Branch point and cut line
Introduction
The comples variable z completes a phase when taken from zero to 2π. While
cpmpleting a phase z remains single valued that means
z = reiθ = reiθ+2πn = reθ ei2πn (1)
Consider a function of this complex variable f (z), for proper mapping be-
tween the z plane and f (z) just like z is single valued over a complete phase,
the function f (z) also has to be. Let’s take an example f (z) = z 2
z 2 = (reiθ+2πn )2 ≡ r2 ei(θ1 +θ2 ) ei2πn (2)
The above equation also leads to single valuedness, as n√is an integer with
n = 0, ±1, ±2, .... Now consider another function f (z) = z
1 1
h i 1 √ iθ
iθ+2πn 2
z =r e
2 2 = re 2 eiπn (3)
This function will have different values depending on whether n is odd or
even. This situation is called branch singularity and different values of f (z)
√ iθ
(here ± re 2 ) is called different branches of f (z). To √ avoid this branch
point singularity one at zero and other at infinity for z (explained later)
we can draw a cut-line starting from zero extending up to infinity.Let’s as-
sume this cut-line is drawn along positive real axis, though not necessary
always, this cut-line is like a great wall that cannot be penetrated. So when
we traverse from θ = 0 we won’t cross the θ = 2π phase. Hence the function
retains its single-valuedness. Branch cut is like a collection of single
valued surfaces glued together, rather than one muti-valued sur-
face.. (Comment:An alternative view is the idea of Riemann surfaces which
will be added later )
Branch point at infinity
√
Continuing with the example of z, to check the nature of the function at
1
z = ∞, put z = and check the function f ( z1 ) at t = 0.
t
1
√ 1 1 1 θ
z −→ √ = √ = √ e−i 2 e−iπn (4)
t reiθ+2πn r
Thus f ( 1t ) has a branch at t = 0, thus f (z) will be having a branch at z = ∞
Some examples
√
Example:-(1) f (z) = z2 − 1
√ 1 1
z 2 − 1 = (z + 1) 2 (z − 1) 2 (5)
z − 1 = r1 ei(θ1 +2πn1 ) and z + 1 = r2 ei(θ2 +2πn2 ) (6)
√ (θ1 +θ2 )
f (z) = r1 r2 ei 2 eiπn (7)
Thus f (z) has branch singualrity at z = ±1.
Now check for infinity.
1
1 12
f (z) = (z 2 − 1) 2 = z 1 − 2 (8)
z
1 For large z i.e z → ∞, the behaviour of this function will be decided
1
by term z outside the parantheses, as 2 −→ 0 for z −→ ∞.
z
2 Then check the nature of z at infinity. We know that z is not having
any branch at infinty.
√
3 So we can conclude that z 2 − 1 have branch points only at z = ±1
but not at z = ∞
p p
Example:-(2) f (z) = z(z 2 − 1) = z(z − 1)(z + 1)
z = r1 ei(θ1 +2πn1 ) (9)
z − 1 = r2 ei(θ2 +2πn2 ) (10)
z + 1 = r3 ei(θ3 +2πn3 ) (11)
1 (θ1 +θ2 +θ3 ) 1
f (z) = (r1 r2 r3 ) 2 ei 2 (ei2πn ) 2 (12)
2
Thus phase of the function varies as eiπn . Hence branch points at z = 0, ±1.
Now check at infinity.
p
2
3
1 12
f (z) = z(z − 1) = z 1 − 2
2 (13)
z
3
For large z f (z) −→ z 2 (14)
3 1 3
z 2 −→ 3 = t− 2 (15)
t2
put t = rei(θ+2πn) (16)
− 32 − 23 −i 3θ −i3πn
t =r e e
2 (17)
3
Thus the function t− 2 has branch point at t = 0 which shows that f (z) has
a branch point at z = ∞
Working rule for branch points
Function f (z) = (z − a)m (z − b)n will have branch point at
1 at z = a if m 6= integer
2 at z = b if n 6= integer
3 at z = ∞ if m + n 6= integer
Reference
1 https://math.mit.edu/classes/18.305/Notes/n00Branch_Points_
B_Cuts.pdf
2 https://web.njit.edu/~matveev/Courses/M656_S14/Complex1.pdf
3 http://math_old.pugetsound.edu/~martinj/courses/spring2008/
m352/CubeRootBranch.pdf
4 http://www1.phys.vt.edu/~ersharpe/spec-fn/ancont2.pdf
