Ask A Scientist Chemistry Archive

name         Linda K.
status       student
age          20s

Question -   We are studying atomic emission/absorption spectra in my
Intro to Chemistry class.  I was wondering if there are any periodic
trends to the spectra?  They seem to be pretty random.
--------------------------------------------------------------------------
Complicated, Yes! Random, No! There are numerous trends in the spectra of
atomic elements of the periodic table -- and their ions, which are also
important in the spectra of stars and galaxies. The definitive web reference
for atomic levels and spectra is the site maintained by the National
Institute of Science and Technology (NIST) which is:
http://physics.nist.gov/cgi-bin/AtData/spass.ksh?@/PhysRefData/AtData/html/ contents.html
    It would be easy to become overwhelmed by this data source, but here is
how you might find out some info.
Name the chemical element by symbol, e.g. H, Li, Cl then a space then an
"upper case" I for the neutral atom, "upper case" II for the singly charged
ion. You can select the energy units you want, eV, cm^-1 etc. and the number
of levels, or the actual spectra -- probably more about "elephants" than you
care to know. The trends are "hidden" because to a first approximation an
electron can "jump" from any level to most any other level. That is not
strictly true, but just a first approximation. What becomes apparent even in
a single electron atom like H or He+(1), or Li+(2) is that the levels, and
hence the spectral lines are almost always several closely spaced
multiplets, for well understood, but rather involved reasons.
    An alternative entry into the spectra of atoms is an introductory text
in "Inorganic Chemistry". I wish that space permitted going into more
detail, but the spectra of atoms, though simpler than molecules, are still
pretty complicated.

Vince Calder
=====================================================
Linda-
     Some kinds of spectra have some hint of periodic trends, but the repeating periods 
	 don't match the periodic table.
All the electrons piled on top of each other in successive shells influence each others' 
energy levels, making a bit of a tangled, shifting jungle.
Anybody who can write a computer program that does a good job of predicting atomic spectra 
just from atomic number,
would probably be quite welcome at some higher universities.

For single deep excitations there is some trend.
Look at the chart of X-ray absorption edges, or X-ray fluorescence, or Auger electron 
emission energies sometime.
The several lines march upwards monotonically vs. atomic number, but the periodicity 
doesn't match the periodic table.
Carbon and Silicon and Germainium have little in common that way.

The one place you might see a periodic-table trend is in valence-1 metals:  Lithium, 
Sodium, Potassium, Rubidium, and maybe Cesium.
Because they tend to act like an inert core with simply one electron around them, and 
only that electron excites.
But I haven't yet looked to see if I find noticeable resemblances or trends there.

Jim Swenson          
=====================================================