« first day (22 days earlier)      last day (163 days later) » 

7:06 PM
@mercio, yes thats about half way to the "case-by-case checking"
So the talk went excellent. All of the "significant" persons found it intrigiung
 
that's some pretty good news
 
They encouraged us to put it to Science, at least to try.
One noticed about the issue of the relative orientation of the O(3) representation
We all agreed that there is a "canonical orientation"
most of all, the tables have been used for these purposes more than 100 years in similar ways
 
and I still don't know why there aren't more of those tables when you pick some other representations into O(2) x O(1)
it's like they are classifying something I haven't pinpointed yet
 
is it correct that any x,y,z permutation of these O(3) reps are isomorphic
?
I guess so
 
yeah you have a notion of isomorphism of representations
namely any change of basis on the representing space
 
7:12 PM
So that means you could make an alternative table where we have the same for Roty
as we have now for Rotz
 
or Rot(x+2y-7z)
 
yeah
but chemically that makes in way no sense
so maybe point groups are still more
than a group and a rep in O(3)
so maybe an additional orientation comes also into play
but I don't see that clearly
 
I want to say there is a page for each isomorphism classes of faithful representation into O(3)
but I don't think that's true :*)
 
mhm
again from the chemistry it would make no sense
 
for example for C24
 
7:16 PM
all these E-type irreps in the higher order C_n
yeah :-)
its clear that the only can refer to z
 
the 24 cyclic group appears twice
on the C24 page and on the S24 page
 
the corresponding representations (Tr=p) are E1+A and E1+B
but they are not considering all the other faithful representations
like E2+A or E3+B
 
I mean those are quite different groups
its like C_s and C_2
 
no, they are isomorphic as groups
 
7:18 PM
I think they ARE C_24 x C_s and C_24 x C_2
Yes so are C_2 and C_s
 
like Cs and C2 and Ci are all the cyclic group of order 2
so they are all isomorphic as groups
 
yes
and I guess these groups are prducts with those
 
so I am wondering why the cyclic group of order 24 only appears twice and not 22 times
 
but maybe C_12 x C_s
maybe the only really isomorphic ones are exactly C_2 and C_s and C_i
and the rest are that are isomorphic are just products of others with them
so for each C_n there are three isomorphic ones
(roughly speaking)
one guy had a very similar idea as I had
 
oh wait some of them are nonfaithful
 
7:22 PM
btw we have only 3 types of generators
1. C_n
 
E5, E7, E10 still should work instead of E1
I think it is because there are outer automorphisms
of the cyclic 24-group
 
what means "outer" here
?
 
automorphisms that are not conjugation by an element of the group
 
so you can think that there are more faithful representations by looking at the tables
but they turn out to be isomorphic to the A+E1 and B+E1 ones anyway
I guess each has their own change of basis
that
when combined with the outer automorphism
okay
E5,E7 and E11*
 
7:26 PM
maybe
 
(and it is probably not a coincidence that those are the integers coprime with 12)
 
i see what you mean
I saw that structure as well
once ago but couldnt express it
One of the guys who probably understood it it best said he found it remarkable that there is this correspondence with the z-component of the angular momentum and mostly that it is conserved under descend of symmetry. While the latter is thank god not that hard, I did'nt confess that we still are working on the former.
I mean on a proper proof
but as far as I see no one of those people would really care, but I do now, since I like to fully understand that.
And some said it fully possible that there is still more to discover.
No one does such kind of considerations nowadays anymore apparently :-)
But maybe its not really more than O(3) + highest order element is along z
 
maybe it's not really more than the representation is reducible
but since there are only finitely many irreducible cases there are probably many equivalent complicated way of stating the distinction
 
What I mean is that its not only O(3) but a physically meaningful O(3).
and QM can be built of position and mometum
algebraically
but position and momentum relate automatically to symmetry of the physical "objects"
When a molecule has a certain symmetry and I say I want to measure the angular momentum of the electrons in that, I will have to decide which component I want to measure.
And that decision can only be made relative to the orientation of the molecule.
 
can't you just rotate your measuring apparatus
:*)
 
7:42 PM
but now the momentum and the angular momentum "live" in the space of the O(3) reps already
I think thats the point somehow
:-)
 
they are operators that transform a certain way under change of coordinates
 
exactly
 
I am not done reading my quantum field theory book
:(
I didn't read today
 
is it also "field"
that qiute tough then
and far above me ... :(
 
it is spooky and magical
 
7:44 PM
I just know some basic concepts but never did anything with that
 
I have seen multiple times people saying how the calculations had no firm mathematical background so it got me curious
 
"renormalization"
 
that kind of thing
 
you get diverging "integrals" = inner products
 
reminds me of that place where he integrated exp(i(p.x))
over all of space
with some 3 vector p
 
7:45 PM
and to get rid of that you divide by another one
 
and then he innocently said "we add an 1/N normalisation constant in front of the whole thing"
 
sounds like theoretical physics
they do that all the time ...
 
and I am still not sure if the author has acknowledged that the integral is far from converging at all
I need to reread a few parts :*)
 
maybe not, they are happy if they get some "formula" out of it
that works
like Euler
and Euler was rocket
 
haha
 
7:47 PM
but it was the same non sense in principle
he just could get around the troubles due to his extreme intuition
it took a while until Analysis was invented that put that stuff on firm grounds
I think Physicists wait for the same with their QFT
 
maybe
if they have some calculations that give the right result, it is good
 
then you just have to remove all of the calculations that give the wrong results
and show tha tthe good result always exist
 
lol
perturbation theory is used often in QFT without any care about convergence, so it is use in chemistry as well
and in deed often it oscillates wildly with the order and eventually diverges. But it still may be the PT2 is better than PT1 than its used
and PT3 is anyway too expensive ...
anyway after I return on Saturday I'll have a fresh look at the MS
 
PT3 is looking at third order terms ?
 
7:54 PM
third order perturbation theory
 
right
 
it contains terms like <a|O|i><i|O|b><b|O|j>/(ea-eb)(ei-ej)
or something
those are coded in the Feynman diagrams
 
3-cycles !
 
@mercio could be!
Oh yeah I forgot, the "guest star" of the meeting was talking about general magnetic interactions with non small magnetic fields
as they occure on neutron stars
you cannot treat these perturbatively but you just solve the full Hamilton equation
however one of his bottom lines was
"paramagnetism is a magnetic Jahn-Teller effect"
Which suggests another intimate connection between JT and paramagntism
it led us both to speculate if there is maybe some more general underlying theory unfying our concepts ...
but thats very hard to bring under one concept at least at the moment I have no idea about that
you see we have a "structural" Jahn Teller effect
giving rise to a certain magnetic response.
And he has a Jahn-Teller effect in the magnetism itself ..
that is induced by magnetic interactions between states
which is pretty wired ...
but maybe
maybe
I got an idea!
 
o..o'
 
8:08 PM
You see the magnetic coordinates are the "spin coordinates"
while we were treating $Q$ so far as our position coordinates
but of course, when going fully relativistic ...
then you need both of them!!
you should study your QM
especially the Dirac equation
I don't know if I ever told you
but the reason I came about this project was I was searching for SOC:=Spin-Orbit coupling effects
SOC is purley relativistic
in a way the "only" really qulatitatively new thing
when going relativistic in chemistry
in vague way
and I was searching for extreme SOC effects in closed shell systems
so I needed to understand particluar non-relativistic things first
namely the JT systems
the ones where I expected the largest SOC effects
that turned out to be true (as some test calculations show)
SO now I have almost understood the non-relativistic case
(this is our current work here)
and then my next step would be the relativistic case
until 5 minutes ago I had no idea what to expect here
but I guess now I have
its what Trygve saw
(the big guy Trygve Helgaker)
@mercio ?
 
it's getting late
 
sure! I also need to sleep, lets continue later!
 
neevr heard of that name
 
theoretical chemist, one the best world wide though I guess hes not that popular in the US but in Europe he is
among Chemists
 

« first day (22 days earlier)      last day (163 days later) »