Re: [Phys-l] experiments with dyes (was: happy equinox)

[John Denker <jsd@av8n.com>]

On 03/23/2008 10:13 PM, Krishna Chowdary wrote:

> I'm interested both in straightforward demonstrations and experiments
> as well as the nontrivial ones and ones that get at misconceptions.
> I'm pretty sure that I have some (maybe lots) of these misconceptions,
> and I'd like to learn more.

A proper discussion of that could take days or weeks.  There
are other folks on the list who know much more about this 
than I do.

Here's a rough outline of ideas that might be worth discussing.  
I'm just throwing ideas out there, without (yet) discussing any 
of them in detail.

Here's a start:  Question:

         What color is the stuff in a bottle of 
         ordinary yellow food coloring?

That sounds like a silly question.  (Indeed it is similar to a 
proverbially silly question:  What was the color of Napoleon's
white horse?)

Hint:  The stuff in a bottle of yellow food coloring is not yellow.

I've had chemistry teachers argue with me about this, saying
that you're supposed to mix the stuff with water, and when
you do that, an "obvious" chemical reaction occurs, making
it turn yellow.  Well, that's not right either.  

All this is easy to explain in terms of basic physics.
  http://www.av8n.com/imaging/dye-spectra.htm
(Thanks to BC for providing the spectra.)

Core take-home message:

  a) Physics class usually analyzes the mixing of light sources.  
  This is a relatively simple phenomenon, and the simple linear 
  theory works pretty well as a starting point.

  b) In contrast, mixing dyes is very complicated.  Any usable
  theory is is nonlinear from the get-go.  In grade school you
  were probably taught a simple paint-mixing theory (using 
  red/yellow/blue primaries).  It's still being taught.  Alas
  it's just wrong.  Experiments don't even remotely match the 
  theory.  I complained about this in kindergarten, and in first 
  grade, and in second grade, and in third grade, et cetera, and 
  never got anywhere.

Professional artists know about this.  They have a lot of ad-hoc
knowledge, but usually not a lot of theory.  They don't spend a
lot of time looking at spectrophotometer traces.

The guys who make printing presses (Heidelberger and such) know
a lot about this, but they tend to keep the details confidential.

Ink-jet printer inks are readily available and probably not too
dangerous to handle (though not as safe as food coloring).

Many artists' supplies, and many of the dyes that are used in 
industry and/or in the biology lab are horrifically toxic, so 
don't get overly adventurous.

You can get all kinds of weird color-shifts if you illuminate
a spectrally-sharp pigment with a spectrally-sharp light source.
I don't offhand know a good dye system for demonstrating this,
but it should be fairly straightforward if you make a white 
illuminant by mixing red, green, and blue laser diodes (and 
compare that with plain broadband white illumination).

With a cheap IR LED and phototransistor you can easily demonstrate 
that blue food coloring is transparent in the IR.

If you have access to a spectrophotometer, you can run the 
percent-transmission curves at various concentrations, and
easily see the nonlinearity.  Before long you will understand
why the underlying physics is expressed in terms of cross 
sections, not % transmission.

Practically every natural dye you can think of (red cabbage,
blueberries, tea, etc.) works as a pH indicator.  
  http://chemistry.about.com/cs/acidsandbases/a/aa060703a.htm
You can do a survey.  This makes a decent science-fair project, 
but lately it's been done to death, so you won't get any points 
for originality ... unless you come up with some novel twist.

Plants typically contain a mixture of various natural pigment
molecules.  They can be separated using paper chromatography.
Many references suggest this as a "home school" or kitchen-table
experiment ... but beware that it is much harder than it sounds.

In olden days, dyeing cloth was a cottage industry.  Amazingly
complex and laborious.  http://users.frii.com/sos/

Plastic cuvettes are cheap (about 20 cents apiece in bulk) and 
are ideal for shining light through a sample of liquid.  Made 
for the purpose.  Nice flat sides.  It's worth scrounging a few.
You can stack cuvettes side by side, so that the light goes 
through them in series.  Compared to mixing the liquids, this 
makes it a lot easier to try many different arrangements.

The office-supply store sells specially-treated acetate suitable 
for printing with inkjet printers.  (Or use a dye-sublimation 
printer on not-so-fancy acetate.)  This is a cheap source of 
"gels".  You can stack them to demonstrate color-mixing.
(And/or borrow some professional-grade gels from the theater
department.)