DIY Alk test discussion thread

[Randy Holmes-Farley]

Used a Pyrex 1 cup measuring cup. Filled it to the 1 cup line as close as I could.

I think a Pyrex measuring cup could easily be off by 10%.

 

[lavoisier]

Thanks for your article Randy. Not only are test kits expensive but I can now attest to inaccurate as well. I ran my test kit and it read Alk of 11. I then ran your procedure twice and found the Alk was 9! At my current Alk level the false reading is not critical but if I got into a situation where my Alk dropped below 7 the test kit would be telling me it was still above 7...not good! You are saving me money again (2 Part articles) and improving my reef keeping. It doesn't get much better.:dance:

 

[Randy Holmes-Farley]

Glad you found it helpful. :)

Happy Reefing!

 

[okeydokey]

Please ignore my bad English

My recipe is similar to yours

KH Formulation
A reagent
0.04g Bromocresol Green
0.008g Methyl Red
90g DI-Water
1.6g 0.1M NaOH

B reagent
10.12g 1N HCl standard solution
428.7g DI-Water

Procedure
1. Add 4ml of sea water in the vial.
2. add 4 drops of A reagent
3. Use syringe
4. Read the KH table

Reference
Ind. Eng. Chem. Anal. Ed., 1941, 13 (7), p 466–470

 

[Randy Holmes-Farley]

Thanks for the info, okeydokey, and welcome to Reef2Reef! :)

I couldn't see past the first page of the paper without paying for it. What pH does that ratio of dyes give for an endpoint?

Why not just use a pH meter?

 

[okeydokey]

The endpoint of pH is approximately 4.6 when using a mixed indicator.

It would be more accurate to use a pH meter, but more convenient to use mixed dyes.

And accurate enough for me.

 

[Randy Holmes-Farley]

OK, sounds good and thanks for posting the procedure!

Happy Reefing. :)

 

[JimWelsh]

Inspired by this thread, I've been doing my Alkalinity tests lately using a Pinpoint pH meter with a BRS "lab grade" probe (about 1 year old now) that I'm calibrating with 7.0 and 4.0 buffers each time I use it. My "syringe" is a 50ml Class A burette. I'm using a stir plate to "swirl" the sample. I'm using 0.0022N HCl. My sample size is 10ml, delivered by a volumetric pipette. This stoichiometry means that I can get three tests out of the 50ml of reagent in the burette, so I'm doing the tests in triplicate. The precision I'm able to achieve is pretty remarkable! The standard deviation of the three readings is consistently less than 0.25%, and averages the theoretical limit of the equipment, which is 0.15%, giving a confidence interval of 0.3%, which means that I can be 95% certain that my Alk is within 0.025 dKh, or 0.009 meq/L of the average of my three measurements.

The accuracy of the method is assured by standardization of the 0.0022N HCl reagent using a Na2CO3 primary standard.

I'm sold on this as a method.

 

[Randy Holmes-Farley]

Inspired by this thread, I've been doing my Alkalinity tests lately using a Pinpoint pH meter with a BRS "lab grade" probe (about 1 year old now) that I'm calibrating with 7.0 and 4.0 buffers each time I use it. My "syringe" is a 50ml Class A burette. I'm using a stir plate to "swirl" the sample. I'm using 0.0022N HCl. My sample size is 10ml, delivered by a volumetric pipette. This stoichiometry means that I can get three tests out of the 50ml of reagent in the burette, so I'm doing the tests in triplicate. The precision I'm able to achieve is pretty remarkable! The standard deviation of the three readings is consistently less than 0.25%, and averages the theoretical limit of the equipment, which is 0.15%, giving a confidence interval of 0.3%, which means that I can be 95% certain that my Alk is within 0.025 dKh, or 0.009 meq/L of the average of my three measurements.

The accuracy of the method is assured by standardization of the 0.0022N HCl reagent using a Na2CO3 primary standard.

I'm sold on this as a method.

Cool.

Have to tested any kits in parallel to see how they perform?

 

[JimWelsh]

Cool.

Have to tested any kits in parallel to see how they perform?

Only the Salifert kit, which I find to be quite accurate compared to the DIY method when the proper endpoint color is identified (more of a lavender color; if you get to an orange-pink then you've overshot), although much less precise, due to the more crude equipment involved.

 

[Randy Holmes-Farley]

Thanks, Jim. :)

 

[JimWelsh]

Adding still more DIY to this test method: I have found that I can use DIY pH calibration buffers made out of Borax and Cream of Tartar to calibrate my Pinpoint pH meter!

The use of Borax as a pH calibration solution has been discussed before in aquarium hobby articles, such as this: http://reefkeeping.com/issues/2005-02/rhf/index.htm, which refers to this: http://web.archive.org/web/20021015005420/www.animalnetwork.com/fish2/aqfm/2000/feb/bio/default.asp. Dr. Bingman's calculations suggest using 1/2 tsp. of Borax in 1 pint of water. My measurements suggest something more like 1 1/4 tsp. in 1 quart of water (I probably didn't compress the Borax as much). As Dr. Bingman says, though, the pH of the resulting solution isn't very dependent upon exact measurements. This solution does need to be protected from atmospheric CO2, so it should be kept in a closed container, hopefully with minimal headspace. It will have a pH of 9.23 at 20C, and 9.18 at 25C.

Another easy and reliable DIY pH calibration solution can be created using potassium hydrogen tartarate, aka Cream of Tartar. Both it and Borax are discussed on page 11 of this document: http://pac.iupac.org/publications/pac/pdf/1978/pdf/5011x1485.pdf. The Cream of Tartar calibration solution is discussed in depth in this document: http://nvlpubs.nist.gov/nistpubs/jres/47/jresv47n6p433_A1b.pdf. A saturated solution at 25C requires about 6.4 grams per liter, which by my measurements is about 1 1/2 tsp. of the powder. The nice thing about this pH calibration solution is that you just need to make a saturated solution. In other words, you don't really need to measure at all -- just add enough to whatever amount of RO/DI water to make sure that it won't all dissolve. Once you are certain that you have a saturated solution, let the remaining crystals settle out, and then pour the clear solution off for use, leaving the undissolved crystals behind. This solution will have a pH of 3.56 at 20C, and 3.55 at 25C.

Now, I know that the instructions for the Pinpoint say that you need to use 7.0 and either 4.0 or 10.0 pH buffers to calibrate it, but in fact, the two calibration screws are just adjusting the bias (7.0 screw) and the slope (4.0/10.0 screw) of the calibration curve. I have found that if I use the Borax solution to adjust the 7.0 screw to the correct value for that buffer, e.g., 9.18 at 25C, and then use the Cream of Tartar solution to adjust the 4.0/10.0 screw to the correct value, e.g., 3.55 at 25C, then the Pinpoint pH monitor can be calibrated reliably using these solutions. I did find that adjusting the one screw tends to throw the other one out of calibration, and they tend to fight each other if you fully adjust each screw to the target value each time. I can more easily and quickly get perfect calibration of both buffers if I only adjust each screw only enough to make up about half the difference each time. For example, I start by setting the 7.0 screw to exactly 9.18 using the Borax buffer. Let's say that I then find the Cream of Tartar solution reads 3.65 instead of 3.55. Instead of turning the 4.0/10.0 screw until the meter reads 3.55, I halve the difference, and adjust it until it reads 3.60. If I switch back to the Borax buffer, I may find that it now reads something like 9.22, so, again, I halve the difference and turn the 7.0 screw until the meter reads 9.20. After just a few back-and-forth cycles like this, I can quickly get both solutions reading correctly.

I have verified the calibration using other pH calibration buffers with values of 4.0, 4.5, 4.65, and 7.0, and in all cases have found agreement to within +/- 0.03 pH units, which is certainly good enough for the DIY alk test. Obviously, these solutions can be used to calibrate other pH meters than just the Pinpoint, as long as the meter allows you to enter/set buffer values, rather than being programmed to use certain specific buffer values.

 

[JimWelsh]

Bumping this ^^^ just because I find it difficult to believe that nobody else finds the idea of easy, inexpensive, reliable DIY pH calibration buffers interesting.

 

[Randy Holmes-Farley]

IMO, these are very nice DIY buffers. Thanks for developing them for us. :)

I suspect that not as many people will want to try to calibrate a meter the way you did the Pinpoint. The two adjustment screws impacting each other is why, I think, pH 7 is nominally selected as one of the points of calibration by many meters. But even if they want to use a pH 7 buffer and then a DIY, the cost is half as much as buying both.

FWIW, I use a meter that allows many buffers for calibration, and I often use 7, 8, 9 and 10. :)

 

[Dnahacker]

And what about the effect of ambient temperature on the result? To do this titration right you would need to use a PH meter with a temperature compensation probe, so with probes you are looking at £500-1000 with probes and calibration standards... You can buy a lot of salifert kits for that...

 

[JimWelsh]

And what about the effect of ambient temperature on the result? To do this titration right you would need to use a PH meter with a temperature compensation probe, so with probes you are looking at £500-1000 with probes and calibration standards... You can buy a lot of salifert kits for that...

You are overstating your case in the extreme. The effect of ambient temperature at the endpoint in the range where virtually everybody in the hobby will be doing this test, i.e., 20-25C, is negligible, (+/- 0.02 pH units at most). I just gave recipes for calibration standards that are so inexpensive as to be virtually free. A $100 Pinpoint pH meter is perfectly adequate for this test.

 

[Randy Holmes-Farley]

And what about the effect of ambient temperature on the result? To do this titration right you would need to use a PH meter with a temperature compensation probe, so with probes you are looking at £500-1000 with probes and calibration standards... You can buy a lot of salifert kits for that...

Are you referring to Jim's pH calibration standards, or the DIY alkalinity test?

In any case, wouldn't you also be worried about the effect of temperature on the dyes in a kit? They will certainly have a temperature dependence on the pKa just as inorganic ions are known to do.

For the DIY alk test, the curve of alkalinity vs pH is fairly flat near the endpoint, and small differences in calibration result in small differences in alkalinity (as I show below)

Here's the curve:

So if you get anywhere between pH 4.0 and 4.5 for the pH measurement of the endpoint, the range of possible alkalinity only ranges from about 3.60-3.75 meq/L (10.1 - 10.5 dKH).

How big of an error could the temp make?

Well, we need to make some assumption about the room temperature. Let's say it is between 20 and 30 deg C, with calibration at a standard 25 deg C.

The Nearnst slope at 25 deg C is 59.16 mV/pH unit
At 20 deg C it is 58.17 mV/pH unit.
At 30 deg C it is 60.15 mV/pH unit.

So for pH 4 we are 3 pH units from the zero point (pH 7) of temperature effects, so the potential error is:

20 (58.17-59.16)*3 = -2.97 mV
25 none
30 (60.15 -59.16)*3 = 2.97 mV

At 58.17 mV/pH units, the error of 3 mV gives an error of 0.05 pH units.

In terms of the predicted alkalinity for that variation, the error in total alkalinity will be quite small. Probably less than +/- 0.05 dKH based on an error of 0.4 dKH when the pH uncertainty is 0.5 pH units (which I showed above).

 

[JimWelsh]

Your assumptions are more conservative than mine, Randy. I was assuming a tighter temperature range for "room temperature", and also an endpoint of 4.5 instead of 4.0. Using your formulae above, but for my assumptions, the error turns out to be 0.021 pH units. Not that I had determined the "0.02 pH units" I claimed above theoretically. My statement above is based on experience and observation.

Temperature-induced error can also be addressed, of course, by a quick calibration immediately prior to doing the test. I do so, and it almost always amounts to just a simple, small bias adjustment (small tweak of the 7.0 screw).

 

[Randy Holmes-Farley]

Your assumptions are more conservative than mine, Randy. I was assuming a tighter temperature range for "room temperature", and also an endpoint of 4.5 instead of 4.0. Using your formulae above, but for my assumptions, the error turns out to be 0.021 pH units. Not that I had determined the "0.02 pH units" I claimed above theoretically. My statement above is based on experience and observation.

Temperature-induced error can also be addressed, of course, by a quick calibration immediately prior to doing the test. I do so, and it almost always amounts to just a simple, small bias adjustment (small tweak of the 7.0 screw).

Well your intuition was highly accurate. :)

 

[JimWelsh]

I have started another thread that is at least partially relevant to this thread: Alkalinity titration endpoint pH & Salifert endpoint color