Professional Documents
Culture Documents
to
Groups, Invariants
and
Particles
Frank W. K. Firk, Emeritus Professor of Physics, Yale University
2000
CONTENTS 2
1. INTRODUCTION 3
2. GALOIS GROUPS 8
LIE GROUPS 78
TRANSFORMATIONS 107
INTRODUC TION
familiar one. In Moder n Phy sics, this noti on ha s evo lved to in clude
sym metri es of an a bstra ct ki nd. These new symme tries play an e ssential
par t in the t heories of the microstruc ture of ma tter. The basic sym metri es
the mselves laws th at go vern the m otions of the g alaxies on the one
num ber o f imperfect obs ervat ions by a small numb er of reno wned
scientists an d mat hemat icians. It was not until the Einsteinian era ,
how ever, that the significance of the symme tries associated wit h the laws
was fully app reciated. The discovery of space-time symmet ries has led to
the widely-held belief that the l aws o f Nat ure c an be deri ved f rom
laws and rest ricted sym metry prin ciples to guide us i n our sear ch fo r the
fundamen tal symmet ries. Fre quent ly, it is impor tant to un derst and w hy
ori entat ion constitutes a st ate w hose symme try p roper ties, or i nvariants,
are to b e stu died. Any tran sformation that leaves th e sta te un chang ed in
sym metry . If the numbe r of conditions that define th e sta te is redu ced
the n the symm etry of th e sta te is incr eased. Fo r exa mple, an o bject
cha racte rized by o blateness alone is symmet ric u nder all transformat ions
In descr ibing the symme try o f a sta te of the most gener al kind (n ot
simply g eomet ric), the algebraic struc ture of th e set of s ymmet ry op erato rs
mus t be given; it is not suf ficient to give the numbe r of opera tions, and
not hing else. The law of c ombin ation of the opera tors must be st ated. It
is the alg ebrai c gro up th at fu lly charac terizes th e sym metry of t he ge neral
sta te.
The The ory o f Gro ups ca me ab out u nexpe ctedly. G alois show ed
tha t an equat ion o f deg ree n , whe re n is an inte ger g reate r tha n or equal to
five can not, in general, be solved by algebraic means. In the cours e of this
gre at wo rk, h e dev eloped the ideas of Lagra nge, Ruffini, and Ab el and
introduc ed th e con cept of a gro up. Galois discussed the func tional
the root s.
The oper ators that tran sform one funct ional rela tions hip i nto 5
ano ther are e lemen ts of a se t tha t is characteristic of th e equ ation ; the set
In the 1 850’s , Cay ley showed that ever y finite g roup is isomorp hic
to a cer tain permu tation gro up. The g eomet rical symm etries of cryst als
are desc ribed in t erms of finite group s. T hese symme tries are discussed in
man y sta ndard work s (see bibliography) and there fore, they will not be
In the b rief period bet ween 1924 and 1 928, Quant um Me chanics
was deve loped. Almost immediately, it was r ecogn ized by Weyl, and by
ana lytical to ol in Quan tum P hysics. T heir ideas have been deve loped over
the deca des in man y are as th at ra nge f rom t he Th eory of So lids to Pa rticle
Phy sics.
by Wigner. T hese group s are know n as Lie Gro ups. They have becom e
par ticularly Nuclear an d Par ticle Phys ics. Fifty yea rs af ter G alois had
met hods of solving different ial equations t hat h ad ev olved over a pe riod of
two hund red y ears. Inf initesimal unit ary t ransformat ions play a key role in 6
par ticle, or system of particles, unde r the Galilean trans forma tion is a basic
par t of every day r elativity. The sear ch fo r the tran sformation that leaves
und er a linear tra nsformation of the s pace-time coord inates, led to the
this tra nsformation, an d its rela ted i nvariants, cann ot be over stated.
This introduc tion to Gr oup T heory , with its emph asis on Li e Gro ups
and thei r app lication t o the stud y of symme tries of t he fu ndame ntal
con stituents of ma tter, has its origin in a one- semester c ourse that I ta ught
at Yale University for more than ten y ears. The cour se wa s dev eloped for
Sen iors, and advan ced J unior s, ma joring in the P hysical Sciences. T he
stu dents had gener ally completed the c ore c ourse s for thei r maj ors, and
had take n int ermed iate level cour ses in Lin ear A lgebra, Real and Com plex
Ana lysis, Ord inary Line ar Di fferential Equa tions, and some of t he Sp ecial
Fun ctions of Physics. Group Theo ry wa s not a ma thema tical requ irement
und ergra duate text books on G roup Theor y and its applications in Phys ics
ten d to be either highly qua litative or hig hly m athem atical. The pu rpose of
this introduc tion is to stee r a m iddle cour se th at pr ovides the stud ent w ith
a s ound mathe matical basis for st udying the symm etry prope rties of t he
fundamen tal p articles. It i s not gene rally appr eciated by Phys icists tha t 7
forms that involve differential operators and their commutators, and these
operators and their algebraic properties have found, and continue to find, a
Guilford, CT.
June, 2000
2 8
GALOIS GROUPS
In the early 19th - century, Abel proved that it is not possible to solve the
Abel did not solve this fundamental problem. The problem was taken up and
y 3 + 3Hy + G = 0
where
H = AC - B2
and
If all the roots are real, a trigonometrical method can be used to obtain
Putting
(s > 0),
gives
s = 2 √(-H)
and
cos3u = -4G/s3.
u = v, and 2π/3 ± v.
In this case,
H = -1 and G2 + 4H 3 = -3.
All the roots are therefore real, and they are given by solving
or,
3u = cos-1(-1/2).
The values of u are therefore 2π/9, 4π/9, and 8π/9, and the roots are
The roots x1, x2, and x3 exhibit a simple pattern. Relationships among
them can be readily found by writing them in the complex form 2cosθ = e iθ +
x1 = e iθ + e -iθ ,
x2 = e 2iθ + e -2iθ ,
and
x3 = e 4iθ + e -4iθ .
x 12 = x2 + 2,
x 22 = x3 + 2,
and
x 32 = x1 + 2. 11
The relationships among the roots have the functional form f(x1,x2,x3) = 0.
Other relationships exist; for example, by considering the sum of the roots we
find
x 1 + x22 + x2 - 2 = 0
x 2 + x32 + x3 - 2 = 0,
and
x 3 + x12 + x1 - 2 = 0.
angle, .
f(x1,x2,x3) → f(x2,x3,x1),
2
f(x1,x2,x3) → f(x3,x1,x2),
and
3
f(x1,x2,x3) → f(x1,x2,x3).
3
The operator = I, is the identity.
(x12 - x2 - 2) = (x22 - x3 - 2) = 0,
and
2
(x12 - x2 - 2) = (x32 - x1 - 2) = 0. 12
x 1 + x22 + x2 - 2 = 0
is valid, then so is
to a field which is closed under the rational operations. If the field is the set
x2 - 3 = 0
we see that the coefficient -3 belongs to Q, whereas the roots of the equation,
permutations of the roots. The symmetry operators formed the Galois group
of the equation.
x n + a 1xn-1 + a 2xn-2 + .. an = 0,
functional relations of the roots are given in terms of the coefficients in the
standard way
x1 + x2 + x3 .. .. + xn = -a1
Rational symmetric functions also can be constructed that involve the roots
and the coefficients of a given equation. For example, consider the quartic
x4 + a 2x2 + a 4 = 0.
x 1 + x2 + x3 + x4 = 0
x 1x2x3x4 = a 4.
rational symmetric function that belongs to the field F of the coefficients, ai, of
coefficients.
has the property that if a rational function of the roots of the equation is
equation
x 6 = 3.
It can be solved by writing x3 = y, y2 = 3 or 15
x = (√3)1/3.
the atomic nature of the equation, itself. (In Chapter 5, it will be seen that a
some insight into the Galois method, however, by studying the group
x 4 + a 2x2 + a 4 = 0
where
µ = (a 22 - 4a4)1/2.
The field F of the quartic equation contains the rationals Q, and the
The relations
x 1 + x2 = x3 + x4 = 0
are in the field F. 16
x1 x2 x3 x4 x1 x2 x3 x4 x1 x2 x3 x4
{ P1 = , P2 = , P3 = ,
x1 x2 x3 x4 x1 x2 x4 x3 x2 x1 x3 x4
x1 x2 x3 x4 x1 x2 x3 x4 x1 x2 x3 x4
P4 = , P5 = , P6 =
x2 x1 x4 x3 x3 x4 x1 x2 x3 x4 x2 x1
x1 x2 x3 x4 x1 x2 x3 x4
P7 = , P8 = }.
x4 x3 x1 x2 x4 x3 x2 x1
irrational expressions of the coefficients, other true relations among the roots
can be formed in the extended field, F'. Consider, for example, the extended
x 12 - x32 = µ is in F'.
x 1 = -x 2 and x3 = -x 4
so that
The permutations that leave these relations true in F' are then
This set is the Galois group of the quartic in F'. It is a subgroup of the set
{P 1,...P8}.
((-a2 - µ)/2) 1/2 to form the field F'' then the relation
{P1, P3}.
This set is the Galois group of the quartic in F''. It is a subgroup of the set
((-a2 + µ)/2) 1/2 to form the field F''' then the relation
The full group, and the subgroups, associated with the quartic equation
distinct elements that it contains). In 5.4, we shall prove that the order of a
subgroup is always an integral divisor of the order of the full group. The
order of the full group divided by the order of a subgroup is called the index 18
of the subgroup.
HG - GH = [H,G] = 0.
the symmetric group, Sn, or a subgroup of Sn, (see 5.6). The Galois method
2. The choice of a maximal subgroup of Hmax(1) . In the above case, {P1, ...P8}
composition indices are given by the ratios of the successive orders of the
groups:
gn/h(1) , h(1) /h(2) , ...h(k-1)/1. 19
to be:
n Composition Indices
2 2
3 2, 3
4 2, 3, 2, 2
5 2, 60
6 2, 360
7 2, 2520
group is solvable.
Galois defined a solvable group as one in which the composition indices are
with (n!)/2 elements. The composition indices are then 2 and (n!)/2. For n >
4, however, (n!)/2 is not prime, therefore the groups Sn are not solvable for n
> 4. Using Galois' Theorem, we see that it is therefore not possible to solve,
m n
B = ∑ ∑ aijxiyj
I=1 j=1
where
x = [x1,x2,...xm], an m-vector,
y = [y1,y2,...yn], an n-vector,
column vector.
B = xTAy
where
a 11 . . . a1n
. . . .
A= . . . . .
. . . .
a m1. . . amn
Q = xTAx.
quadratic form
x' = Mx
where
x = [x,y],
x' = [x',y'],
and
|i j|
M = .
|k l|
oblique coordinate system in which the new x'- axis has a slope (k/i), and the
y′
[i+j,k+l]
[j,l]
x'
[0,1] [1,1]
x′
[i,k]
[0,0] [1,0] x
binary quadratic:
Q(x,y) = xTDx
where
|a h|
D= ,
|h b|
where
a' h'
D' = ,
h' b'
and
Now,
Q'(x',y') = (Mx)TD'Mx
= xTMTD'Mx
MTD'M = D.
(detM)2detD' = detD
because
detMT = detM.
apart from a factor (il - jk)2 that depends on the transformation itself, and not
Mathematics.
= ∑ aix1n-ix2i
forms.
follows
The coefficients
f(x1,x2) = 0
f'(x1',x2') = 0.
r wI(ao',a1',...an') = I(ao,a1,...an)
transformed. The degree of the invariant is the degree of the coefficients, and
the exponent w is called the weight. In the example discussed above, the 25
It is found that the Jacobian of two binary quadratic forms, f(x1,x2) and
∂f/∂x1 ∂f/∂x2
∂g/∂x1 ∂g/∂x2
The determinant
∂2f/∂x12 ∂2f/∂x1∂x2
,
∂ g/∂x2∂x1 ∂ g/∂x2
2 2 2
Example:
To show that
= xTDx
where
x = [x,y],
and
a ox + a 1y a 1x + a 2y
D= .
a 1x + a 2y a 2x + a 3y
i j
M=
k l
then 27
f(x,y) = f'(x',y')
if
D = MTD'M.
detD = (detM)2detD',
detD = φ(x,y)
and
detD'= φ'(x',y')
so that
φ(x,y) = (detM)2φ'(x',y'
= xTEx,
where
Also, we have
therefore
xTEx = (detM)2xTMTE'Mx
so that
E = (detM)2MTE'M.
detE = (detM)4detE'
x' = Mx.
4 29
Eve nts o f finite extens ion and du ration are part of t he ph ysical
wor ld. It will be conv enient to intro duce the n otion of ide al ev ents th at
hav e neither exten sion nor d uration. Ideal even ts ma y be repre sented as
mat hemat ical points in a spa ce-time ge ometr y. A part icular eve nt, E, is
described by the f our c ompon ents [t,x, y,z] where t is the time of th e eve nt,
and x,y, z, ar e its thre e spa tial coord inates. The tim e and spac e coo rdinates
are refe rred to ar bitrarily chosen ori gins. The spat ial mesh n eed n ot be
Car tesian.
axis, be reco rded as th e eve nt E'[t ',x'] by a seco nd ob serve r O', movi ng at
con stant spee d V a long the x -axis. We supp ose t hat t heir clocks are
syn chron ized at t = t' = 0 w hen t hey c oincide at a co mmon origin, x = x' =
0.
t' = t
and
x' = x - Vt,
be writt en
E' = GE
whe re 30
1 0
G = .
-V 1
t = t'
and
x = x ' + V t'
or
E = G-1E'
whe re G-1 is the inverse Ga lilean ope rator . (It und oes t he effect of G).
length.
algebraic for m. We not e, ho wever , the key role played by acc eleration in
v' = -V + v,
whe re a and a ' are the accelerations in the two frames of refer ence. The
mov ing in emp ty sp ace a t v = c ≅ 3 x 10 8 m/ s, it does not fit the facts. All
4.2 . Lor entz invar iance and Einst ein's spac e-tim e
sym metry .
It was E instein, above all others , who adv anced our under stand ing o f
the true natu re of spac e-time and rela tive motion. We sha ll see tha t he
mad e use of a symm etry argum ent t o find the chan ges t hat m ust b e mad e
rap idly moving obj ects and o f beams of light. H e rec ognized an
whe re
1 0 1 0
-1
G = an d G = .
−V 1 V 1
incorpor ated this principle in his the ory. He a lso r etained th e lin earit y of
(Eq uispaced interv als of time and distance in on e inertial fram e rem ain
equ ispaced in any other iner tial frame). H e the refor e symmetri zed th e
t' 1 -V t
= .
x' -V 1 x
cor recte d by intro ducing the invariant spee d of light, c a postu late in
ct' = ct − Vx/c
instead of th e scaling factor of unity that appe ars in the Galilean equat ions
whe re
γ - βγ
L = , and E = [ct,x ]
- βγ γ
E = L-1E'
whe re
γ βγ
-1
L = .
βγ γ
This is the i nvers e Lor entz trans forma tion, obta ined from L by chan ging
LL-1 = I
or
γ - βγ γ βγ 1 0 34
= .
-βγ γ βγ γ 0 1
γ2 − β2γ2 = 1
the refor e,
Pre viously, it was show n tha t the spac e-time of Galileo an d New ton
+γ2(1 + β2)x' 2
invariant und er L:
bec ause
γ2(1 − β2) = 1.
and
xµ = [x0,-x1,-x2,-x3].
E µ = [ct,-x,-y,-z]
E µT E µ = (ct)2 − (x 2 + y2 + z2).
x' µ = Lxµ
whe re
γ - βγ 0 0
- βγ γ 0 0
L= .
0 0 1 0
0 0 0 1
Imp ortan t con sequences of th e Lor entz trans forma tion are t hat
intervals of time measured i n two different iner tial frames are not the s ame
∆l' = ∆l/γ
defined by
= [γc,γvN] .
Vµ Vµ = (γc)2 - (γvN)2
= (γc)2(1 - (vN/c)2)
= c 2.
(In form ing t he sc alar produ ct, t he tr anspose is unde rstoo d).
of its r ole as an invariant in an isolated system. We the refor e int roduc e the
pos sible Lore ntz i nvariants involving this new q uantity. The c ontra variant
Pµ = mVµ
P µ Pµ = (mc)2.
Now ,
Pµ = [mγc, mγvN]
the refor e,
P µ Pµ = (mγc)2 − (m γvN)2.
Writing
The quan tity Mc2 ha s dimensions of ener gy; w e the refor e wri te
E = Mc2
c2Pµ Pµ = E 2 − (p c)2 = E o2
whe re
E = γE o = E o + T,
whe re
T = Eo(γ − 1) ,
Pµ = mc.
P'µ = LPµ .
For rela tive motion along th e x-axis, this equat ion is equ ivalent to the
equ ations
E' = γE − βγcpx
and
ν' = γν − βγν
= γν(1 − β) 39
This is the r elativistic Dop pler shift for the f reque ncy ν', measured i n an
pro posed by d eBrog lie in the earl y 192 0's. He u sed t he fo llowing
arg ument .
written ((ω/c)ct − k•r), and t his has th e for m of a Lor entz invariant
whe re Eµ is the event 4-ve ctor, and Kµ is the "freq uency -wavenumbe r" 4-
vec tor.
E µ [ct ,r]
(deBroglie)
rea lized that ther e cou ld be only one value for the c onsta nt of
pro porti onality if the Planck-Einstein resu lt for pho tons E = h ω/2π is but a
Pµ ∝ K µ
or
Pµ = (h/2π)Kµ .
E = (h/2π)ω
and
p = (h/2π)k .
In these rema rkable equ ations, ou r not ions of pa rticles an d wav es ar e 41
for ever merge d. T he sm allness of the value of P lanck's co nstan t pre vents
us from obser ving the d uality dir ectly; how ever, it is clearly obser ved a t
K µ Kµ = (ω/c, k)[ω/c,-k]
ωo2 = ω2 − (k c)2.
ωdω − kc 2dk = 0
the refor e,
v G = dω/dk = kc 2/ω.
The deBr oglie invariant invo lving the produ ct of the phase and group
velocity is t herefore
We see t hat
v φvG = c2 = E/M
or,
velocity.
This res ult p layed an i mport ant p art i n the deve lopment of Wave
Mechanics.
We shall find in later chapt ers, that Loren tz tr ansformati ons form a
The elements of the set {±1, ±i}, where i = √-1, are the roots of the
equation x4 = 1, the “fourth roots of unity”. They have the following special
properties:
1. The product of any two elements of the set (including the same two
3. A unique element of the set exists such that the product of any
element of the set and the unique element (called the identity) is equal to the
that the product of the element and its corresponding element (called the
The set of elements {±1, ±i} with these four properties is said to form
a GROUP.
need not be the case. For example, the set of integers Z = {.., -2, -1, 0, 1, 2,
...} forms a group if the law of composition is addition. In this group, the
identity element is zero, and the inverse of each integer is the integer with the
1 --closure
and
2 --associativity,
and to contain:
and
4 --inverses.
concept. In the first example, the group elements are real and imaginary
numbers, in the second, they are positive and negative integers, and in the
third, they are matrices that represent linear operators (see later discussion).
the diagonal.
A group of elements that has the same structure as an abstract group is
rotations in the plane about its center, and under reflections in the three
Rz (2π/3)(123) → (312) = a
Rz (4π/3)(123) → (231) = a 2
RI (123) → (123) = b
Positive rotations are in an anticlockwise sense and the inverse rotations are in
e a a2 b c d
e e a a2 b c d
a a a2 e d b c
a2 a 2 e a c d b
b b c d e a a2
c c d b a2 e a
d d b c a a2 e
In reading the table, we follow the rule that the first operation is written on 47
the right: for example, ca2 = b. A feature of the group D3 is that it can be
involving {b, c, d}. The set {e, a, a2} forms a group called the cyclic group
of order three, C 3. A group is cyclic if all the elements of the group are
ak an-k = a n = e,
The group D 3 can be broken down into a part that is a group C3, and a
part that is the product of one of the remaining elements and the elements of
D3 = C 3 + bC 3 , b ∉ C3
= cC3 = dC 3.
Let
elements
This is called the left coset of Hm with respect to g k . We note the important
facts that all the elements of gk hj, j=1 to m are distinct, and that none of the
elements gk hj belongs to Hm .
divisible by m, hence n = Km, where the integer K is called the index of the
where
gj2 ∈ Gn ∉ Hm , 49
gk3 ∈ Gn ∉ Hm , gj2 Hm
group S3) and let Hm = C 3 = {123, 312, 231}, the cyclic group of order
three. The elements of S3 that are not in H3 are {132, 213, 321}. Choosing
gk = 132, we obtain
and therefore
S 3 = C 3 + gk2 C3 ,K = 2.
conjugate to a and the class conjugate to b are the same. If a is not conjugate
by itself.
so that
ba = ab.
If Hm is a subgroup of Gn, we can form the set
aHm a-1 = H m
(all conjugate subgroups of Hm in Gn are identical to Hm ), 51
Now,
and
The decomposition of S3 is
S3 = H 3 + bH 3 = H 3 + H 3b.
H 3b = H 3c = H 3d.
e=e
eae-1 = ea = a
aaa-1 = ae = a
a 2a(a2)-1 = a2a2 = a
bab-1 = bab = a2
cac-1 = cac = a2
dad-1 = dad = a2
decomposed by classes:
All left and right cosets are not equal: Hm = {e, b} is therefore not an
= Hm + aHm + a2Hm .
order 1: {e}
order 6: S 3 .
5.6 Permutations
A permutation of the set {1, 2, 3, ....,n} of n distinct elements is an
n!
π2π3 = π1
and
π3π2 = π1
whereas
π4π5 = π3
and
π5π4 = π2.
s 1 s2 . . .s n 1 2 . . n 1 2 . . n
= .
t1 t2 . . .tn s 1 s 2 . . sn t 1 t 2 . . tn
s 1 s 2 . . . sn
π-1 = | |
1 2 . . . n
g1 g2 . . gn
πi = .
g ig1 gig2 . . gign
g1 g2 . . gn
πj = ,
g jg1 g jg2 . . gjgn
where gi ≠ g j, then
g1 g2 . . gn
πjπi = .
(g jgi)gi (gjgi)g2 . . (gjgi)gn
This is the permutation that corresponds to the element gjgi of Gn. 56
the full symmetric group of order n! that contains all the permutations of the
has its origin in the theory of differential equations, we shall introduce the
system as follows:
v´ = R v,
where
cosθii´ . .
= cosθij´ . .
cosθik´ . cosθkk´
where i, j, k, i´, j´, k´ are orthogonal unit vectors, along the axes, before and
v´ = Rv(φ)v,
where
Rv(φ) transforms a point P[x, y] in the plane into a “new” point P´[x´, y´]:
Rv(δφ) = 1 −δφ
δφ 1
and
x´ = x + δx and y´ = y +δy
so that
We note that
Rv(δφ) = 1 0 + 0 -1 δφ
0 1 1 0
= I + iδφ
where
i= 0 -1 = Rv(π/2).
1 0
so that
a quantity that differs from the identity I by a term that involves the
= iRv(φ) .
This is Lie’s differential equation.
Rv(φ) φ
∫ dR (φ)/R (φ) = i ∫ dφ
v v
I 0
so that
ln(Rv(φ)/I) = iφ,
or
Rv(φ) = Ieiφ , the solution of Lie’s equation.
Previously, we obtained
We have, therefore
Ieiφ = Icosφ + isinφ .
δx = Of(x, y; δφ) ?
and
δy = O(yδφ) = (x∂/∂y − y∂/∂x)y∂φ = xδφ . 61
δxi = Xxiδφ , i = 1, 2
where
Rv(φ) = eiφ,
and therefore
Rv2(δφ) = (I + iδφ)2
= Rv(2δφ).
This result agrees, as it should, with the exact solution of Lie’s differential
equation.
A finite rotation can be built up by exponentiation of infinitesimal 62
rotations, each one being close to the identity. In general, this approach has
the plane about the origin of a cartesian reference frame with a momentum p
is
Lz = r × p = rpsinφnz
where nz is a unit vector normal to the plane, and φ is the angle between r
and p.
components of p.
replacing
is Planck’s constant.
and therefore
X = iLz Q/(h/2π), 63
and
x´ = x - yδφ
and
y´ = y + xδφ .
= I + 2πiδφL z /h)f(x, y)
= e2πiδφLz/h f(x, y)
The invatriance of length under rotations follows at once from this result:
If f(x, y) = x2 + y2 then
∂f/∂x = 2x and ∂f/∂y = 2y, and therefore
= f(x, y) = x2 + y2 = invariant.
This is the only form that leads to the invariance of length under rotations. 64
z
z′ y′
µ about x
y y
x x, x′
x′ x′′ x′
z′′ z′′′
y′′ y′′′
φ about z´´
For infinitesimal rotations, the total rotation matrix is, to 1st-order in the δ’s:
1 δφ −δθ
Rc (δµ, δθ, δφ) = −δφ 1 δµ .
δθ −δµ 1
The infinitesimal form can be written as follows: 65
1 δφ 0 1 0 −δθ 1 0 0
Rc (δµ, δθ, δφ) = −δφ 1 0 0 1 0 0 1 δµ
0 0 1 δθ 0 1 0 −δµ 1
where
0 0 0 0 0 -1 0 1 0
Y1 = 0 0 1 , Y2 = 0 0 0 , Y3 = -1 0 0 .
0 -1 0 1 0 0 0 0 0
To 1st-order in the δ’s, we have
The algebraic properties of the Y’s are important. For example, we find
0 0 0 0 0 -1 0 0 -1 0 0 0
[Y1, Y2] = | 0 0 1|| 0 0 0| − | 0 0 0|| 0 0 1|
0 -1 0 1 0 0 1 0 0 0 -1 0
= −Y3 ,
[Y1, Y3] = Y2 ,
and
Jk = -i(2π/h)Yk , k = 1, 2, 3.
Their commutators are obtained from those of the Y’s, for example
or
and therefore
δy = y´ - y = -xδφ + zδµ
δz = z´ - z = xδθ - yδµ .
For example, if i = 1
+ δθ(-z∂/∂x + x∂/∂z)x
3
Rc (δµ, δθ, δφ) = I + ∑ (∂Rc /∂αi)| ⋅ δαi .
i= 1 All αi’s = 0
7 68
dir ectly to t he an gular mome ntum opera tors of Qu antum Mech anics.
Imp ortan t al gebra ic pr opert ies of the matr ix re prese ntations o f the
ope rator s also wer e int roduc ed. In th is chapter , we shall cons ider the
x´ = f(x; a) or x´ = T a x
the x’s and t he a’ s to any r equired or der. These fun ctions nec essarily
the smallest numbe r req uired to c harac terize the tran sformation,
com pletely.
= f(x ; c)
cλ = χλ (a ; b) , λ = 1 to r,
and
2. To e very trans forma tion there corr esponds a unique inverse that
belongs to th e set :
We have
Con sider the trans forma tion x → x´ unde r a f inite chan ge in a si ngle
x´ + dx´
a + da
a “di fferential”
x (a = 0)
We have
x ´ + d x´ = f(x; a + d a)
a + da = χ(a; δa).
But
χ(a; 0) = a, (b = 0)
the refor e
a + da = a + ∂χ(a; b)/∂ b δa
b= 0
so that
d a = ∂ χ(a; b)/∂ b δa
b= 0
or
δa = A(a)da.
Therefor e
dx ´ = u (x´)A(a)da,
leading to
dx ´/u(x´) = A(a )da 71
so that
x 0
U (x´) - U(x) = s .
A t ransformat ion o f coo rdinates (new v ariables) there fore trans fers all
x 1´ a 0 x1
x 2´ = 0 a 2 x2
the refor e
x 1´ ( 1 + δa) 0 x1
x 2´ = 0 (1 + δa) x2 .
2
To 1st-order in δa w e hav e
x 1´ = x1 + x1δa
and
x 2´ = x2 + 2x2δa
or
δx1 = x 1δa
and
δx2 = 2x 2δa. 72
∫ dx /x ∫ da ∫ dx /2x
x1´ a x2´ a
1 1 = and 2 2 = da ,
x1 0 x2 0
so that
lnx 1´ - ln x1 = a = ln (x1´/x1)
and
ln (x2´/x2) = 2a = 2ln(x1´/x1)
or
U´ = (x2´/x1´2) = U = ( x2/x12) .
7.2 Det ermin ation of t he fi nite equat ions from the i nfini tesim al
for ms
x 1´ = φ(x1, x 2 ; a)
and
x 2´ = ψ(x1, x 2 ; a),
and let the i dentity co rresp ond t o a = 0. 73
f(x1´, x2´) in a Maclaurin series in the p arame ter a (at definite v alues of x 1
and x 2):
whe re
and
= {(∂f/∂x1´)(dx1´/da) + (∂f/∂x2´)(dx2´/da)}| a= 0
the refor e
= Xf(x1, x 2).
Con tinuing in this way, we h ave
Xnf is the symb ol for ope rating n-t imes in succession o f f w ith X.
and
x 2´ = x2 + aXx2 + (a2/2!)X2x2 + = ...
If x1 a nd x 2 ar e definite values to which x 1´and x 2´ r educe for the i dentity 74
equ ations
the n
= x 1 + ax 1 + (a2/2!)x1 + ...
= x 1ea .
Also, we find
x 2´ = x2ea .
x 1´ = bx 1, a nd x 2´ = bx 2.
Let
gro up, a nd let a=0 give the identity. A fu nction F(x 1, x 2) is ter med a n
If
it is necessary fo r
XF = 0,
Con sequently,
F(x1, x 2) = co nstan t
is a solution of
dx 1/u(x1, x 2) = dx 2/v(x1, x 2) .
the refor e G(1) ha s only one basic invariant, an d all othe r pos sible invariants
and the differential equation tha t giv es th e invariant fun ction F of the
so that
dx1/x2 + dx 2/x1 = 0.
x 12 + x 22 = cons tant,
This met hod c an be gene ralized. A gro up G(1) in n-va riables de fined
by the e quation
Let (x1, x 2, . ..xn) b e the coor dinates of a po int in n-space and let a be a
par amete r, in depen dent of th e xi’s. As a var ies, the p oint (x1, x 2, . ..xn) w ill
nec essary and sufficient con dition tha t F(x 1, x 2, . ..xn) b e an invariant
invariant cur ve if
XF(x 1, x 2, x 3, . ..xn) = 0.
8 78
where
r
dx i = ∑ u iλ(x)δaλ , the Lie form.
λ=1
r
= ∑ δaλ Xλ F
λ=1
The operator
r
I + ∑ Xλδaλ
λ=1
Theory of Lie Groups. The Xλ’s are differential operators. The problem is
where the coefficients uiλ are functions of the variables x1, x2, and f(x1, x2)
The difference is
- X2u11∂f/∂x1 - X2u12∂f/∂x2.
≡ [X1, X2]f.
Xa = uia∂f/∂xi
n
= ∑ u ia∂f/∂xi , ( a sum over repeated indices).
i=1
(∂ckτσ /∂aρ)uik = 0
in which 80
written
= c kρσXk.
The commutators are linear combinations of the Xk’s. (Recall the earlier
The ckρσ’s are called the structure constants of the group. They have the
properties
c kρσ = -ckσρ ,
Lie made the remarkable discovery that, given these structure constants,
1961).
8.1 The rank of a group
of a group, Xi:
[A, X] = 0.
For example, consider the orthogonal group, O+ (3); here
A = αiXi i = 1 to 3,
and
X = xjXj j = 1 to 3
so that
[A, X] = αixj[Xi, Xj] i, j = 1 to 3
= αixjεijkXk .
-α2 α1 0 x1 0
α3 0 -α2x2 = 0 .
0 -α3 α2 x3 0
x j = αj (j = 1, 2, 3)
so that
A = X.
operators, Jk:
Jk = -i(h/2π)Yk (k = 1, 2, 3).
0 0 0 0 0 -1 0 1 0
Y1 = 0 0 1, Y2 = 0 0 0 , Y3 = -1 0 0 .
0 -1 0 1 0 0 0 0 0
The square of the total angular momentum, J is
3
J2 = ∑ Ji2
1
= (h/2π)2(-2I).
The operator J2 with this property is called the Casimir operator of the
group O+ (3).
with the elements of the set of irreps of a given group, forms the set of
Casimir operators of the group. All Casimir operators are constant multiples
Ci = aiI.
Mat rix r epres entat ions of linear opera tors are i mport ant i n Lin ear
Algebra; we s hall see t hat t hey a re eq ually impo rtant in G roup Theor y.
can be f ound in wh ich each e lement is associated with the corre spond ing
G n = {g1,...gk,....gn},
and
The subject o f Gro up Re prese ntations forms a ver y lar ge bra nch o f
Gro up Th eory. The re ar e man y sta ndard work s on this topic (see the
bibliography), each one cont aining num erous definitions, lemmas and
the orems . He re, a rath er br ief accoun t is given of s ome o f the more
importan t res ults. The read er sh ould delve into the deeper asp ects of th e
sub ject as th e nee d ari ses. The subject will be intr oduce d by considering
rep resen tations of the rotat ion g roups , and thei r cor respo nding cyclic 84
gro ups.
9.1 The 3-di mensi onal representa tion of ro tatio ns in the plane
c osφ -sinφ
Rv(φ) = .
s inφ cos φ
tri angle invariant unde r rot ations in the p lane is of orde r thr ee, and ea ch
= 1 0 , -1 /2 -√3/2 , -1/2 √3 /2 .
0 1 √3/2 -1/ 2 , -√3 /2 -1 /2
Con sider an e quilateral tria ngle located in the plane and let t he
coo rdinates o f the thre e ver tices P 1[x, y], P 2[x´ , y´] , and P 3[x´ ´, y´´] be
P 13 → P 23[P 3, P 1, P 2] i s given by
P 23 = D2(3) P 13,
whe re
0 0 1
(3)
D2 = 1 0 0 ,
0 1 0
P 13 → P 33[P 2, P 3, P 1] i s given by
P 33 = D3(3) P 13
whe re
0 1 0
(3)
D3 = 0 0 1 .
1 0 0
dim ensio nal r epresentat ion of the original 2-dimensional re prese ntation
is repre sented by
Rn(m) = {R1(m) , R2(m) , . ....Rn(m) 86
whe re
Rn(m) ~ Gn,
P 1d = [P 1, P 2, . ..P d],
and let
be an m- vecto r, wr itten in n ormal orde r, in which the comp onent s are each
d-vector s. Introd uce t he m × m matri x ope rator Dk(m) (gk) such t hat
whe re Pkm is the kth ( cyclic) permuta tion of P1m , and Dk(m) (gk) is called
= T(x)T(y),
cha nge o f var iable in t he co rresp onding sub stitutions. Re prese ntations
different rep resen tations. All repres entat ions that are e quivalent to S ar e
equ ivalent to each othe r, an d the y for m an infinite class. Two equivalent
S(g) 0
P = , (g ∈ G)
0 T(g)
of dimension s + t is c alled the dir ect s um of the matri ces S(g) and T(g),
sam e), we can obta in a third by a dding them dire ctly. Alt ernat ively, let P
A(x) 0
0 B(x)
whe re A(x) and B(x) are s × s and t × t matri ces, respe ctively. (The 0’s 88
A(x)A(y) 0
= .
0 B(x)B(y)
rep resen tations. The r epres entat ion P is said to b e dec ompos able, w ith
rep resen tation int o a f inite numb er of indecompo sable comp onent s. (It
sho uld b e not ed th at th e pro perty of i ndeco mposa blity depe nds o n the field
of the r epres entat ion; the r eal field must somet imes be ex tende d to the
A(x) 0
P(x) =
E(x) B(x)
whe re A(x), and B(x) are matri ces o f dimensions s × s and t × t 89
An irreduci ble rep resen tation is one t hat c annot be r educed. E very
dec ompos able matri x is reduc ible (E(x) = 0), where as a reduc ible
of irred ucible com ponen ts. The c ompon ents are d eterm ined uniquely, up
to an eq uivalence. The set of distinct irr educible r epres entat ions of a finite
gro up is (in a given field) an inv arian t of the group . Th e com ponen ts fo rm
9.4 Sim ilari ty an d uni tary trans forma tions and matri x
importan t res ults in th eTheory of Matr ices. The proo fs of thes e sta temen ts
Q-1AQ = B ,
the n the matr ices A an d B ar e rel ated by a sim ilari ty tra nsformatio n.
If Q is unit ary ( QQ† = I: Q† = (Q*)T , t he he rmitian co njugate) 90
com mute.
the eigenvalues of A.
tra nsformation.
rep lacing the elements of Ld by their positive squar e roo ts. Let U be a
Ld = UDU -1.
S = Ld-1/ 2U,
th en SDiS-1 is uni tary. (This prop erty can b e dem onstr ated by co nsidering
tra nsform the orig inal matri x rep resen tation Gn in to di agona l form. E very
unitary matri x is diagonalizable, and there fore every matr ix in ever y finite
tra nsformation. I nvoking th e res ult o f the prev ious section, w e nee d onl y
D(ν)(R)A = AD(µ)(R) ∀ R ∈ Gn
or
Let the µ co lumns of A be writ ten c1, c2, . ..cµ, t hen, for a ny ma trices
an
µ µ µ
AD(µ) = ( ∑ D(µ)k1ck, ∑ D(µ)k2ck, . ..∑D(µ)kµck).
k= 1 k= 1 k= 1
the refor e µ
D(ν)cj = ∑ D(µ)kjck
k= 1
and ther efore the µ c-vectors span a sp ace t hat i s invariant und er th e
irr educible set of ν-dimensional matri ces { D(ν)}. The c-vectors are
In the s econd case, the hermi tian conjugates D(ν)1†, . ..D(ν)n† an d D(µ)1†,
...D(µ)n† also ar e irreduci ble . Furt hermo re, s ince D(ν)i(R)A = AD(µ)i(R)
D(µ)i†A† = A†D(ν)i† ,
D(ν)(R) gives a r epres entat ion o f G t hat i s equ ivalent to D(µ)(R). Th ese t wo
set s of matri ces are ge nerally different, w herea s the ir st ructu re is the same.
We wish, ther efore, to answer the ques tion: what intr insic prop erties of the
mat rix r epres entat ions are i nvariant u nder coord inate tran sformations?
Con sider
∑ [ CD(R)C-1]ii = ∑ CikDkl(R)Cli-1
i ikl
= ∑ δklDkl(R)
kl
χ(R) = ∑ Dii(R)
i
χ(µ)(R) or [ µ; R].
D(R) = D(U)D(R)[D(R)]-1
the refor e
χ(S) = χ(R).
We shall cons ider those Lie group s tha t can be d escribed b y a f inite
x i´ = f i(x1,...xn; a1,...ar) = f( x; a) .
tra nsformations:
x = f( x´; a).
The gen eral linea r gro up GL (n) in n-d imensions is given b y the set
of equat ions
n
x i´ = ∑ aijxj , i = 1 to n ,
j= 1
The grou p con tains n 2 pa ramet ers t hat h ave v alues cove ring an infinite
All linear gr oups with n > 1 are non-abelian. The group GL(n ) is
relation ther efore exists am ong t he n2- p arame ters so th at th e num ber o f
the n the rest ricted gro up is called th e ort hogon al gr oup, O (n), in n-
whe re
a11 a 12 a 13
O = a21 a 22 a 23 .
a 31 a32 a 33
We have
This invariance imposes six conditions on t he or iginal nine par amete rs, a nd
If the x i’s and t he aij’s of th e gen eral linear gro up GL (n) are
com plex, and the t ransformat ions are r equired to leave xx† in variant in the
com plex space, the n we obtain the unit ary g roup U(n) in n-dimensions:
There ar e 2n2 in depen dent real param eters ( th e rea l and imag inary part s of
tha t
∑j |a ij|2 = 1,
|aij|2 ≤ 1 fo r all i, j.
The para meter s are limited t o a f inite rang e of values, and the refor e the
the n we obtain the special unita ry group SU (n) in n-d imensions. We have
The dete rminantal condition reduc es th e num ber o f req uired real
par amete rs to (n2 - 1). We shall see th at th e gro ups S U(2) and S U(3) play
Pau li sp in ma trice s
a b
U =
c d
whe re a, b, c , d ∈ C. 97
a* c*
†
U = ,
b* d*
and ther efore
ad - bc = 1 .
c = - b*, a nd d = a*.
a b
U= .
-b* a*
str uctur e
1 0 δa δb 1 + δa δb
Uinf = + = .
0 1 -δb* δa* -δb* 1 + δa*
1 + δa* + δa = 1,
or 98
δa = -δa*.
so that
1 + δa δb
Uinf = .
-δb* 1 - δa
(The factor o f two has been intro duced for later conv enience).
mat rices
1 0 0 1 0 -i 1 0
, , , .
0 1 1 0 i 0 0 -1
as follows
a b 1 0 0 1 0 -i 1 0
=A +B +C +D ,
c d 0 1 1 0 i 0 0 -1
whe re
a = A + D, b = B - iC, c = B + iC, and d = A - D.
We then have
a b (a + d) 1 0 (b + c) 0 1 i(b - c) 0 -i (a - d) 1 0
= + + + .
c d 2 0 1 2 1 0 2 i 0 2 0 -1
The ’s are t he Pau li sp in-ma trices:; they are t he ge nerat ors o f the grou p
j
SU( 2):
0 1 0 -i 1 0
1 = , 2 = , 3 = .
1 0 i 0 0 -1
10. 6 Co mmuta tors of th e spi n mat rices and struc ture const ants
section, we c onsider th e com mutat ors o f the gene rator s of SU(2); the y are
[ 1, 2 ] = 2 i 3, [ 2, 1 ] = -2i 3,
[ 1, 3 ] = -2 i 2, [ 3, 1] = 2 i 2,
[ 2, 3 ] = 2 i 1, [ 3, 2 ] = -2 i 1.
con stant mult iplied by the v alue of th e rem aining mat rix, thus
[ j, k] = εjkl2i l .
whe re th e qua ntity εjkl = ±1, d epend ing o n the perm utations o f the indices.
P = xT = x j j, j = 1, 2, 3
x = [x1, x 2, x 3] a nd = [ 1, 2, 3]:
the refor e
x3 x 1 - ix2
P = .
x 1 + ix2 -x3
We see t hat
x3 x 1 - ix2
† T
P = (P*) = = P,
x 1 + ix2 -x3
TrP = 0,
and
Ano ther matri x, P´, can b e for med b y car rying out a similarity
tra nsformation, th us
P´ = UPU †, (U ∈ SU (2)).
A s imilarity trans forma tion leaves bot h the trac e and the deter minant
unc hange d, th erefore 101
T rP = TrP´,
and
d etP = det P´.
xxT = x´x´T,
or
x 12 + x22 + x32 = x 1´2 + x2´2 + x3´2 .
dim ensio nal or thogo nal t ransformat ion t hat l eaves xxT in variant.
We have seen that the b asic form of th e 2×2 m atrix repr esentation
of
th e gro up SU (2) is
a b
U = , a, b ∈ C; | a|2 + |b| 2 =1 .
-b * a*
1 0
x1 = and x2 = .
0 1
We then have
a
x1´ = Ux1 = = ax1 - b*x2 ,
-b *
and
b
x2´ = Ux2 = = bx1 + a*x2 ,
a*
and ther efore
x´ = Utx.
u ´ = au + bv
and
v ´ = -b*u + a*v ,
c´ = Uc .
fro m tho se of the basis vect or x — the t ransformat ion m atrices ar e the
u2, u v, an d v 2,
whe re
and
m = j, j - 1, ...-j .
(The factor 1 /√{(1 + m) ! (1 - m)!} is chosen to make the r epres entat ive
mat rix u nitary). 103
by defining t he tr ansformati on
so that
We then have
a2 √ 2ab b2 f 11 f11´
-√ 2ab* |a| - |b| √2a*b f 0 = f 01´
2 2 1
follows
Let
f mj(u, v) = u j + mvj - m , m = j, j - 1, ...-j. 104
√{(j + m) !(j - m)!}
For a gi ven v alue of j, ther e are 2j + 1 linearly ind epend ent p olynomials,
element U of SU(2 ):
The deta ils of this gen eral case are g iven in Wigner’ s classic text. He
dem onstr ates the i rredu cibility o f the (2j + 1)-dimensional rep resen tation
tha t |a| 2 + |b|2 = 1 mus t nec essarily be a con stant ma trix, and there fore, by
10. 9 Re presentati ons o f rot ation s and the conce pt of tens ors
ort hogon al gr oup O (3) and th eir c onnec tion to an gular mome ntum
The comp onent s of a sym metri c tra celess 2-index tenso r con tains 5
and we h ave
and a se t of 5 ind epend ent c ompon ents of a symme tric trace less tenso r.
9=1 ⊕3 ⊕5 .
oth er su bsets of 2 -index ten sors can b e fou nd th at pr eserv e the ir identities
no subsets of tens ors t hat p reserve th eir i denti ties under the defin ing
sym metri c and anti -symmetric part s is impor tant in th e Qua rk Mo del of
elementa ry pa rticles.
The repr esentations of the o rthog onal group O(3) are found to b e
par ticle of f inite mass can always be descibed in its rest fram e (all inertial
fra mes a re eq uivalent!), and ther efore the particle can be char acter ized by
rot ations. A ll known p articles have d ynamical states that can be de scribed
dimension of the i rrep is (2j + 1 ) the n the part icle spin is found t o be
pro porti onal to j. In Parti cle Physics, irreps with values of j = 0 , 1, 2,... and
with j = 1/2, 3/2, ... are f ound that corre spond to t he fu ndame ntal boson s
two dime nsional gr oup S U(2) have the s ame L ie algebra . In the case of
the grou p SU( 2), t he (2 j + 1 )-dimensional r epres entat ions are a llowed for
bot h int eger and h alf -integer va lues of j, wher eas, the r epres entat ions of
integer repre sentations, and ther efore they are (spinor) t ransfomations
The squa re of the invariant interval s, bet ween the o rigin [0, 0, 0, 0]
of a spa cetime coo rdinate sy stem and a n arb itrar y eve nt x µ = [x0, x 1, x 2,
The lower ind ices can b e rai sed u sing the m etric tens or
so that
s2 = xT x = x´T x´ .
The prim ed an d unp rimed column ma trices (contrav ariant vec tors) are
x´ = Lx .
xT x = (Lx)T (Lx)
= xTLT Lx .
LT L = .
This is the def ining pr opert y of the L orent z tra nsformations. 108
x → x´ → x´´
whe re
x´ = L1x ,
and
x´´ = L2x´.
x´´ = L2(L1x)
= L2L1x
= Lc x
whe re
(L2L1)T (L2L1) =
or
L1T(L2T L2)L1 =
so that
L1T L1 = , (L1, L2 ∈ L)
the refor e
Lc = L2L1 ∈ L .
Any numb er of succ essive Lor entz trans forma tions may be ca rried out to
d et(LT L) = det
we obtain
so that
de tL = ±1.
(LT L)–1 = –1
,
or
L–1 –1
(LT)–1 = –1
. 110
–1
Using = , a nd re arran ging, gives
L–1 (L–1)T = .
This res ult shows that the i nvers e L–1 is always a membe r of the s et L.
1. If L1 an d L2 ∈ L , then L2 L1 ∈ L
2. If L ∈ L , then L–1 ∈ L
and
The set of all Lor entz trans forma tions ther efore form s a gro up.
Spa tial rotat ions in tw o and thre e dim ensions ar e Lor entz
tra nsformations in which the time -component rema ins unchan ged.
give x´ t hen
x´Tx´ = xTRTRx 111
= invariant unde r R.
The action of R on any three -vector pr eserv es length. The set of all 3×3
dis conne cted:: one w ith d etR = +1 an d the othe r wit h det R = -1. The
and
all poin ts in a 3- dimensional car tesian sys tem t hroug h the orig in th en
-1 0 0
P = 0 -1 0 .
0 0 -1
fro m the set of tr ansformati ons w ith d etL = +1 to thos e with det L = -1; we
say that the set { L: d etL = +1} i s dis conne cted fr om th e set {L: d etL = -
1}.
L → L µν
whe re µ = 0,1,2,3 labels the r ows, and ν = 0,1,2,3 labels the c olumns the n
L 00 ≥ +1 or L 00 ≤ -1.
and
{L↓-} = {L: d etL = -1, L 00 ≤ -1}. 113
11. 4 Pa rity, time -reversal and o rthoc hrono us tr ansfo rmati ons
P = {P: r → -r, t → t}
and
T = {T: r → r, t → -t }
= di ag(-1, 1, 1, 1} .
PT
↑
L + L↓-
P T
L↑- L↓-
Fig . 3 Con necti ons b etween the disc onnec ted p arts of th e Lor entz
tra nsformatio ns
The prop er or thoch ronou s tra nsformations ar e in the g roup L↑+ . We 114
see that it is not nece ssary to c onsider th e com plete set {L} o f Lor entz
tra nsformations — we ne ed co nsider onl y tha t sub set { L↑+ } t hat f orms a
sho wn clear v iolations under the parity tra nsformation, P an d vio lations
How ever, not a single experiment has b een c arried out that show s a
violation of the p roper orth ochro nous trans forma tions, {L↑+ }.
12 115
ISO SPIN
pro perties: mass, charg e, spin, parity, and thei r ele ctric and magne tic
mom ents. In o ur on -going que st fo r an under stand ing o f the true natu re of
the fund ament al particles, and th eir intera ctions, ot her i ntrin sic proper ties,
with nam es such as “iso spin” and “stra ngene ss”, have been discovered .
The intr insic prop erties are defined b y qua ntum numbe rs; for ex ample, the
Aφ = a φ
In this chapt er, w e sha ll discuss the first of t hese new p roper ties to
The building b locks of n uclei are proto ns (positively charg ed) and
neu trons (neu tral). Nu merou s exp eriments o n the scat tering of proto ns by
pro tons, and proto ns by neut rons, have show n tha t the nuclear f orces
bet ween pairs have the same stren gth, provi ded t he an gular mome ntum
and spin stat es ar e the same . Th ese o bserv ations for m the basis of an
importan t con cept — the cha rge-i ndepe ndenc e of the n ucleo n-nuc leon
for ce. (Corr ections for the coulomb e ffects in proto n-pro ton s catte ring
its mass is a lmost equa l to the m ass of the prot on. (Recent me asure ments
give
and
the ory o f nuc lear forces in which he c onsidered the n eutro n and the proto n
intrinsic var iable, later ca lled isospin, that p ermit s the char ge st ates (+, 0 ) of
the nucleons to be distinguished. Thi s new vari able is needed (in addition
to the t raditional spac e-spin var iables) in the descr iption of nucleon-
In nuclei, pr otons and neutr ons b ehave in a rema rkably sym metri cal
way : the bind ing energy of a nucleus is closely propo rtion al to the numbe r
of neutr ons a nd pr otons , and in light nuclei (mass nu mber <40), the
num ber o f neu trons can be eq ual t o the numb er of prot ons.
obs erved spec tra o f ato ms th at co uld n ot be unde rstoo d using th e the n
num ber, is the orbit al an gular mome ntum quant um nu mber, m is the
mag netic quan tum n umber, and ms is the spin quant um nu mber.
so that, eith er
or
Let
ΨI = ψa (1)ψb(2) (1 an a, 2 in b)
and
system s pends 50% of its time in ΨI an d 50% of i ts time in ΨII . The t wo-
par ticle syst em is cons idered to be a lin ear c ombin ation of ΨI an d ΨII :
or
(The coe fficient ( 1/√2) norm alizes the sum of th e squ ares to be 1).
Exc hanging 1↔ 2 leaves Ψsym m un chang ed, w herea s exc hanging pa rticles
If two p articles are in ΨS, b oth p articles can ex ist in the same stat e with
can not e xist in th e sam e qua ntum state. El ectro ns (fermions, spin = (1/2)h)
Ψ(χ, q ) = ψ(χ)φN(q) ,
whe re
χ = χ(r, s )
label.
For two nucleons, we wr ite 119
or
= ψ4(χ1, χ2)φN(n1)φN(p2).
If we re gard the p roton and neutr on as different stat es of the same object,
pri nciple to cover the new c oordi nate: the total wave function is the n
We shall need thes e res ults in later d iscussions of t he sy mmetr ic and ant i-
Nuc lei are bo und s tates of n eutro ns an d pro tons. If the n umber of
pro tons in a nucleus is Z an d the numb er of neut rons is N then the m ass
num ber o f the nucleus is A = N + Z. S ome n uclei are natur ally unsta ble.
β-decay — a pr ocess that typi fies the f undam ental “wea k int eract ion”) .
or, we c an ha ve
A
Z XN → A
Z-1 XN-1 + e+ + νe ( β+ - decay).
e– + A
Z XN → A
Z+1 XN+1 + νe .
νe + A
Z XN → A
Z-1 XN-1 + e+
and
νe + A
Z XN → A
Z+1 XN-1 + e– .
The deca y of the fre e pr oton has n ot be en ob serve d at the p resen t tim e. 121
The expe rimental limit on th e half-life of the p roton is > 10 31 ye ars! Many
cur rent theor ies of the micr ostru cture of m atter pred ict t hat t he pr oton
dec ays. If, howev er, t he life-time is > 10 32 - 1 033 ye ars t hen t here is no
rea listic pos sibility o f observing the deca y dir ectly (The limit is set b y
The fund ament al β-decay is tha t of the f ree n eutro n, first o bserv ed in
194 6. T he pr ocess is
This mea sured life-time is o f fundamen tal import ance in Pa rticle Phy sics
and in C osmology.
tha t we have a 2-state system in which the trans forma tion p ↔ n occur s:
n → p + + e– + νe ,
p → n + e+ + νe .
φ = φp ,
+ n φ = 0, (eliminates a p roton )
+ p
and 122
φ = φn ,
- p φ = 0, (eliminates a n eutro n).
- n
1 0
φ(p) = a nd φ(n) = ,
0 1
0 1 0 0
+ = and - = .
0 0 1 0
0 1 0 1
+ n φ = = , φn → φp,
0 0 1 0
and
0 1 1 0
φ =
+ p = ( + re moves a pr oton) .
0 0 0 0
To make the p resen t alg ebraic des cription analogous t o the two- state
com binations of th e ± :
0 1
1 = + + - = = 1, a Paul i mat rix,
1 0
and
0 -i 123
2 = i( - - + )= = 2, a Paul i mat rix.
i 0
1 0
3 = = 3 , a Paul i mat rix.
0 1
relations
whe re th e factor o f(1/2) is intro duced beca use o f the 2:1 homom orphi sm
t = /2
pro jection of t on the 3rd a xis, t3. The e igenvalues, t3, o f t3 co rresp ond t o
the prot on (t 3 = +1/2) and neutr on (t 3 = -1/2) stat es. The n ucleon is said to
be an iso spin doubl et wi th isospin quan tum n umber t = 1/2. (The numb er
Q N = q(t 3 +( 1/2)) = q or 0,
whe re q is the pro ton c harge . (It is one o f the grea t uns olved prob lems of
Par ticle Phys ics to und ersta nd wh y the char ge on the proto n is equal to t he
com ponen ts
= (Z – N)/2 .
= q( T3 + A/2) .
T3 = −T, −T + 1,...0,...T + 1, T .
[H, T] = 0.
effects in nu clei was first applied to the obser ved p roper ties of th e lowest-
lying st ates in th e thr ee nu clei with mass numbe r A = 14: 14C, 14N, and 14O. 125
The rela tive energ ies of the stat es ar e sho wn in the following diagram:
6
0+ T = 1, T3 = 1
0+ T = 1 , T3 = 0
0+ T = 1 , T3 = -1 1+ T = 0 , T3 = 0
0
The spin and parity of the g round stat e of 14C, the f irst excited st ate o f 14N
cha racte rized by T = 1. The grou nd st ate o f 14N h as spin and par ity 1 + ; it
is an isospin singlet (T = 0 ).
nuc leons and nuclear isobars using the conc ept o f isospin, and the a lgebra
Yukawa (1935), first pr oposed tha t the stro ng nu clear forc e bet ween
a p air o f nuc leons is c arried by mas sive field particles called mesons.
Yukawa’s meth od wa s a m aster ful development of t he th eory of th e 126
the “mes on wa vefunction” the n the Yukawa differential equa tion for t he
mes on is
∂µ∂µ ψπ + (E0/hc)2ψπ = 0.
whe re
∇2 → (1 /r2)d/dr(r2d/dr)
The stat ic (time-indepe ndent ) solution of t his equation is read ily checked to
be
Ψ(r) = (-g2/r)exp(-r/rN)
whe re
so that
1 /rN2 = (E π0/hc)2
r = r N = h/mπc ≈ 2 ×10-13 cm .
qua ntum) . In the case of th e electrom agnet ic field, the m ass of the field
qua ntum (the photo n) is zero , and ther efore the force has an infinite ran ge.
The meso ns co me in thre e cha rge s tates: +, -, and 0. The meson s 127
hav e int rinsic spins eq ual t o zer o (th ey ar e field particles an d the refor e the y
are boso ns), and t heir rest energ ies are me asure d to be
t = 1, t3 = ±1, 0.
num ber b y I, we sh all follow this conv ention fro m now on.
The thir d com ponen t of the i sospin is an ad ditive qua ntum numbe r.
The comb ined values of the i sospin pro jections o f the two particles, one
isospin doublet.
3 ⊗ 2 = 4 ⊕ 2.
(This is the rule for f ormin g the prod uct ( 2I3(1) + 1)⊗(2I3(2) + 1).
13 128
mat ter s eemed to b e com plete. We requ ired neutr ons, proto ns, electrons
and neut rinos, and meso ns an d pho tons. Our opti mism was short- lived.
By 1953, excited s tates of t he nu cleons, and mor e mas sive meson s, had
bee n discover ed. Some of th e new part icles had completely unex pecte d
pro perties; for ex ample, in the i ntera ction betw een p roton s and π-mesons
Pr oton (p+ )
Sig ma (∑ + ) Pion
(π0 )
❊ Kao n (K+ ) ❊
Pion
(π + )
Pi on (π + )
⇑ ⇑
Initial interaction Fina l decay
lasts ~10 -23 se conds ta kes ~ 10-10 se conds
Gell-Mann, an d ind epend ently Nishijima, pro posed that the kaons (heavy
mes ons) were endow ed with a new intrin sic proper ty no t affected by t he
str ong f orce. Gel l-Mann called t his proper ty “s trang eness”. S trang eness
p + (S = 0) ∑ + (S = –1 )
π0 (S = 0)
❊ K + (S = +1 ) ❊
π+ (S = 0)
π+ (S = 0)
⇑ ⇑
∆S = 0 ∆S = 1
interact ion, they are d efined for ever. All subs equen t exp eriments involving
the se ob jects have been cons istent wit h the orig inal assignment s.
scheme t hat c lassified the s trong ly interac ting particles into family gro ups
the ory, concerned with the n ature of p ossible building blo cks o f the
nuc leons and meson s. T hey w ere c oncer ned, howev er, w ith t he inclusion
arr angem ents of pa rticles in an abs tract spac e de fined by t heir electric
cha rge a nd st range ness. The comm on fe ature of e ach family was chosen
to be th eir intrin sic spin; the f amily of s pin-1/2 baryons (str ongly 130
interact ing p articles) has e ight membe rs: n 0, p + ,∑± ,∑ 0 ,Ξ– ,Ξ0 , and Λ0 .
cha rged proto n exists, and n o pos itively-charged Ξ ex ists). If the positions
of these eight par ticles are given in charg e-strangen ess space, a re marka ble
Λ0
∑– ∑+ –1
∑ 0
+1
–2
Ξ –
Ξ 0
Cha rge –1
There ar e two part icles at t he ce nter, each with zero char ge an d zer o
str angen ess; they are t he ∑0 an d the Λ0. ( They have different rest masses).
They stu died the s truct ure o f oth er fa milies. A part icularly 131
importan t set of p articles consists of all baryo ns with sp in 3/2. A t the time ,
Ξ0, a nd Ξ-1 . They have the following patte rn in char ge-strange ness space:
–1
∑*– ∑ *0 ∑+
–2
Ξ* –
Ξ* 0
–3
Ω –
The symm etry patte rn of the family of spin-3/2 b aryon s, shown b y the
sug gest that a ten th me mber of th e fam ily should exist. F urthe rmore , if
the symm etry has a phys ical basis, the tent h mem ber s hould have spin-3/2,
cha rge – 1, st range ness –3, a nd it s mas s sho uld b e abo ut 15 0MeV great er
tha n the mass of t he Ξ0 pa rticle. T wo ye ars a fter this suggestion, the tent h
mem ber o f the family was identified in high ener gy pa rticle collisions; it
dec ayed via weak intera ctions, and pos sessed the pred icted prop erties.
mod el was used to obtain sym metri es of meso ns, also groupe d by their
spins.
str angen ess”. Alt hough the model was shown not to be valid, Ikeda e t al.
system t hat i nvolved th e gro up SU (3). The notion tha t an under lying
ear ly 19 60’s. (Sp ecial Unit ary G roups were used by J . P. Elliott in the
The prob lem facing Part icle Physicists, at the t ime, was t o find the
app ropri ate g roup and its fu ndame ntal repre sentation, and to co nstru ct
higher-d imensional repr esentations tha t wou ld account for the w ide variety
has its origin in the c harge -independe nce o f the nucleon-nucleon for ce.
When app ropri ate, we sh all discuss the symm etry prope rties of p articles in
a g roup chara cterized b y its thre e gen erato rs, t he Pa uli spin m atrices.
Two-state systems, such as t he electro n wit h its quan tized spin-up a nd sp in-
dow n, an d the isospin states of n ucleons an d nuc lei, can b e tre ated
qua ntitatively using th is gr oup. The symme tries of n ucleon and meso n
end owed not o nly w ith p ositive, n egative, a nd ze ro ch arge but a lso with
str angen ess. A th ree-state system was ther efore cons idered nec essary, at
the very least; th e mos t pro mising can didate was the group SU(3 ). We
shall discuss the infinitesimal form o f this gro up, a nd we shall find a
The quan tities δαj ar e rea l and infinitesimal, and the 3 ×3 m atrices j ar e
the linearly independen t gen erato rs of the group . Th e rep eated index, j,
U = exp{iαj j/2}.
1 0 0
u = 0 , v = 1 , an d w = 0 .
0 0 1
0 1 0 0 1
A+ v = u, 0 0 0 1 = 0 .
0 0 0 0 0
We can introd uce o ther opera tors that trans form the s tates in p airs, thus 135
0 0 0
A– = 1 0 0 ,
0 0 0
0 0 0 0 0 0
B+ = 0 0 1 , B– = 0 0 0 ,
0 0 0 0 1 0
0 0 0 0 0 1
C+ = 0 0 0 , C– = 0 0 0 .
1 0 0 0 0 0
and the group SU(2 ), th e non -singular, trac eless, her mitian mat rices 1, a nd
1 = + + – = 1 an d 2 = i( 1 – 2) = 2.
The gene rator s of SU(3) are forme d fro m the oper ators A±, B±, C± by
0
1
0 = A+ + A– ≡ 1, a gene rator of S U(3).
0 0 0
whe re
0
2
2 = 0 ,
0 0 0
and
C+ + C– = 4 , C+ – C– = –i 5,
B+ + B– = 6 a nd B + – B– = i 7 .
The rema ining gene rator s, 3 an d 8 ar e tra celess, diagona l, 3×3 m atrices:
0 1 0 0
3 = 3 0 , 8 = 0 1 0 .
0 0 0 0 0 -2
Tr ( j k ) = 2δjk.
1 0 0
8 = (1/ √3) 0 1 0 .
0 0 –2
If we pu t F i = /2. we f ind
i
A± = F 1 ± iF 2 ,
B± = F 6 ± iF 7,
and 137
C± = F 4 + iF 5 .
A3 + B3 + C3 = 0.
The last cond ition mean s tha t onl y eight of the nine opera tors are
independ ent.
The gene rator s of the g roup are r eadily sho wn to obey the Lie
[F i, F j] = ifijk F k , i,j,k = 1 to 8.
whe re th e qua ntities fijk ar e the non- zero struc ture const ants of th e gro up;
fijk = –fjik ,
The comm utation re lations [F i, F j] c an be writ ten i n ter ms of the opera tors
[F 3, F 8] = 0 .
each wit h zer os in the third row and c olumn; the y obe y the comm utation
Par ticles tha t exp erience th e str ong n uclear int eract ion are ca lled
had rons; t hey a re se parat ed into tw o set s: the bar yons, w ith h alf-intege r
spins, and th e mes ons w ith z ero o r int eger spins. Pa rticles th at do not
exp erience th e str ong i ntera ction are called lep tons. In or der t o qua ntify
–1, and all other particles are a ssigned L = 0. It i s a p resen t-day fact ,
based up on nu merou s obs ervat ions, that the total bary on an d lepton
num ber i n any inte raction is cons erved . Fo r exa mple, in t he de cay o f the
n 0 = p+ + e– + νe 0
B = +1 = +1 + 0 + 0
L = 0 = 0 + 1 + (–1) .
The fund ament al symmetr ies in Nat ure r esponsible for these cons ervat ion
laws are not known at t his time. These con serva tion laws may, in all
In discussing the patte rns o f bar yon f amilies in char ge-strange ness
space, w e wish to incorporat e the fact that we a re de aling with bar yons
tha t int eract via the s trong nuclear f orce in wh ich iso spin an d strangen ess
are cons erved . We ther efore choo se to desc ribe their patt erns in isospin- 139
hyp ercha rge s pace, wher e the hype rchar ge Y is defined to i nclude bot h the
str angen ess and th e bar yon a ttrib ute o f the part icle in an additive way:
Y = B + S.
The diagonal opera tor F 8 i s the refor e assumed to be dire ctly associated
F 8 = (√3/2)Y.
ope rator s com mute with I3 an d Y, n o oth er ad ditive qua ntum numbe rs ar e
we found that the commu tation rel ations of the g enera tors of th e gro up
wer e sat isfied not only by t he fu ndame ntal 2×2 m atrices bu t also by
mat rices of h igher dime nsion [(2J + 1) ⊗ (2 J + 1 )], w here J can have the
values 1 /2, 1, 3/2, 2, ....The J-values cor respo nd to the spin of th e par ticle
tra nsformation pro perties). In t he 2×2 r epres entat ion, both cov arian t an d
´ = Uij
i j ,
whe re
a1
= ,
a2
and
ii) con trava riant spinors (with uppe r ind ices) are writt en as
whe re
= (b1, b 2).
The co- and c ontra -variant s pinor s are tran sformed with th e aid of t he an ti-
ij
sym metri c ten sors ij an d . For e xample,
j
i = ij
b2
i = .
- b1
The high er-dimensional repre sentations are built up f rom t he fu ndame ntal
j
for m by taking ten sor p roduc ts of the fundamenta l spinors i , , or i
whe n a t ensor prod uct o f spi nors has b een b roken down into its symmetric
and anti -symm etric part s, it has been decom posed into irre ducib le 141
ori ginal discussion of the m ethod , and de S wart in Rev. Mo d. Ph ys. 35,
(1963) for a detailed discussion of te nsor analysis in the stud y of the i rreps
of SU(n))
i⊗ j= i j = ( i j + )/2 + (
j i i j - )/2
j i
There ar e fou r ele ments associated wit h the prod uct ( i,j can ha ve va lues 1
and 2).
and tran sforms as an ob ject that has spin J=1. (There are 2J + 1 me mbers
of the s ymmet ric set). The anti-symme tric part has o ne element , and
the refor e tra nsforms as an o bject with spin J = 0. T his r esult is familiar in
the theo ry of angu lar m oment um in Quan tum M echanics. The explicit
J3 = +1: 1 1
J=1 J3 = 0 : (1/√2 )( 1 2 + 2 1 )
J 3 = –1 : 2 1
and
pro ducts are then writt en in term s of their symm etric and anti-symme tric
par ts in orde r to obtain the irre ducible re prese ntations. For examp le, the
j
pro duct i , i,j = 1,2,3, can be w ritte n
j j
i = ( i - (1/ 3)δij k
k
) + (1/3)δij k
k
,
in which the trace has been separ ated out. The trace is a zero -rank tens or
with a sin gle co mpone nt. The o ther tenso r is a tra celess, symmetr ic tensor
with eig ht in depen dent compo nents . Th e dec ompos ition is w ritte n
3 ⊗ 3 = 8 ⊕ 1.
follows:
3⊗3 = 6⊕3,
in which the symme tric tenso r has six compo nents and the a nti-symmet ric
3 ⊗ 3 ⊗ 3 = 10 ⊕ 8 ⊕ 8 ⊕ 1 ,
and
8 ⊗ 8 = 27 ⊕ 10 ⊕ 10 ⊕ 8 ⊕ 8´ ⊕ 1 .
The appe aranc e of the o ctet “8” i n the 3 ⊗ 3 de compo sition (re call 143
the obse rved octet of s pin-1/2 baryons ), and the decu plet “10” in th e tri ple
pro duct 3 ⊗ 3 ⊗ 3 de compo sition (re call the o bserv ed de cuplet of spin-3/2
an d
8 u = βuu, 8v = βvv, a nd 8w = β ww ,
a 0 0 b 0 0
N N
3 = 0 – a 0 , an d 8 = 0 b 0 .
0 0 0 0 0 –2 b
If
and
N N
3 w = 0w , 8 w = –2bw.
space, t hus:
β
2b
–2a –a a 2a α
−b
–2 b
This weight d iagram for the fundamenta l “3” repr esentation of S U(3) was
Although the octet and decup let p atter ns of hadr ons o f a g iven spin
major pr oblems rem ained that resu lted in a great deal of s cepticism
with the fund ament al tr iplets 3, 3 of SU(3 ) tha t exh ibit the s ymmet ry of
independ ently, Zwe ig, propos ed th at th ree f undam ental enti ties do ex ist
tha t cor respo nd to the base states of SU(3), and that they form boun d
sta tes o f the hadr ons. That such enti ties have not b een o bserv ed in the
fre e sta te is simply re lated to t heir enorm ous b inding ene rgy. The three
ent ities were called qua rks by Gell-Mann, and ace s by Zwei g. T he Ge ll-
Mann ter m has surv ived. The anti -quar ks ar e associated with th e
and d fo r the up- and d own-isospin sta tes, and s for stran geness) have
Lab el B Y I I3 Q= I3 +Y /2 S = Y − B
s –1 /3 2/3 0 0 + 1/3 +1
d –1 /3 –1/3 1 /2 +1 /2 + 1/3 0
I3
I3
u d
s
These diagram s hav e the same rela tive forms as t he 3 an d 3 we ight
diagrams of S U(3).
up of th e simplest poss ible struc tures , nam ely quark– anti-quark pair s. T he
cov ariant and cont ravar iant 3-spinors intro duced in t he pr evious section
and
The form al operato rs A±, B±, a nd C±, introd uced in section 13. 3.1,
are now viewed as opera tors that trans form one fla vor (t ype)of qua rk in to
A± ≡ I±(I3) → I3 ± 1 ,
and
C± ≡ V±(V3) → V3 ± 1, ca lled the V-spin oper ator. 147
I+ (–1/2) → 1 /2 : d → u
I–(+1 /2) → –1 /2 : u → d
U+ (–1/2) → 1/2 : s → d
U–(+1 /2) → –1 /2 : d → s
V+ (–1/2) → 1/2 : u → s
and
The quar ks ca n be chara cterized b y the thre e qua ntum numbe rs I3, U 3, V3.
Their po sitions in the I3-U3-V3 - space again sho w the unde rlying
sym metry :
U3 -V 3
+1/2
-I 3 - 1/2 + 1/2 I3
+1/ 2
s( 0, -1/2, 1/2)
V3 -Y -U 3
The memb ers o f the octe t of meson s with JP = 0– ar e for med f rom q q-pairs 148
tha t bel ong t o the fund ament al 3, 3 re prese ntation of the quark s. T he π0
and η0 me sons are l inear comb inations o f the q q(bar) s tates, thu s
K0 d s Y K+ us
d u
π– d u π0 π+ ud
-1 η0 +1 I3
u d
s
K– s u K0 s d
The none t for med f rom t he te nsor produ ct 3 ⊗ 3 i s split into an octe t
tha t is even under the label exch ange of tw o par ticles, and a s inglet tha t is
3⊗3 = 8⊕1
whe re th e “1” is
→ π0
Omitting the tenso r pro duct sign, norm alizing th e amp litudes, and ch oosing
oct et wo uld h ave t he sa me ma ss. This is not qui te th e case; th e sym metry
mas ses of the quar ks, d erived fro m the mass differenc es of hadr on-pairs, is
not the same as th e “cu rrent -quar k” ma sses that appear in the
fundamen tal t heory . Th e discrepa ncy b etween the effective mass es an d the
3 ⊗ 3 ⊗ 3 = (6 ⊕ 3) ⊗ 3
= 10 ⊕ 8 ⊕ 8´ ⊕ 1
in which the states of the 10 ar e sym metri c, th e 1 is anti symme tric, and the
dec ompos ition is a ssociated with the o bserv ed de cuplet of spin-3/2 b aryon s.
following diagram:
d dd ~ du d ~ u ud uuu
~ sds ~ sus
sss
The prec ise make-up of each state, labelled by ( Y, I, I3,) is given in the
The gene ral t heory of t he pe rmuta tion group of n enti ties, and its
rep resen tations, is out side the s cope of th is introdu ction. Th e use of t he
u +2/3
d –1/3
s –1/3
The char ges o f the anti -quar ks ar e opp osite in s ign t o the se va lues.
Ext ensive rev iews of th e 3-q uark model and its application to t he
Gell-Mann and Zwei g, it was recog nized that the model was not c onsistent
with the exte nded Pauli prin ciple when applied t o bou nd st ates of th ree
qua rks. For examp le, the st ructu re of the spin-3/2 ∆ + st ate is such tha t, if
each qua rk is assigned a spin sq = 1/2, the t hree spins must be a ligned ↑↑↑
to give a net spin of 3 /2. (It is assumed that the r elative or bital angu lar
rea sonable assumpt ion t o mak e, as it c orres ponds to m inimum kin etic
ene rgy, and t herefore t o a s tate of lowest total ener gy). The quark s are
fer mions, and ther efore they must obey the gener alized Pau li Principle;
the y can not e xist in a completely aligned s pin state when they are i n an s-
sta te th at is symm etric unde r par ticle (qua rk) e xchan ge. The uni tary sp in
later ca lled “colo r”, a prop erty with profo und c onsequences. A quar k
with a c ertain flavor p ossesses color (red, blue, gre en, say) t hat
cor respo nds t o the trip let r epres entat ion o f ano ther form of SU (3) —
nam ely S U(3)C, w here the s ubscr ipt C different iates the group from that
qua rks ( that possess anti-color) have a tri plet repre sentation in SU (3)C th at
is the c onjugate r epres entat ion (the 3). Alth ough the S U(3)F sy mmetr y is
the color sin glet of a baryo n occ urs i n the deco mposition
S U(3)C = 3 ⊗ 3 ⊗ 3 = 10 + 8 + 8´ + 1 .
1 = (RR + BB + GG)/√3 .
Although the hadro ns ar e colorless, certain obse rvable qua ntities are
dir ectly rela ted t o the numb er of colors in the model. Fo r exa mple, the
pur ely electromagn etic decay of t he ne utral pion , π0, into t wo ph otons
π0 = γ + γ,
has a lifetime tha t is found to b e closely propo rtion l to the squ are o f the
num ber o f col ors. (A dler (1970) gives Γ = h/τ = 1(eV) (no.of co lors)2 .
The meas ureme nts o f the lifetime give a value of Γ ~8 eV, consistent with
Ncol s = 3. S ince these earl y mea surem ents, refined e xperiments have
dem onstr ated that there are three , and only thre e, co lors associated with 153
bet ween the t heore tical mode l and the measured v alue of R, over a wi de
The color att ribut e of the q uarks has been respo nsible for the
qua ntum chrom odyna mics. It is a field theo ry in which the quar ks ar e
field ar e called glu ons; t hey p ossess col or, t he at tribu te of the sourc e of the
field. Consequent ly, they c an interac t wit h eac h oth er th rough the color
the sour ce of the field, nam ely electric charge. The phot ons, there fore, do
different col or. For e xample, in the interaction bet ween a red quar k and a
blue qua rk, t he co lors are e xchan ged. This requ ires that the e xchan ged
and gluo ns. Three of t hese are r ed-red, blue-blue, a nd gr een-green that do
sym metri c in the c olor labels, and thi s com bination is the singlet state of
the grou p SU( 3)C. Eight gluo ns, each w ith t wo co lor indices, ar e the refor e
res onanc e at a cen ter-o f-mass ene rgy o f 3.1 GeV. The lifetime of th e
longer t han e xpect ed fo r a s tate forme d in the s trong inte raction. The
res onant stat e is called the J/ψ. It wa s quickly realized t hat t he st ate
bou nd st ate c c, wh ere c is a four th, m assive, qu ark e ndowe d wit h one unit
of a new quan tum n umber c, c alled “cha rm”. The quant um nu mbers
Sou nd th eoret ical argum ents for a four th qu ark, carry ing a new
qua ntum numbe r, ha d bee n put forw ard s evera l years be fore the
sta tes o f the “cha rmoni um” s ystem has been obser ved, and t heir decay
pro perties st udied. De tailed com parisons h ave b een m ade w ith
The inclusion of a char med q uark in th e set of q uarks mean s tha t the
app lications have been made using the model. Th e fun damen tal
an d
dec ompos itions of the v arious ten sor p roduc ts. The m ost u seful are:
4 ⊗ 4 = 15 ⊕ 1,
4 ⊗ 4 = 10 ⊕ 6,
an d
15 ⊗ 15 = 1 ⊕ 15sym ⊕ 15ant i ⊕ 20sym ⊕ 45 ⊕ 45 ⊕ 84. 156
The “15” includes the n on-charmed (JP = 0– ) m esons and the f ollowing
In order to d iscuss the bary ons, it is nece ssary to i nclude the quar k spin,
and ther efore the group must be e xtend ed to SU(8 )F. Relatively few
bar yons have been studied in deta il in this exte nded framework.
9.4 , 10. 01, a nd 10 .4 Ge V, and wer e int erpre ted a s bou nd st ates of an other
qua rk, t he “b ottom ” qua rk, b , and its anti-partn er, t he b. Mes ons c an be
The stud y of the w eak d ecay modes of t hese states is curre ntly fashionable.
qua rk, c alled the “top” quar k, t. It is a massive en tity with a mas s almost
We have seen that the q uarks inte ract stron gly v ia gluon e xchan ge.
They also tak e par t in the w eak intera ction. In an e arlier dis cussion of
isospin, the group gene rator s wer e int roduc ed by cons idering th e β-decay
n 0 → p + + e– + ν0 .
We now k now t hat, at th e microsco pic level, this proc ess involves th e 157
tra nsformation of a d-q uark into a u-q uark, and the p roduc tion of th e
ν0
– 1
W e–
d u
neu tron, n 0 d(– 1/3) → u( +2/3) prot on, p +
u u
d d
The carr iers of th e Weak For ce, W± , Z 0, were first identified in p-p
collisions at high cent er-of-mass ener gy. The p roces ses involve
qua rk–an ti-quark i ntera ctions, and the dete ction of t he de cay e lectrons a nd
pos itron s.
e+ e–
0
Z
u( +2/3) u ( –2/3)
ν 0
W+ e+
u( +2/3) d (+1/3)
ν 0
W– e–
d( −1/3) u (−2/3)
(R ecall that the range of a forc e ∝ 1/ (mass of c arrier); the W and Z mass es
tac kle Grand Unified Th eories, re quires a k nowledge o f Qua ntum Field
Theory t hat i s not expe cted of re aders of t his introd uctor y boo k.
14 159
and
whe re p i is the (canonical) mom entum and qi is a (g enera lized) coo rdinate,
the n
n
{ , H} = ∑1 (∂ /∂q i)(∂ H/∂p i) – (∂ /∂p i)(∂ H/∂q i) .
of the m otion .
(d /dt) = { , H}
is replaced b y
If the D irac Brack et is zero, the quan tum m echan ical quant ity is
suitable choice of the ord er of the q’s a nd p’ s in the P oisson Bra cket).
14. 2 In finit esima l uni tary trans forma tions in Q uantu m Mec hanic s
U = I + iδαX/h ,
whe re δα ia real infinitesimal param eter, and X is an h ermit ian operat or.
tra nsformation:
→ ´ = U U–1
= (I + iδaX/h) (I – iδaX/h)
= + i(δaX – δaX)/h
= + i(F – F)/h.
whe re
F = δaX.
δ = ´–
= i[F, ]/h
δ = i[−Hδt, ]/h,
or
–δ = i[H, ]δt/h ,
so that
d /dt = i[ , ]/h
is therefore deduc ed fr om th e uni tary infin itesi mal t ransformat ion o f the
ope rator .
This app roach was taken by S chwinger i n his form ulation of Quan tum
Mechanics.
|F| = Hδt i s dir ectly rela ted t o the gene rator , X, o f a Q uantu m
with eve ry sy mmetr y tra nsformation of the s ystem an h ermit ian operat or
equ ation s of motio n. I f the equa tions of m otion are uncha nged by th e
con stant as t he sy stem evolves with ti me. As a well-known exam ple, if the
equ ations of motion of an ob ject are i nvariant u nder trans lations is spac e,
BIBLIOGRAPHY
The following book s are typ ical of th ose t hat a re su itable for
Arm stron g, M. A., Gro ups a nd Sy mmetry, S pring er-Verlag, New York,
198 8.
Bur ns, G erald, Int roduction to Gr oup T heory, A cadem ic Pr ess, New York,
197 7.
Fri tzsch, Har ald, Qua rks: the S tuff of Ma tter, B asic Books, New York ,
198 3.
Jon es, H. F., Gro ups, Repre senta tions and Physi cs, A dam H ilger, Bristol,
199 0.
Car ter, Roger ; Seg al, Graeme ; and Macd onald, Ian, Lec tures on L ie
Gro ups a nd Li e Alg ebras, C ambridge U niversity Press , Cam bridg e, 19 95.
Com mins, E. D ., and Buc ksbaum, P. H., Wea k Int eractions of Le ptons and
Eisenhar t, L. P., Con tinuo us Gr oups of Tr ansfo rmati ons, D over, New
York, 19 61.
Gibson, W. M., and Pollard, B. R., Sym metry Prin ciple s in Eleme ntary
Par ticle Phys ics, C ambridge U niversity Press , Cam bridg e, 19 76.
Lichtenb erg, D. B., Uni tary Symmetry a nd El ement ary P artic les,
Lip kin, Harry J., Lie Grou ps fo r Ped estrians, N orth- Holland, A mster dam,
196 6.
Lom ont, J. S., Ap plica tions of F inite Grou ps, D over, New York, 1993 .
196 1.
Mec hanic s of Atomi c Spe ctra, A cadem ic Pr ess, New York, 1959.