CHEMICAL APPLICATIONS OF GROUP THEORY F. A, Cotton 3rd Edition, 1990 Answers to Problems and Exercises Copyright, 1990 F. A. CottonChapter 2 1. any similarity transform ve set vp, say A”/BA, may be rearranged because of- the commutative property of Abelian groups to A”‘AB which equals B. Im other words every possible similarity transform on every ment converts that element into itself. 2. I€ in a group of ofder h any element, X, has a cycle (1.e., X, X?...x = £) such that n momobmo omomaE mo aurea nowPeRon >pmownoza ep moonok gE |e a [a a |B ¢ Jc dp jo rF iF « [c¢ a ta omomeam> womarmas omomm> mo ourmmang womromon prmowmeZa mp wagman E A | aA BIB c |e dp [oD F iF on a ola i. ge |z a 8 ¢ DF G a a [a c £ B HU GD F Bp |B £ ¢ AG H F D ¢ c B A E F ODOR G p |p G H F EC AB re |F HG Dc EBA ¢ |G F DH BA E& CG He JH DF G A BCG E7 8. For G,°2) each element is in # class by itself. Subgroups are: (2.3), (B,D), (E,6) (B,A,B,C) For 6.) the classes are E; A,B; ; BAF; G,H. The subgroups are (Bc) (B,A,B,0), (E,A,D,F), (E,C,6,0) For 6,4) each element is in a class by itself because the group is Abelian (symetrical about the diagonal - all multiplications therefore commute). Subgroups of orders 2 and 4 are possible and several of each exist: (EA), (E,B), (E,6),... and (E,B) (E,A,B,C), (E,A,D,F), (E,B,D,G) and (E,C,D,H) 9. We may proceed as we did for the groups of order 8. Begin with the cyclic group, G,°2), vhose miltiplication table can be written straightforwardly: This group has two E’s on the diagonal. ‘The remaining possibilities to consider are those with four and six E's on the diagonal. We already have the multiplication table for G,°2), the one with four diagonal E's. We now turn to the task of working out che muleiplication table for the one with six, a task we shall find to be impossible. We begin with the following partial multiplication table, which incorporates the assumption:We may now assume that BA=C. Since the elements B, C, D and F are still arbitrary we can make this sumption arbitrarily. We could equally well have assumed BA = Dor F. All ve are saying {s that the product BA is not, B, Aor E. We then proceed by left-miltiplying or right-auleiplying both sides of this and subsequent equations and making use of the fact that each element is its own inverse. BAC BA=C BBA= BC BAA = Ca A= BC B= CA With these results included, the table is now: Now, the only way to complete row 2 is to accept that DA - F and FA ~ From these equalities we derive soue others: DA=F FA=0 DDA= DF FFA = FD Aw DF A= 9D In a similar way, the remaining places in column 3 aust be filled by BD - F and BF = D. Adding all these new results we have:We now find that it is impossible to go further without violating the rules of group theory. The product CD cannot be C, Dor E; it must therefore be A or F, But we cannot put A or F into the D row because both are already there. a 10. For group G,“!) che multiplication table is symmetrical about the diagonal and fs thus Abelian. Therefore, each element is in a cli a by itself, There is one subgroup of order 2, namely, (E,B). For group ¢,‘2) che table {s again symmetrical and each element is ina by itself, There are three subgroups, (B.A), (E,B), and (B,C). For Gy), the cyelic subgroup of order 6, as for all cyclic groups, fs Abelian and each element is ina class by itself. There are two subgroups based on the concept chat che group Ls the cycle of some generating element X, namely XK, x, XS, Kt, XS, xe Clearly, X* = £, hence (E,X*) is a subgroup, and X?X? = X¢ and X‘X? - x*, hence (E, X*, X*) form a subgroup. In the table as worked out in exercise 9, we would write these as (E,C) and (£,3,D).=10-- Chapter 3 Set A 3-1 A3.2 43.3 43.4 33.3 33.4 B35 33.6 33.7 33.8 83.9 83.10 B3.11 Sen (_?5 Pag (yi S5y (C5) (2) 55 (6) Dyy Ce) Dgy 4) Cy Ce) ® 5 (@) Sy Om) GC Cy Son 55 - see C5, 2 3 4 Sq - Sgr Sg’ (Cz)+ Sg+ Sg* (Cy)s Sg” Dag and 0, T and Ty a Tf 9 = 30°, 0,; Otherwise 0 Dyy The pentagons ere not regular P34 Dra 3. 8 (6,75 837. SQ° CEDSet ¢ -il- GG, Cy, G) Sy (4) 6, (5) Dp, (6) Dy, (7) Cy, (8) G, () Dy 0) Cy, AY Cy (12) Dy, (13) C, 4) C15) as Og by Gyr er Ops ds Coe er Cys fe Ky Cys Ly Dogs me Gye Set D GD Cy, (2) Cy, 3) Cy 4) Dog 8) Cy) Dy 0) Dg 1) Day Set E 254 o @ &y @) (3) De, (9) @) Deg (10) ) 06 ap (6) Dy 12) as) Coys Be Ci By Cys te Dope Jo Coys (3) Dy, 2) ¢, (6) Cy (7) Cy,cvs Chapter 4 1, (1) Two rotations about the same axis. Choose the z axis. A rotation by an angle @ or # {s given by the matrices cosa -sina cosB -sing sina cosa sing cosf ‘Their product, multiplied in the order given is (— + sinasing -cosasing - el sinacosf + cosasind -sinasing + eosacosp and when multiplied in teverse order is on + sinasing -sinacos# - cosasing cosasing + sinacosp -sirasin? + cosacoss Except for trivial rearrangement these are identical. (2) Reflections through planes 4 to each other. Take xz and yz planes 1 ofa d o pio y 9-1 o ly fo et xz, yz yz xz,13 (3) Inversion and any reflection or rotation. The inversion matrix is a unit matrix with a minus sign: f-1 0 0 0-1 0 0 0-1 A unit matrix commutes with gl] other matrices. (4) Two Cp rotations about | axes, Such rotations are described by @iagonal mactices. All pairs of diagonal aatrices commute. (3) Rotation and at reflection, Take rotation about z by a and reflection in the xy plane. The following matrices clearly commute. cosa -sina | {0 1 0 0 sing cosa}/0 0 1 0 ° o /f2 0 0 -1 2. The operations are £, Cy, 0,7, oY + as shown below.1a For the two one-dimensional representations, the characters themselves are the complete matrices, For a general point, 3-dimensional matrices will be generated, but all of them will be of the form: since no operation chenges the z coordinate. We may therefore consider only che x and y coordinates. For Cy, Cy? and 0‘! we can write down by inspection: [ev - sing costa sinda -1 0 Jae cone] fotze conte] | oa cy ey ® vhere @ - 2n/3, The simplest vay to get the matrices for o,(?) and o, 7) ia to use the multiplication properties of the group, t.e., 2°?) = co, and 0) = ¢420,(). when the corresponding matrices are multiplied ve get: ~cosa sing ~ees2a—sintal and sine cosa sinta costa gD oo? Note that these matrices have characters of 0. Of course, they aust since the one for o,‘7) has x= 0, All operations in the sane class are required to have the same characters. If all we want is the character for a class,1s we can just choose the easiest one to work with and forget the others. AlL the nece ry matrices are given in the text or are obvious. One simply has to recognize which are inverse to which (e.g., Cy and Cy? are inverses) and grind through all the similarity transforms. Note first that only one operation in each class needs to be used to get the character for that class, 4. The matrices for the various C,, C, and C, rotations as they affect « general point, x, y, z, which have been discussed in the text will give the representations and hence the characters for T, directly. For the more complex functions we proceed as follows: A C,(2) operation applied to the set of functions xz, yz, xy will give the set -yz, xz, -xy. To express this in a matrix equation, we clearly require: dD -1 0 xz “yz 1 0 0 ye} = xz o 0 xy 7"y ‘The chatacter is -1. For aC, operation (the one which permutes x, y. z to z, x. y) we can see that the set xz, yz, xy becomes yx(-xy), ax(-xz), yz. Again the pertinent matrix equation is written and the character (0) determined. o 0 1] | xz Loy 10 of tye, = | xe 0 1 of | xy ye For the two kinds of C, operations (1.e., C, = C,? and the class of six) we have:16 G,(z) applied to xz, yz, xy gives -xz, -yz, xy. The matrix will clearly be diagonal with elements -1, -1, 1 and x = For the C, axis that lies between the +x and +y axes in the xy plane the set xz, yz, xy becomes -yz, -xz, xy. The matrix equation is o -2 Of [xz “ye 10 0) dyz) = |-xz 0 0 1 |xy xy and the character is 1. 5. Because of the commutative property of an Abelian group, each operation ts in a cli by itself. Thus, for an Abelian group of order h there are h irreducible representations (rule 5). The only set of h integers whose squares add up to h? is a set of I's. 6. The group C, is a cyclic group. Its representations must consist of the ¢? where ¢ = exp(2xip/4) = cos2mp/4 + isin2rp/4. Since a cyclic (hence Abelian) group of order 4 gust have 4 one-dimensional representations, we have7 From the trigonometric functions, all «P with p = 4n are equal to 1 and those with p = 2n are equal to -1. Those with p = 4n + 1 and 4n + 3 will be { or -1, respectively. We thus obtail c? oo a 7. The approach here is to write down the matrices for Cy, Cy? and C,(x) by inspection, using a general point x, y, z. Obtain those for the other Cy operations by matrix multiplication. it will be found that these will factor into a set of two-dimensional representations that form the E representation, Since the other two representations are one-dimensional, their matrices consist entirely of #1 and can be written down by Inspection. Note. If you are really alert, you will observe that all three C, operations are in the same cla and therefore have the same characters. You can thus work only with the “obvious” matrices for Cy, Cy? and C,(x).18 8. For the two dihedral planes, the results are 1 0 0 o 10 0 0 -L ‘This matrix commutes with each of the others and thus ve need examine only two products, e.g., 1 0 of jo 1 0 ao 10 o 1 of 2 0 of = |2 0 Oo o 0 ‘This matrix must describe the tvo-fold rotation about an axis along the line of intersection of the two planes, and it can be seen that it does.19 Chapter 5 1, The characters of the direct products are: Pon EB 2Cy 3 Cy 3y’ 36," 1 25, 2%, oy oy 3, B Polo o’oaoa- toa. 1 - ‘ig * 7g Mees, FoF Pod obo po oyoaar rai 2 Bay E, 2-2 -lo2 0 0 - 2 1 2 @ au * Fag EXE, 4-1 1 - 0 0 4 1 -t 4 0 0 ag * Pau BgXBe Do 1 -4 0 9 4 2 1 4 0 oO am By and they reduce as follows: The important points to note here are: (1) when any representation is taken in dizect product with the vorally symmetric one, it goes into itself. (2) any one-dimensional representstion times itself is the totally symmetric representation. (3) the products of ux g or g x U must be u. (4) more generally, any direct product containing sn odd number of u's will be u; otherwise it will be g, 2. Since ¥*, belongs to A, or Ay, (for C,, oF Dag) we need only examine the various binary direct products for the representations to which the algebraic functions belong end the representations to which ¥*,, may belong,20 and see if they are or contain A or A,,. This is really a trivial task. From the character table one finds the representation to which the function belongs and then ¥1, must belong to the same one. Thus z requires A,, (x,y) require E, (x? + y?) requires A,, and so on in C,.. 3. Im Ty the character table shows that the set, (R,, R,, RZ) belongs to che T, representation. Therefore, the direct product of the representations of the upper and lover states for an allowed transition must be of contain the T, representation. Only those direct products of dimension 23 need be examined. The results are tabulated below. It is seen that there are 7 possible pairs of states that could participate in magnetically allowed transitions in a tetrahedral molecule. Ty £ 8, 3C, 65, boy irred. components AxT 3 0 al oat T ApxT, 3 0 at 1 u A,xT, 3 0 aA 1 ™ A,xT 3 0 - 1 ty ExBo @ 21 4 0 o A, +A, +E ExT, 6 0 2 0 ° +t ExT, 6 0 2 0 ° T+, Tet 9 0 to. L A, +E+T +7, Txt 9 90 Dob 2 A, + E+ T+ T, xT 9 9 2.1 1 AP + E+ TM +Ta Chapter 6 1. (a) It should be obvious by inspection that there are four sets of orbitals, members of which can be exchanged only among themselves by the symmetry operations of Dy, These are (1,4,5,8), (2,3,6,7). (9,10) (b) Each of these sets must be used as the basis of a reducible representation of Dj, and then each of these representations must be reduced. We must refer to the indicated coordinate system so we know which plane and axis is which. We shall go through the first set of orbitals, (1,4,5,8) in detail. For E, each orbital goes into itself x = 4. For all Cy rotations, i, o(xy) and o(yz), each orbital goes somewhere else. Therefore, the character for each of these operations {s 0. Reflection in the xy plane changes each px orbital ints the negative of itself, thus giving a matrix with -1 at each diagonal position and “4. The results for all three sets ave as follow: Day E Ca(z) Caly) Gel) t _ofay) o(xz) _oyz) 14,5,8 o 0 o a 0 ° ° 1 3,6,7 4 0 ° ° ok 0 ° "3,10 2 0 2 ° Oo 22 o 2 r. 1,4,5,8 7 12,3,6,7 7 Bag * Bag * Au * Pre T3107 Bag * Pia (ce) The ten normalized SALCSs can be written down by inspection of the character table since all representations are one-dimensional. For @ 3), combination of functions 1,4,5 ané 8 we require a linear combination that