Titration Calibration or Certification?

A. Yeah... that isn't the easiest to answer directly is it?

What I have done here when the question has arisen is to make sure that my supplier (presumably a lab) can show me some traceable or verifiable certificates for my lab chemicals (presumably they have to certify the concentration - that is all I mean).

Then, make a known sample using lab grade chemicals and titrate it a few times to discover how accurate your readings are and to make sure that they are centered on the correct level.

You may have to use more than one operator doing the titrations and do some statistical analysis of the results (look up R&R studies - Repeatability and Reproduceability Studies and the gauge controls that were, in the past, required by QS9000) before your customer is satisfied with your answer, but that is the basic method.

Good luck!

A. Hi Robin,
I have heard (but not seen) auto titration dispensing units for various titrants. This does not include determining end point colors, turbidity, etc. You may want to talk to the customer to let them know you don't have this type of equipment. They probably have a QA person that may not understand laboratory apparatus, and methods of titrations. Some equipment in the lab can be calibrated, and some can't. If a flask or graduated cylinder says 100 mls, we trust that is what it is.
If they insist on documentation of titration calculations and bath additions, maybe that would satisfy them. Some customers demand that the standards be supplied by a laboratory supply house, instead of your lab tech or chemist making them up. That can get expensive as you know. Good luck with the customer.

A. There is an ISO specification that many companies are gradually going to. I think that the number is 15020.
Normally, if you are using burettes, you will uses a calibrated balance to weigh a specified percentage of burette dispensed distilled water. A 50 ml burette might have 5 - 10 samples taken. Pipette [pipettes on eBay or Amazon [affil links] are full quantity and in this case, you would pay for class A glassware. DI water is also weighed.
Another kicker is you normally have to do a 2 point calibration of the balance before and after use.

A. One common method is the known addition, whereby you make a standard addition of the component being measured to the sample after the titration is performed and then further titrate to the new endpoint. For example, when looking to measure % HCl, you would titrate the sample to the endpoint, make a known addition of HCl, then titrate further to the new endpoint. You can then calculate the % recovery of the known addition as a QC indicator of the titration.

Another method is to titrate a known standard concentration of the analyte in question. You can then adjust the calculations of the titration to account for variations in the titrant solution.

A. I've been in your situation before and have a few ideas that may be useful. First, have you tried convincing your customer that this process doesn't require calibration? If your glassware is like mine it has a qualification marked on it; mine has TD 20C, meaning it's set To Deliver the marked volume at a temperature of 20C. As long as your pipettes and burettes are not chipped or broken they are designed to maintain their volumes and don't need to be calibrated. The titration method is based on a chemical reaction that hasn't changed in recorded chemical history. Therefore there isn't anything that can fall out of calibration, so the process doesn't require calibration.

Not all customers will be swayed by the above argument. If the customer is adamant about getting calibration data for the titration process then you can go about verifying the calibration of your glassware. At room temperature one milliliter of distilled water weighs 1.00 grams. Fill your burette or pipette with distilled water and discharge a measured amount onto a tared container on a scale (assuming, of course, that the scale has been calibrated). Record the weight. Do this at least 10 times for every piece of glassware. You now have enough data to calculate a standard deviation or percent error or whatever you need to make your customer happy.

If that isn't good enough then you can prepare some plating solutions at known nickel or tin concentrations and titrate those. Once again you should repeat the test at least 10 times and develop whatever statistical data your customer requires. You may need to prove that your chemicals are certified in their concentrations. A certificate of compliance or some other form of written verification from the chemistry supplier should cover that requirement.

Good luck with your customer relations. I hope that was helpful.

A. A Certificate of Analysis (C of A) for the standard EDTA you use might do for documentation.

Or, you could purchase standard solutions of nickel and tin from a supply house. 10 g/l solutions, (or 1.33 oz/gal) are available for use as ICP standards. Then, you could document how your observed values match up with the true, and generate control charts, if you like.

Or, you could have a known plating solution prepared by some outside vendor, and use that the same way.

The common theme is documentation.

A. Robin
A chemist would not have to ask. This is a serious point. If you are processing parts for a critical application like pacemaker circuits, you have to take the control of bath chemistry seriously. Electroplating is a high tech business and chemical control is another. Neither can be picked up from books or the internet. I think you should take advice on bath control from your supplier and/or a competent analytical chemist.
The direct answer to your question is to buy in certified titration standards. No chemist would find it cost effective to standardise his own. It takes time and skill to get accurate results.

A. Robin,

One good way is to make lab solutions that mimic your process baths as close as possible. But make sure that you know the exact metals concentrations in your lab solutions.

Then titrate these lab solutions and see what your accuracy and reproducibility is. One big benefit of this approach is that it will help account for interferences.