Difference between a "cartoon" flame test and a "real" flame test? How do chemists do flame tests correctly?

Question

This well-received answer to Why does the same electron transition release photons of different frequencies for some elements? begins with the following intriguing passage:

I am glad that you updated the question because it highlights a very common misconception. First of all the JavaLab Flame Test is completely wrong for both copper, calcium and many more salts especially the Group II and transition metal salts. Consequently, the NIST reference does not apply (see below). Secondly, it is a cruelty to science if a cartoon flame test is being shown instead of the real flame test. Search Youtube for real flame tests.

Rather than each of us search YouTube for it, I think it would be great if a short explanation of what chemists would consider a "real flame test" were captured forever in a Stack Exchange answer. So to that end I am asking:

Question: What is the difference between a "cartoon" flame test and a "real" flame test? How do chemists do flame tests correctly?

The next paragraph mentions a "pocket spectrometer" that one can actually make from a DVD (this is fun, I've tried it on various neon lights but never a flame) and there are discussions of spectral databases in the question and answer(s) there. Are these necessary ingredients of what a chemist would consider a "real" test?

Answer 1

Your interest is greatly appreciated. As a first step, let's clarify that flame tests as an analytical tool are obsolete. No professional chemist will use them to identify a given unknown sample. However, flame absorption and emission spectroscopy continue to be used routinely today at a more sophisticated level. There were so many elements discovered through flame emission that its demonstration is still worthwhile today. As for the cartoon flame test, I was simply referring to the Java app that depicted a fictitious flame test. It had some inaccurate atomic emission depictions as well as misleading electronic transitions "backed up" by the NIST database.

Flame tests of Group I, Group II, and some other selected elements should be performed using a platinum wire that has been dipped clean several times in HCl soln. Introduce the wire into the flame and heat it until no color appears in the flame. We all know that sodium is everywhere so this process ensures that the Pt wire is clean. Moreover, it is ensured that we use an oxidizing flame, that is, the flame should be very light blue to near-invisible with a clear blue cone inside. This is adjusted by checking the air flow to the Bunsen burner. Then one would dip the clean Pt wire in a concentrated solution of a chloride of Group I, II and some selected transition metals. Chlorides salts decompose and volatilize easily in a relatively low temperature flame of a Bunsen burner (methane and air) and immediately color the flame.

There are different levels of flame tests:

(a) At an elementary level, the flame is viewed with the naked eye and the flame color is noted.

(b) Next level is the use of cobalt blue glass to block sodium emission and sometimes iron contamination

(c) The third and perhaps the most effective level is to use a prism/grating spectroscope to watch the spectrum on a calibrated scale. However in that case, there must be a continuous supply of the sample to the flame. There are various methods to achieve that.