pH meter slowly climbs when testing pH of tap water

[afr0byte]

So, I was just taking the pH of my tap water (I took it with the aerator out, if that matters.). The pH started at about 6.5 and slowly, over the course of about 5 minutes, climbed to ~7.12 where it seemed to more or less stabilize. That makes sense, since that's what Ward Labs had said my water pH was. Does anyone have any idea why the pH would slowly rise like this? My only guess is that there's some CO2 disolved in the water that's slowly coming out of solution. I suppose there could be something wrong with my probe, but I'd calibrated it right before I took the reading, and I didn't have to adjust it much.

This is my water profile from Ward Labs, if it helps:

87 Ca
9 Mg
5 Na
4 Cl
21 Sulfate
214 alk as CaCO3
261 bicarb

pH 7.1

 

[ajdelange]

As we discussed in the current slaked lime thread, carbo distributes itself according to the pH. At pH 7.1 16.1% of it is H2CO3. With an alkalinity of 214 the total carbo is 5.276 mmol/L resulting in 48.6 mg/L H2CO3. That's more than the water can hold i.e. it is supersaturated with H2CO3 and carbon dioxide will leave the water over time (a couple of days) and the pH will rise until the water is no longer super saturated. It will probably finish up around 8.

Now that's just part of the story. pH meter electrodes are not exactly rapid response devices. When you move an electrode from one medium to another whatever it was in last adheres to the bulb and junction and needs to be rinsed away with sample. If you are doing things right the previous medium was a squirt of DI water and you should have blotted it away but that is still a different medium from the sample and the comment still applies. It definitely speeds things up to swish the electrode through the sample to accomplish this removal of the previous stuff but that alone doesn't insure quick response. It takes time for the glass of the pH bulb and the reference junction to come to equilibrium with the solution. Over time the bulb, if used in brewing, will become coated with proteins which should be removed from time to time with an enzymatic cleaner. If this isn't done the response of the electrode will slow. The junction frit can be replaced in some meters and this is a good idea too. But eventually the electrode just runs out of steam (slope degrades, response gets too slow) and it has to be replaced. Manufacturers make a big deal out of trying to speed response by things like clever junction design (I have one that renews the junction at every measurement), glass formulations and so on. Even so they take time.

Electrodes seem to be subject to hysteresis or to at least have a mind of their own. Sometimes the response is quick and some times it is slow. You must make the decision as to when to take the reading. Some modern meters take that responsibility from you and have an algorithm that freezes the reading if it doesn't change by more than x pH in y seconds. Some let you set x and y or one or the other. You must be very careful with this. In the lab I have the meter send it's readings to the computer where I plot them and print them to the screen as they come in. I can, thus, graphically see the change in pH over time. If the computer asks the meter for a reading when the meter thinks the pH reading is unstable it marks it with an *. I have screen fulls of data without * but in which the plot shows pH slowly approaching an asymptote. As noted above with your water I wouldn't expect that asymptote to be reached in less than about 3 days.

 

[afr0byte]

As we discussed in the current slaked lime thread, carbo distributes itself according to the pH. At pH 7.1 16.1% of it is H2CO3. With an alkalinity of 214 the total carbo is 5.276 mmol/L resulting in 48.6 mg/L H2CO3. That's more than the water can hold i.e. it is supersaturated with H2CO3 and carbon dioxide will leave the water over time (a couple of days) and the pH will rise until the water is no longer super saturated. It will probably finish up around 8.

Now that's just part of the story. pH meter electrodes are not exactly rapid response devices. When you move an electrode from one medium to another whatever it was in last adheres to the bulb and junction and needs to be rinsed away with sample. If you are doing things right the previous medium was a squirt of DI water and you should have blotted it away but that is still a different medium from the sample and the comment still applies. It definitely speeds things up to swish the electrode through the sample to accomplish this removal of the previous stuff but that alone doesn't insure quick response. It takes time for the glass of the pH bulb and the reference junction to come to equilibrium with the solution. Over time the bulb, if used in brewing, will become coated with proteins which should be removed from time to time with an enzymatic cleaner. If this isn't done the response of the electrode will slow. The junction frit can be replaced in some meters and this is a good idea too. But eventually the electrode just runs out of steam (slope degrades, response gets too slow) and it has to be replaced. Manufacturers make a big deal out of trying to speed response by things like clever junction design (I have one that renews the junction at every measurement), glass formulations and so on. Even so they take time.

Electrodes seem to be subject to hysteresis or to at least have a mind of their own. Sometimes the response is quick and some times it is slow. You must make the decision as to when to take the reading. Some modern meters take that responsibility from you and have an algorithm that freezes the reading if it doesn't change by more than x pH in y seconds. Some let you set x and y or one or the other. You must be very careful with this. In the lab I have the meter send it's readings to the computer where I plot them and print them to the screen as they come in. I can, thus, graphically see the change in pH over time. If the computer asks the meter for a reading when the meter thinks the pH is stable it marks it with an *. I have screen fulls of data without * but in which the plot shows pH slowly approaching an asymptote. As noted above with your water I wouldn't expect that asymptote to be reached in less than about 3 days.

Wow, thanks for all of the info. I suppose I'll use my cleaning solution. I do pour some distilled water over the probe and try to wick away excess between each reading. The probe seems to jump close to 4 quickly, when I calibrate with the 4 solution, and close to 7.01 when I calibrate with the 7.01 solution. Presumably the H2CO2 would come out of solution faster if I heated the water to 77F? My tap water is about 50-55F.

 

[mabrungard]

Also don't forget that the water is at about 4 atmosphere's of pressure in the distribution system piping and it probably had excess CO2 if it came from a groundwater source. This means that the water would start 'off-gassing' as soon as it exited the tap. There have been times when I would draw a glass of water and the water became cloudy with tiny bubbles. After a minute or so, the bubbles make their way to the surface and the water became clear again. Given that reduction in carbonic acid from the CO2 escape, the pH would be rising in that situation. I never measured it though.

You can see a similar phenomena as you heat your water for brewing. I find that around 180F, bubbles form on the side of the kettle, showing that the CO2 and other gases are being driven from solution. The pH rises slightly in that case too.