Sciencemadness Discussion Board

water "ionizers" pH anomaly

chemrox - 21-8-2012 at 20:41

I tested two so called water "ionizers" with pH paper using three different types that indicated colorimetrically in the ranges 0-11, 0-13, 0-10. They all indicated the water that was supposed to be 9.5 pH to be ~5-6. OK so far. But then after checking my pH papers against buffer solutions, I calibrated my pH meter between 7.01 and 10.20 and checked the ionized water with it. It indicated pH value of 9.14 for water that was supposed to be 9.5 and tested 5-6 with paper. I've never run into this and don't know what it could mean.

blogfast25 - 22-8-2012 at 04:31

It means very little, because pH is a logarithmic scale: pH = - log [H3O+].

This means that at pH 9.14, [H3O+] = 10<sup>-9.14</sup> = 7.2 10<sup>-10</sup> (mol/L) and for pH = 9.5, [H3O+] = 10<sup>-9.5</sup> = 3.2 10<sup>-10</sup> (mol/L). It's a negligible difference. Your ioniser probably works as intended.

If you really want to be that precise, check your pH meter once more, e.g. against Cream of Tartar (very OTC). For paper, 9 is a bit of a difficult area: precise at best to +/- 1 pH point.



[Edited on 22-8-2012 by blogfast25]

[Edited on 22-8-2012 by blogfast25]

chemrox - 22-8-2012 at 11:18

In the world of water chemistry the difference between ph 5 and ph 9 is a pretty big difference. Much bigger for example than municipal NPDES discharge permits allow. As to checking with cream of tarter why bother? I've got ph standards in my lab. The colorimetric methods agree with them. The anomaly is this:

The colorimetric reagents work with the pH buffers but don't read the water from the "ionizer" (not mine!!) The pH meter reads the "ionizer" as putting out as advertized. I haven't run into anything like this before. We use the paper on streams all the time. In fact we've got a set of indicators in the range of 5-8 for stream work. Sadly, we're seeing lots of lower pH values recently. (I don't believe in ionized water, wetter water or structured water by the way. Also I don't own the "ionizer.")

blogfast25 - 23-8-2012 at 05:08

Quote: Originally posted by chemrox  
In the world of water chemistry the difference between ph 5 and ph 9 is a pretty big difference.


I didn't say otherwise.

Clearly something is the 'ionised' water is messing with the indicators in the paper. Form here it's almost impossible to guess what it could be.

'Ionisers' are about as useful as 'karma crystals'.

watson.fawkes - 23-8-2012 at 07:27

Quote: Originally posted by blogfast25  
Clearly something is the 'ionised' water is messing with the indicators in the paper. Form here it's almost impossible to guess what it could be.

'Ionisers' are about as useful as 'karma crystals'.
Whether these devices be useful or not, it seems that chemrox may have uncovered a real physical effect rhere. Whether that effect was the one intended or not is more or less irrelevant to the scientific question about what's actually occurring. Just because the people promoting such devices are quacks doesn't mean they do exactly nothing. A disparity between two methods of pH testing is evidence that something may be happening. A pH meter senses the electrical characteristics of a solution; pH paper senses reactivity. It's certainly conceivable that there's sometimes a difference between them.

So I did one internet search and a tiny amount of reading on what these devices do. "Separating the positive and negative ions" and that kind of claptrap. There's only one thing that occurs to me that's plausible, and it's easy to test. Perhaps there's some effect that's metastable, where the relaxation time to the equilibrium state is being extended. I see two ways to test this.

The first is to use time. Take ten contemporaneous samples and run the pH tests against them in a time series, say, daily. It's possible, even likely, that the pH paper starts behaving as expected at some point. I'd guess any effect doesn't last too long, say, on the time scale of hours. This is a low-effort test, requiring just a bunch of vials and a bit of attention.

The second is to use heat. Think of this as a way of accelerating time. Bring a sample up to some fixed temperature, cool it off to a reference temperature, and test. If there's something that's reaction-rate dependent, the Arrhenius equation would indicate that heat would bring it to equilibrium state sooner.

unionised - 23-8-2012 at 11:21

The buffering capacity of any water that's clean enough to drink will be very low and so it will be difficult to measure. Ordinary pH paper or pH meters just won't do the job.

What you have discovered is more likely to be a limit of your measurement techniques than anything more important.

blogfast25 - 23-8-2012 at 12:13

Watson:

Unless Chemrox’s paper measurements are erroneous in some way, something is happening but I didn’t dispute that. It’s just that I haven’t got the foggiest idea of what is happening.

The only way to make progress is to at least understand well what the ioniser precisely does to the water (and not what is claimed is being done to the water).

watson.fawkes - 24-8-2012 at 05:25

Quote: Originally posted by unionised  
The buffering capacity of any water that's clean enough to drink will be very low and so it will be difficult to measure. Ordinary pH paper or pH meters just won't do the job.

What you have discovered is more likely to be a limit of your measurement techniques than anything more important.
What are the systematic errors possible in each of these measurements? Measuring very pure water with electricity has some problems with low conductivity, but with an electrical pH measurement of around 9 it seems not so likely that this is the issue. Assuming for the moment that this is the more accurate method, what is pH paper measuring that causes the disparity?

This page from a manufacturer of pH instruments has a suggestion to add high-purity KCl to the solution to increase the ionic strength of the test solution. It seems like another simple test to run to see if the effect is preserved with an alteration that shouldn't alter measured pH.
Quote: Originally posted by blogfast25  
The only way to make progress is to at least understand well what the ioniser precisely does to the water (and not what is claimed is being done to the water).
It's one way, but not the only way. If existing theory covers what's happening, then it will be fruitful. If it doesn't, then theory a dead end and new experiment is needed.

This is an interesting case to me. Assuming basic experimental soundness, there's either some new phenomenon or some little-known way that a measurement fails to be accurate. Neither of the two measurement techniques measure pH directly, but intermediated through some other physical effect. If the intermediary effect isn't behaving as expected, that's interesting to know, as it's significant to metrology and calibration, a field I have a particular interest in. If it's a new phenomenon, I assume that's of general interest.

unionised - 24-8-2012 at 09:06

Have a look at this sort of thing
http://www.thermo.com/eThermo/CMA/PDFs/Articles/articlesFile...

And you may want to think for a while what the effect of dilution would be on the pH of a solution of universal indicator (currently buffered at 5.0).

To a good approximation, it stays at 5.

hyfalcon - 24-8-2012 at 09:15

I've noticed a PH of about 5.5 when I checked my de-ionized water from my filter. I only test 9-12 ppm dissolved solids on this water so I'm sure that getting an accurate reading is impossible as it is.

gutter_ca - 24-8-2012 at 12:24

pH 5.5 -ish is normal with DI due to dissolved CO2 from the atmosphere.

DerAlte - 24-8-2012 at 21:24

Blogger:
Quote:
This means that at pH 9.14, [H3O+] = 10-9.14 = 7.2 10-10 (mol/L) and for pH = 9.5, [H3O+] = 10-9.5 = 3.2 10-10 (mol/L).

Unionised:
Quote:
The buffering capacity of any water that's clean enough to drink will be very low and so it will be difficult to measure. Ordinary pH paper or pH meters just won't do the job. What you have discovered is more likely to be a limit of your measurement techniques than anything more important.

Watson.Fawkes:
Quote:
This page from a manufacturer of pH instruments has a suggestion to add high-purity KCl to the solution to increase the ionic strength of the test solution. It seems like another simple test to run to see if the effect is preserved with an alteration that shouldn't alter measured pH.


I think the answer is latent in the above posts.

Consider the quantity of indicator needed to react to produce the required color difference. Let’s use Blogger’s numbers calculated for us above. Assume a liter of test water. This gives us [H3O+] = 10-9.14 = 7.2* E-10 mols available for reaction. The only suitable indicator I could find (CRC) was Thymol Blue, FW 466, used on it’s second indication at pH 8.9-9.2, transition Yellow - - > Blue.

For the reaction H+ + In- < -- > HIn (In = indicator here) the amount of indicator that can be changed in color is the quantity above. In actual mass this is then 466*7.2* E-10 g, ie about 3μg. Unless the amount on the paper is less than this, the color change would be masked by unreacted yellow. I have no idea how much there is on a paper, but it’s probably a lot more than this.

As for the pH meter,it all depends on the current needed to operated. I cannot be bothered to work out a relation here, but essentially the method would involve how much current the EMF sensing devices uses, the junction PD expected and the quantity of electrons in faradys required to charge the capacitors of the system and overcome leakage. All relative unknowns. All I can say is that the response might be slow but modern pH meters probably give the right number since leakage currents in the sensing amps are way down in the picoamperes. I would trust it rather than the papers, under the conditions of about a liter volume.

This might explain why pH paper cannot give reliable results. Try Watkin.fawkes’s suggestion; and also try leaving the indicator paper in a stream from the ‘ioniser’ (whatever that is) for a long time. It should not then suffer from the above problem as adequate ions become available.

Regards
Der Alte

unionised - 25-8-2012 at 04:29

pH paper won't work because it's more strongly buffered than the material you are seeking to measure.
pH electrodes won't work because they leach buffer solution through the liquid junction.
So you are comparing two methods that won't work.

Do you expect consistent results from that?

blogfast25 - 25-8-2012 at 05:05

Quote: Originally posted by unionised  
pH paper won't work because it's more strongly buffered than the material you are seeking to measure.
pH electrodes won't work because they leach buffer solution through the liquid junction.
So you are comparing two methods that won't work.

Do you expect consistent results from that?


Strange (even though I understand what you're referring to). I've a two point calibration pen-type pH meter which works really quite well and readings for tap water and DIW are always close to 7. Conductivity would probably be a much better measure of 'purity', though.

Adding the indicator to the water should work though, no? A Methyl Orange indicator solution is typically about 0.003 M. Now you add 1 drop (0.05 ml) to, say, 20 ml. So now you have 0.0000075 M MO (- log C = pC = 5.1). Hmmm... that's still in the order of magnitude of [H3O+], [OH-].

Somebody's opened a can of worms here...;)

[Edited on 25-8-2012 by blogfast25]

unionised - 25-8-2012 at 08:16

"I've a two point calibration pen-type pH meter which works really quite well and readings for tap water and DIW are always close to 7. "
Is there any evidence that those readings are actually correct?
Pure water rapidly dissolves a little CO2 and becomes slightly acid so, unless t's in a recirculating system, the pH of DIW should be near 5.

watson.fawkes - 25-8-2012 at 08:16

Another suggestion, in the spirit of divide and conquer. Honestly, this should have been the first thing that occurred to me.

Test the water going into the device with both methods and see if you have the same inconsistency in measurement. If the input water is showing the same results, it would be pretty clear that the measurement techniques available to you are insufficient to tell if the devices is doing anything.