[Golden_4_Life]

I am having major problems with the "%w/w determination, ZnO in Ointment" which involves sple digestion in acid +alcohol, then rendering the sol'n basic via NH3Cl/OH, and ultimately indication with Eriochom B T as sol'n transits from purple to very briefly lavendar and finally to the aqua-blue endpoint upon stochiometric equivalence with free Zinc (aq).

The problem is:  that my high level spikes sample have a high quant bias of 10% to 15% inaccuracy; but the RSD's on triplicate are all <1%. In contrast my low level spikes have no bias inaccuracy and the RSD's on the triplicate reps are all <1%.  Why are my high end spikes giving inaccuracy whereas my low end spikes are exceptionally accurate?

My brain is wrapped around this problem so much tha I can no longer think straight!

[marquis]

Suggestion.

The United States Pharmacopoeia (USP) has a monograph on zinc oxide.  The titration is different from the one you mention. However, it has been validated and is reliable. I would recommend investigating the USP method and see if you have the chemicals to do the titration.

I don't know the composition of your ointment.  If it is mostly a grease, the hydrocarbon portion could be ashed off and leave the zinc oxide.  The hydrocarbon would leave little residue.  Then titrate per the USP method.

[Golden_4_Life]

Thanky Marquise for that info.  I had previously perused the USP and saw that they used ''ashing' process which was a bit out of my range of capability as we have many organic solvents and no muffle furnace.  But the approach is similar chemically, and I appreciate your advice on how the interference from hydrocarbons would be alleviated after combustion.  I will aim to implement it and compare with my observations.  Thanky Marquise

[marquis]

Be careful with the ashing.  The first time, you might want to try charring on a bunsen burner and then putting in the muffle furnace.  Using a small amount of sample is also a good idea.

I never had any problems with ashing the zinc oxide mixtures.  But when you are using oils, the possibility exists.

Good luck.

[Golden_4_Life]

Unfortunately our lab doesn't have access to a muffle furnace; so I will try ashing via crucibe x bunsen flame. It'll likely be quite smoky.  Yet, I wondered if even after conducting this more elaborate test, that my basic problem of the high end bias in my samples will sustain?  I mean, it is bafffling that only the upper level spikes are showing inaccuracy.  And, I believe that the EDTA titration method is linear throughout the high-end range that I'm using.  Yet,l I will be sure to work under the fume hood.

[marquis]

When you heat on the bunsen burner, heat gently.  This will turn the organic portion of the zinc oxide blend brown and then black.  Work the heat up on the bunsen burner slowly.  It will take time, but you should be able to avoid the smoke.

I can't explain the high RSD.  I suspect the method, but can't see any obvious errors.  That's the reason the USP method was suggested.  We did assays on relatively pure zinc oxide (95% pure). We didn't see that kind of variability.

[Golden_4_Life]

OK, so your results do prove that the assay is linear up through ~95%w/w ZnO.  My highest standard was prepared as 6% w/w zinc oxide added to other compounds to make the ointment mixture.  Today I was just informed by the prep line manufacturer that they added the incorrect amount of zinc oxide. That explains why my values were biased on the upper end.  I knew it was not the assay---since the other spike levels (less conce'd ones) were quanting OK.

Thanky, Marq!

[Golden_4_Life]

Question:  In doing a titration, I have some general questions about the "Indicator" solution.
1) about how many drops do you customarily add in?  Our SOP says    5 drops.  But is there a general rule--just add enough to get the solution to change color?

2

[marquis]

One of the titration errors is called "indicator error".  It takes a small, but measurable amount, of titrant to shift the color of the indicator.  The general rule is to use the minimum amount of indicator needed to get a clear end point.

Generally, the amount of titrant needed is very small.  You can adjust for this by running a blank (same reagents as normal, just no sample) and adjusting the titration volume based on the blank. 

Had a weird thought. Your original method used EDTA titrations.  These methods need to be done in an alkaline environment (say pH 10 for aqueous solutions). Had a similar problem to yours with EDTA. When using a relatively large amount of sample (a calcium salt, in this case), the calcium salt shifted the solution to the acid side.  The titration values got funny.  Adding a small amount of base before the titration solved the problem.   

[Golden_4_Life]

I am using ammonium chloride + ammonium hydroxide mixture reagent , pH ~10.25 as a means of making the pre-titration solution alkaline. Also, I hoped that this reagent could also act as an 'anti precipitating' measure to keep Zinc in  solution.
Good observation though Marquis!

?Is five drops too much or too many of indicator?

[marquis]

It depends on the concentration of the indicator.  Usually, five drops is within the normal range.

[Quaff]

I am having major problems with the "%w/w determination, ZnO in Ointment" which involves sple digestion in acid +alcohol, then rendering the sol'n basic via NH3Cl/OH, and ultimately indication with Eriochom B T as sol'n transits from purple to very briefly lavendar and finally to the aqua-blue endpoint upon stochiometric equivalence with free Zinc (aq).

The problem is:  that my high level spikes sample have a high quant bias of 10% to 15% inaccuracy; but the RSD's on triplicate are all <1%. In contrast my low level spikes have no bias inaccuracy and the RSD's on the triplicate reps are all <1%.  Why are my high end spikes giving inaccuracy whereas my low end spikes are exceptionally accurate?

My brain is wrapped around this problem so much tha I can no longer think straight!

I'd be inclined to put the sample in a kjeldahl with some sulfuric and heat it to digest the organics and then do a dilution and run AA or ICP.  In the alternative, the Zn could be precipitated and a gravimetric method used.  Depending on other materials present it could also be possible to dissolve a gram or two of the sample in 30 ml of amyl alcohol, hit it with an excess of an EDTA salt, add about 100 ml of IPA and back titrate with aqueous mag chloride.

[Golden_4_Life]

Back-titration is a good deductive suggestion, would I use the same indicator or a different one?

[Potatomuffin]

Use Methyl Orange.  The pH change should be sharp so determine the endpoint based on the immediate change to pumpkin orange.

[Borek]

Use Methyl Orange.  The pH change should be sharp so determine the endpoint based on the immediate change to pumpkin orange.

pH changes pH during EDTA titration? No idea what you refer to.