4
$\begingroup$

I am trying to figure out how the degerenate 5-fold $d$ orbitals is split in a $D_{3h}$ crystal field. A practical case would be a subtitutional $\rm Fe$ impurity in the hexagonal graphene lattice. What I am supposed to do is:

  1. find the character table of $D_{3h}$ group and the symmetry operations;

  2. calculate the character of these symmetry operations for the full rotational group; and

  3. use the decomposition formula to work out how many times each irreducible representation of group $D_{3h}$ is contained in the reducible representation of the full rotational group.

step (1): The first step is pretty easy to do

Character table for group <span class=$D_{3h}$" />

step (2): The character expressions of these symmetry operations of the full rotational group are as follows: $\chi^{(l)}(C_n)=sin[(l+1/2)\alpha]/sin[\alpha/2]$ for rotation $C_n$ with $\alpha=2\pi/n$
$\chi^{(l)}(\sigma_h)=\chi^{(l)}(\sigma_v)=\chi^{(l)}(C_2\bigotimes i)=(-1)^l\times\chi^{(l)}(C_2)$
$\chi^{(l)}(S_n)=\chi^{(l)}(C_{n/2}\bigotimes i)=(-1)^l\times\chi^{(l)}(C_{n/2})$ for improper rotation $S_n$
In the present case,
$\chi^{(2) \mathrm{reducible}}(E)=lim_{n\rightarrow\infty}\chi^{(2)}(C_n)=[l+1/2]\alpha/[\alpha/2]=2l+1=5$
$\chi^{(2) \mathrm{reducible}}(\sigma_h)=\chi^{(2) \mathrm{reducible}}(\sigma_v)=(-1)^{2}\times\chi^{(2)}(C_2)=sin[5/2\times\pi]/sin[\pi/2]=1$
$\chi^{(2) \mathrm{reducible}}(C_3)=sin[5/2\times2\pi/3]/sin[\pi/3]=sin[5\pi/3]/sin[\pi/3]=-1$
$\chi^{(2) \mathrm{reducible}}(S_3)=(-1)^2\times\chi^{(2)}(C_{3/2})=sin[5/2\times2\pi/(3/2)]/sin[1/2\times2\pi/(3/2)]=sin[10\pi/3]/sin[2\pi/3]=-1$
$\chi^{(2) \mathrm{reducible}}(C_2^\prime)=sin[5/2\times\pi]/sin[\pi/2]=1$

step (3): The decomposition formula for reducible representation:
$a_j=1/h\sum_k{N_k\chi^{\Gamma_j}(C_k)^*\chi^{\mathrm{reducible}}(C_k)}$
where $h=12$ is the order of group $D_{3h}$, $\Gamma_j$ the irreducible representation of $D_{3h}$, $C_k$ the class $k$ with $N_k$ elements.
By appying the decomposition formula to the irreducible representation $A^\prime_1$ of $D_{3h}$, one has
$a_{A^\prime_1}=1/12\times[\chi^{A^\prime_1}(E)^*\times\chi^{(2) \mathrm{reducible}}(E) + \chi^{A^\prime_1}(\sigma_h)^*\times\chi^{(2) \mathrm{reducible}}(\sigma_h)+ N_{C_3}\times\chi^{A^\prime_1}(C_3)^*\times\chi^{(2) \mathrm{reducible}}(C_3)+ N_{S_3}\times\chi^{A^\prime_1}(S_3)^*\times\chi^{(2) \mathrm{reducible}}(S_3) +N_{C_2^{\prime}}\times\chi^{A^\prime_1}(C_2^{\prime})^*\times\chi^{(2) \mathrm{reducible}}(C_2^{\prime})+ N_{\sigma_v}\times\chi^{A^\prime_1}(\sigma_v)^*\times\chi^{(2) \mathrm{reducible}}(\sigma_v)]=1/12\times[5+1-2-2+3+3]=2/3$
In principle, $a_{A^\prime_1}$ should be either zero or an positive integer. But I got a fractional number. I repeated the calculations several times but got the same result. I appreciate it if anyone can help me out with this puzzle.

$\endgroup$
5
  • $\begingroup$ Do you have a reference for the $\chi^{(l)}(O)$ equations? $\endgroup$
    – Hans Wurst
    Commented Nov 13 at 11:07
  • $\begingroup$ @HansWurst Chapter 5.2 (Characters for the full rotation group) of "Group Theory: Application to the Physics of Condensed Matter" by Mildred S. Dresselhaus (google.com.sg/books/edition/Group_Theory/…). $\endgroup$
    – bitsoal
    Commented Nov 13 at 12:27
  • $\begingroup$ I just derived the characters myself and I get $\chi(S_3)=1$ instead of -1. Using this character you get the expected $a_{A'_1}=1$. All other characters were the same, so I guess there is your error. $\endgroup$
    – Hans Wurst
    Commented Nov 13 at 13:03
  • $\begingroup$ @HansWurst, below is my detailed calculations. Could you please help to point out where I made the mistake? $\chi^{(2)}(S_3)=(-1)^2\times\chi^{(2)}(C_{3/2})=\chi^{(2)}(C_{3/2})$. $C_{3/2}$ corresponds to the rotation of $2\pi/(3/2)=4\pi/3$. Thus, $\chi^{(2)}(C_{3/2})=sin[5/2\times4\pi/3]/sin[1/2\times4\pi/3]=sin[10\pi/3]/sin[2\pi/3]=sin[-2\pi/3]/sin[2\pi/3]=-1$ $\endgroup$
    – bitsoal
    Commented Nov 13 at 13:21
  • $\begingroup$ See my answer. Your calculations were in principle all correct, except for mistaking the $S_3$ operation. Due to this you applied the wrong equation for the character. $\endgroup$
    – Hans Wurst
    Commented Nov 13 at 13:29

1 Answer 1

2
$\begingroup$

I think I spotted your error. You assumed that $S_3$ is a rotation and inversion, which is not the case in the $D_{3h}$ group. This leads to the wrong character for $\chi(S_3)$. The correct $S_3$ operation is a combination of the $C_3$ rotation and the $\sigma_h$ reflection, which leads to a character $\chi(S_3)=1$

$\endgroup$
1
  • $\begingroup$ Many thanks! You are right. For odd numbers, $S_n$ is a rotation $2\pi/n$ followed by the horizontal reflection $\sigma_h$. For even numbers, it is instead a rotation $4\pi/n$ followed by inversion. $\endgroup$
    – bitsoal
    Commented Nov 13 at 13:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.