. Atmospheric $$CO_2$$ Interaction.
2. Soluble Silica Speciation.

Atmospheric $$CO_2$$ interaction?:

I first wondered if it was $$CO_2$$ getting in and causing precipitation, so I tried two identical solutions (which I knew had previously precipitated) and exposed only one to atmosphere and they both precipitated at (almost) exactly the same time (the capped one took a bit longer to cool down). So I figured that $$CO_2$$ ingress is not a severely detrimental factor.

Speciation Curves:

Moving on and thinking it may be related to pH, I've been focusing on chemical speciation in the hope to work this problem out. To start, I reviewed some literature first here and there to see if there were existing silica speciation curves. I then made these speciation curves based on the literature as seen below:

I figured that the less stable silica species, like $$Si(OH)\circ O_{3(aq)}^{3-}$$ are more reactive due to their unstable charge and thus likely to precipitate and used a pH meter to find that the solution was extremely alkaline (i.e. $$pH \geq 14$$). Note that the above speciation curves were made with the following equilibria calculations seen below in Equations 1-5 which I back calculated from the mentioned literature assuming the basis of aqueous dissolution of amorphous silica (i.e. $$SiO_{2(s)} + 2H_2O_{(l)} \rightleftharpoons Si(OH)_{4(aq)}$$):

Where, in general, the transitioning of silica species can be defined by Equations 6 below:

Each of the reactions above had both their product (with subscripts $$y_1$$ and $$y_2$$) and reactant (with subscripts $$x_1$$ and $$x_2$$ for silica and a for protic species quantified by linking with their equilibrium values as per the Equation 7 below:

I then worked on calculating the idealised carbonate system speciation to make certain I understood the maths and chemistry fundamentals. When applying the same logic to the silica system with the above equations and equilibria constants I got these curves:

These curves are entirely different to what I've seen in literature. I'm not sure what I'm doing wrong here or what other things I could try to quantify the system better so that I can mitigate precipitation occurring with the silica system. I thought understanding the pH of everything would be a good place to start, hence why I'm here.

---UPDATE 24/11/22---

I've been trying to understand this better from the POV of the following silicate solubility ternary plot. I've noticed that some of my solutions precipitate outside of precipitation zones (i.e. in the middle of zone 3 and not zone 2), I believe this is because of the different ways of manufacturing sodium silicate in industry. Unfortunately I have not found a link to the plot without a paywall :/ .

• Please clarify: 1) The picture(s) show an electric stirrer which is not on a hotplate. Is this the correct item? 2) What sodium silicate (liquid? solid?) are you using, and what are the amounts of water, NaOH and the sodium silicate in your pictured mixture? 3) Is there any aluminum in the mixtures shown? Aug 24, 2022 at 18:33
• Apologies. I have edited the question with more information. (1), it is an electric stirrer no hotplate. (2) Liquid, refer to edited parts. (3) No aluminium at all :) Aug 25, 2022 at 8:02
• The diagrams do nor consider silicate condensation, especially at acidic pH and or higher concentration. I would expect rather prediminance diagrams with regions of dominant silicate form. Aug 25, 2022 at 12:29
• Aug 25, 2022 at 13:53
• Aug 25, 2022 at 15:37

A lot of data is available online from commercial suppliers. PQ offers a handbook (free: Ref 1) which may explain some of the effects that you are observing.

Since you know the composition of the solution you are starting with, you can locate your mixture on Figure 3:

The ratio numbers are SiO2/Na2O by weight, and are for commercial liquids.

Another figure (#5, below) shows a very interesting viscosity minimum for certain concentrations and SiO2/Na2O ratios. While the data brochure does not address temperature as a variable, or precipitation, because that is undesired in a commercial product, you could very well assume that a sharp upward trend of the viscosity suggests a boundary line between two different behaviors, perhaps between clear and fluid versus much less fluid (gel-like) and perhaps also not so clear. Throw in the temperature variable, and you may be able to imagine what is happening to your mixture as you cool it.

The variability of your results indicates that you may be very close to a boundary line, or a metastable boundary region, so stirring or scratching or cooling might "precipitate" a phase change.

• Thanks for the great reply! My solutions both are around the SiO2/Na2O [wt%/wt%] value of 0.9. My supplier does not have original viscosity measurements. Regardless, I'll have to go about doing these viscosity measurements myself on the solutions I've created to check whether this is the answer which explains the precipitation. It'll be hard to see the ball drop amongst the white cloudiness so I may have to do the height difference of the whole beaker by listening to the ball drop to the bottom. Any other suggestions on this? Aug 26, 2022 at 3:29
• Viscosity suggestion: start with a tared glass rod; immerse it a certain distance into the mixture; pull it out and let it drain for 15 sec +/-; weigh it. The weight increase is a measure of viscosity; this can be done to compare viscosities. A similar method uses a cup/beaker: fill it, empty it, drain it, weigh it. The increase (stuck to the inside of the container) is a measure of viscosity. You need to make sure there are no special effects (like, don't use a non-stick teflon beaker). Aug 26, 2022 at 13:31
• The viscosity has value only because it can be compared to some other viscosity. So, if the falling ball method will not work to give a time in seconds because you cannot see the ball, take a (clear) mixture you can analyze that way, get a time in seconds, then test it by a weighing test on a rod or a cup. If you make another mixture with double the time to fall, the weight test will also increase (maybe not exactly double, but you can establish a correlation). The viscosity of these mixtures is complex. You should be able to get +/- 10 - 20% estimate of real viscosity. But not on gels. Aug 27, 2022 at 2:39
• I wonder if equilibration is slow at room temperature so that metastability is common. This would suggest that the method of preparation (e.g., time at temperature, rather than temperature alone, or even the rate of cooling) could determine the final state (clear or turbid). And mixing. There are so many ways to do the experiment that you will have to visualize what's happening from a molecule's eye view! Nov 24, 2022 at 15:37
• A fixed (relatively short) time period could give repeatable results, but follow at least one experiment out till no further change occurs, to insure that you are selecting 1) a useful time (i.e., remains clear) and 2) not very sensitive to variations (i.e., if you are at 80-90% of cool equilibrium, any further variations should be small). This allows you to do many experiments, but if they must remain clear for-ever, you must run the chosen recipe out for a long time (maybe not for-ever), for proof of stability. Learn how to identify the knife-edge! Nov 25, 2022 at 16:04