[PFScience]

Is there a way to predict (without reference to another test) what colour a metal ion will display in a flame test?

For example - if I wanted to predict what colour Metal X would display, could I use energy gap informaton and equations to get the wavelength of light emitted and thus the colour?

Can anyone do the maths and show why, for example, Lithium burns with a red flame?

[Corribus]

In principle, you can determine (or look up) the unique emission lines for each element. But the actual wavelengths you observe will depend on a lot of things, including the temperature of the flame. Moreover, it's also hard to guess ahead of time how various lines will add up and translate into a single observed color - which isn't always easily determined from a set of light wavelengths. You might be able to use a color chart to help you do it, but... why? The flame test behavior of just about every metal is already known. Why not just look it up?

[PFScience]

In principle, you can determine (or look up) the unique emission lines for each element. But the actual wavelengths you observe will depend on a lot of things, including the temperature of the flame. Moreover, it's also hard to guess ahead of time how various lines will add up and translate into a single observed color - which isn't always easily determined from a set of light wavelengths. You might be able to use a color chart to help you do it, but... why? The flame test behavior of just about every metal is already known. Why not just look it up?

Fair comment.... that would be far easier! I think I was just trying (probably unnecessarily!) to delve deeper into the maths aspect of it.

Cheers