Is there a property that defines how much water a hygroscopic salt can absorb?

[seanspotatobusiness]

Is there some property that I can look up which would tell me how much water a hygroscopic salt can absorb (per unit mass of salt; for example anhydrous lithium chloride) before it's saturated and won't absorb any more?

 

[jim mcnamara]

I think you want the equilibrium moisture content. But is not a constant, it depends on environmental conditions like temperature. So to compare compounds you want EMC data for a given temperature, for example.
This is not what you seem to have described - exactly.
https://en.wikipedia.org/wiki/Equilibrium_moisture_content

What are you trying to do? Your ultimate goal, not your methods.

 

[seanspotatobusiness]

Thanks for responding.

My goal is to find out the size of container I need for 30 g of several salts I wish to use for calibrating some humidity sensors with the assumption that I fail to close them tightly enough and they continue to absorb water from the air. I don't want them to be able to overfill their containers and make a mess. Of course I could just use a ridiculous size of container but they're much more expensive and I need several and I would like to be able to use a modest-sized plastic food container for doing the calibrating inside so the smaller, the better. It would also be nice to know what volume of air could have the moisture removed it. I calculated earlier that 100 L of air at 100% RH only contain 4 g of water so it should be possible to reuse the salt many times provided I can keep an air-tight seal on the bottle.

The salts I'm interested in are: LiCl, Mg2Cl.6H2O, K2CO3, NaBr, NaCl, KCl and K2SO4 but I think that the first three or four are more likely to be a problem because they're more hygroscopic than the others (I put them in order of decreasing hygroscopicity).

 

[jim mcnamara]

AFAIK you can calibrate humidity sensors for one correct RH value and then you are good to go from there.
This post used table salt and a plastic bag to check for a known RH from wet NaCl (constant 75%RH) in an open jar lid. May be a lot cheaper to do:

https://www.stevejenkins.com/blog/2014/06/how-to-calibrate-a-hygrometer-humidity-sensor-using-the-salt-test

Interesting kitchen table science.
PS: you can ovendry and then reuse your salts in the list, or in the above example, table salt.

 

[seanspotatobusiness]

I want to do something more like https://www.allaboutcircuits.com/projects/how-to-check-and-calibrate-a-humidity-sensor/ where they do linear regression in a spreadsheet to get equations which correct the readings from their humidity sensors.

I know you can heat the lithium chloride to dry it but I think I will lose some transferring it between containers. I plan to keep the salts in polypropelene since that's compatible with all seven chemicals but cannot be heated to the temperature required. I was thinking I would try to dry it under vacuum instead but I might just have to accept the loses when the time comes.

 

[jim mcnamara]

Good choice. Let's ping @TeethWhitener and @Borek -- they may have a pchem approach your problem.

 

[TeethWhitener]

I’m a bit skeptical about the link in post #5. I haven’t read it in detail, but I noticed that they fit 4 points to a cubic equation. This is like fitting 2 points to a line. You will always get a perfect fit, but that in itself is meaningless.

A quick google search gives the accuracy of (properly calibrated) humidity sensors at somewhere between 0.5% and 5%, so I’m not convinced that a slightly leaky seal will be such a big deal, unless you’re leaving the apparatus in ambient for hours to days. Especially if you repeat each measurement several times and calibrate to the average.

 

[seanspotatobusiness]

I'm intent on using seven points instead of four but they do say that four is the minimum to obtain a cubic equation. It will be interesting to see how the values I obtain if I use four values compare to using seven. Maybe using only four will still give an equation that enables accurate correction for the other three.

I'm not concerned about a leaky seal during the calibration but in storage between calibrations. I'm interested in seeing how much the sensors deviate from their calibration after a year, two years, etc. I still need to buy the small polypropylene bottles though and I'm trying to decide whether I need 30 or 60 ml bottles. I'm assuming they won't pull in more than their own weight in water?

I'm also wondering whether it's really necessary to add water to the lithium chloride before use - the instructions say that it is, but wouldn't that only be useful if the ambient RH was less than 11%? I can guarantee that it won't be so it just seems like a waste of my salt (and/or time and effort dehydrating it).

 

[TeethWhitener]

Ok I’m extremely confused as to what you’re trying to do. The point of the lithium chloride for calibration is that the vapor pressure of water over a saturated solution of lithium chloride is ~11% that of the vapor pressure of pure water. The vapor pressure of water over a dry bed of lithium chloride is whatever the heck the humidity happens to be.

 

[seanspotatobusiness]

You're saying that you have to add water to lithium chloride to get it start absorbing more water from the air?

I thought that if you put dry lithium chloride in an enclosed space, it will absorb water from the air, bringing down the RH to 11.3%.

 

[Borek]

I thought that if you put dry lithium chloride in an enclosed space, it will absorb water from the air, bringing down the RH to 11.3%.

No, in such a case RH can get lower. It is saturated solution that you have to aim at.

 

[TeethWhitener]

Again, what you’re saying is overall very confusing to me. Do you want to use dry lithium chloride as a desiccant? Or do you want to use an LiCl solution as a calibration standard for a humidity sensor?

As for 7 points vs 4, I’m not sure what kind of accuracy you’re going for, and what accuracy your humidity sensor is rated for. At any rate, I doubt you’ll need to worry about any nonlinearity in the response of your sensor. It’s easier just to fit your points to a line.

As for whether the calibrations will deviate after a few years, yes probably. You’ll likely need to recalibrate periodically to get accurate readings.