Funciones elementales

Credit: This notes are 100% from chapter 4 of the book entitled A First Course in Complex
Analysis with Applications by Dennis G. Zill and Patrick D. Shanahan. Jones and Bartlett
Publishers. 2003.

In this chapter we shall define and study a number of elementary complex analytic functions.
In particular, we will investigate the complex exponential, logarithmic, power, trigonometric,
hyperbolic, inverse trigonometric, and inverse hyperbolic functions. All of these functions will
be shown to be analytic in a suitable domain and their derivatives will be found to agree with
their real counterparts. We will also examine how these functions act as mappings of the
complex plane.

Exponential and Logarithmic Functions

Complex exponential function
It is defined as
ez = ex cos y + iex sin y (1)

Theorem (4.1): Analyticity of ez The exponential function ez is entire and its derivative
is given by:
d z
e = ez (2)
dz

Modulus, Argument, and Conjugate Modulus and argument: From w = ez = ex cos y +

iex sin y = r(cos θ + i sin θ) we have

|ez | = ex > 0 ⇒ ez 6= 0 for all complex z (3)

arg(ez ) = y + 2nπ, n = 0, ±1, ±2, · · · (4)

Conjugate:

(ez ) = ex cos y − iex sin y = ex cos(−y) + iex sin(−y) = ex−iy = ez̄ (5)

Theorem (4.2): Algebraic Properties If z1 and z2 are complex numbers, then

(i) e0 = 1
(ii) ez1 ez2 = ez1 +z2
(iii) eez21 = ez1 −z2
z

(iv) (ez1 )n = enz1 , n = 0, ±1, · · ·

1
 Figure 1: The fundamental region ez (from the book)

Periodicity The most striking difference between the real and complex exponential functions
is the periodicity of ez . Analogous to real periodic functions, we say that a complex function f
is periodic with period T if f (z + T ) = f (z) for all complex z.
The complex exponential function ez is periodic with a pure imaginary period 2πi
ez+2πi = ez (6)
ez+2nπi = ez (7)
for n = 0, ±1, · · · . Thus, the complex exponential function is not one-to-one, and all values ez
are assumed in any infinite horizontal strip of width 2π in the z-plane. That is, all values of
the function ez are assumed in the set −∞ < x < ∞, y0 < y ≤ y0 + 2π, where y0 is a real
constant. In Figure 1 we divide the complex plane into horizontal strips obtained by setting y0
equal to any odd multiple of π. If the point z is in the infinite horizontal strip −∞ < x < ∞,
−π < y ≤ π, shown in color in Fig. 1, then the values f (z) = ez , f (z + 2πi) = ez+2πi ,
f (z − 2πi) = ez−2πi , and so on are the same. The infinite horizontal strip defined by:
−∞ < x < ∞, −π < y ≤ π (8)
is called the fundamental region of the complex exponential function.

The Exponential Mapping

Because all values of the complex exponential function ez are assumed in the fundamental
region, the image of this region under the mapping w = ez is the same as the image of the
entire complex plane.
In order to determine the image of the fundamental region under w = ez , we note that this
region consists of the collection of vertical line segments z(t) = a + it, −π < t ≤ π, where
a is any real number. Then, w(t) = ez(t) = ea+it = ea eit , −π < t ≤ π, where w(t) defines a
circle centered at the origin with radius ea . Because a can be any real number, the radius ea
of this circle can be any nonzero positive real number. Thus, the image of the fundamental
region under the exponential mapping consists of the collection of all circles centered at the
origin with nonzero radius. In other words, the image of the fundamental region −∞ < x < ∞,
−π < y ≤ π, under w = ez is the set of all complex w with w 6= 0, or, equivalently, the set
|w| > 0 (the point w = 0 is not in the range of the complex exponential function). See Fig. 2

2
 Figure 2: The image of the fundamental region under w = ez (from the book)

There was nothing particularly special about using vertical line segments to determine the
image of the fundamental region under w = ez . The image can also be found in the same
manner by using, say, horizontal lines in the fundamental region.

Exponential Mapping Properties .

(i) w = ez maps the fundamental region −∞ < x < ∞, −π < y ≤ π, onto the set |w| > 0.
(ii) w = ez maps the vertical line segment x = a, −π < y ≤ π, onto the circle |w| = ea .
(iii) w = ez maps the horizontal line y = b, −∞ < x < ∞, onto the ray arg(w) = b.

Complex Logarithmic Function

In complex analysis, the complex exponential function ez is not a one-to-one function on its
domain C. To see why the equation ew = z has infinitely many solutions, in general, suppose
that w = u + iv is a solution of ew = z. Then we must have |ew | = |z| and arg(ew ) = arg(z). It
follows that eu = |z| and v = arg(z), or, equivalently, u = ln |z| and v = arg(z). Therefore, if
ew = z ⇒ w = ln |z| + iarg(z). (9)

Because there are infinitely many arguments of z, it gives infinitely many solutions w to the
equation ew = z. The set of values given above defines a multiple-valued function w = G(z)
which is called the complex logarithm of z and denoted by ln z, that is,

ln z = ln |z| + iarg(z) (10)

By switching to exponential notation z = reiθ we obtain the following alternative description
of the complex logarithm:
ln z = ln r + i(θ + 2nπ) (11)
with n = 0, ±1, ±2, · · · .
The complex logarithm can be used to find all solutions to the exponential equation ew = z
when z is a nonzero complex number.

3
Examples: Find all complex solutions to each of the following equations:

1. ew = i

2. ew = 1 + i

3. ew = −2

Solution: For each equation ew = z, the set of solutions is given by w = ln z:

1. For z = i, we have |z| = 1 and arg(z) = π/2 + 2nπ. Thus, w = ln i = ln 1 + i(π/2 + 2nπ) =
(4n+1)π
2
i, with n = 0, ±1, ±2, · · · .
√
2. √ For z = 1 + i, we have |z| = 2 and arg(z) = π/4 + 2nπ. Thus, w = ln(1 + i) =
ln 2 + i(π/4 + 2nπ), with n = 0, ±1, ±2, · · · .

3. For z = −2, we have |z| = 2 and arg(z) = π + 2nπ, thus w = ln(−2) = ln 2 + i(π + 2nπ),
with n = 0, ±1, ±2, · · · .

Theorem (4.3): Algebraic Properties of ln z If z1 and z2 are nonzero complex numbers

and n is an integer, then
(i) ln(z1 z2) = ln z1 + ln z2
(ii) ln zz21 = ln z1 − ln z2
(iii) ln z1n = n ln z1

Principal Value of a Complex Logarithm The complex function Lnz defined by:

Lnz = ln |z| + iArg(z) (12)

= ln r + iθ , −π < θ ≤ π (13)

is called the principal value of the complex logarithm.

It is important to note that the identities for the complex logarithm in Theorem 4.3 are
not necessarily satisfied by the principal value of the complex logarithm. For example, it is not
true that Ln(z1 z2 ) = Lnz1 + Lnz2 for all complex numbers z1 and z2 (although it may be true
for some complex numbers).

Lnz as an Inverse Function Because Lnz is one of the values of the complex logarithm
ln z, it follows for z 6= 0 that, eLnz = z. This suggests that the logarithmic function Lnz is an
inverse function of exponential function ez . Because the complex exponential function is not
one-to-one on its domain, this statement is not completely accurate. The exponential function
must first be restricted to the fundamental region on which it is one-to-one in order to have a
well-defined inverse function, that is,

eLnz = z = x + iy if − ∞ < x < ∞ , −π < y ≤ π (14)

Lnz as an Inverse Function of ez If the complex exponential function f (z) = ez is defined

on the fundamental region −∞ < x < ∞, −π < y ≤ π, then f is one-to-one and the inverse
function of f is the principal value of the complex logarithm f −1 (z) = Lnz.

4
Analyticity The principal value of the complex logarithm Lnz is discontinuous at the point
z = 0 since this function is not defined there. This function also turns out to be discontinuous
at every point on the negative real axis. This is intuitively clear since the value of Lnz a point
z near the negative x-axis in the second quadrant has imaginary part close to π, whereas the
value of a nearby point in the third quadrant has imaginary part close to −π. The function Lnz
is, however, continuous on the set consisting of the complex plane excluding the nonpositive
real axis. Therefore Lnz is a continuous function on the domain |z| > 0, −π < arg(z) < π.
Put another way, the function f1 (the principal branch of ln z) defined by f1 (z) = ln r + iθ
is continuous on the domain |z| > 0, −π < arg(z) < π for f1 where r = |z| and θ = arg(z).
The nonpositive real axis is the branch cut and z = 0 is a branch point. The branch f1 is
an analytic function on its domain.

Theorem (4.4): Analyticity of the Principal Branch of ln z The principal branch f1

of the complex logarithm defined by f1 (z) = ln r + iθ is an analytic function and its derivative
is given by:
1
f1′ (z) = (15)
z
Because f1 (z) = Lnz for each point z in the domain |z| > 0, −π < arg(z) < π, it follows
from Theorem 4.4 that Lnz is differentiable in this domain, and that its derivative is given by
f1′ . That is, |z| > 0, −π < arg(z) < π then:
d 1
Ln(z) = (16)
dz z

Logarithmic Mapping The complex logarithmic mapping w = Lnz can be understood in

terms of the exponential mapping w = ez since these functions are inverses of each other. The
following summarizes some of these properties.
(i) w = Lnz maps the set |z| > 0 onto the region −∞ < u < ∞, −π < v ≤ π.
(ii) w = Lnz maps the circle |z| = r onto the vertical line segment u = ln r, −π < v ≤ π.
(iii) w = Lnz maps the ray arg(z) = θ onto the horizontal line v = θ, −∞ < u < ∞.

Complex Powers
If α is a complex number and z 6= 0, then the complex power z α is defined to be:

z α = eα ln z (17)

In general, eα ln z gives an infinite set of values because the complex logarithm ln z is multiple-
valued. When n is an integer, however, the expression it is single-valued (in agreement with
fact that z n is a function when n is an integer).

Examples: Find the values of the given complex power:

(a) i2i
(b) (1 + i)i
Solution:
(a) Since ln i = (4n+1)π
2
i then
i2i = e2i ln i = e−(4n+1)π (18)
(8n+1)π
(b) Since ln(1 + i) = 21 ln 2 + 4
i then

(1 + i)i = ei ln(1+i) = e−(8n+1)π/4+i(ln 2)/2 (19)

5
Properties: .
(i) z α1 z α2 = z α1 +α2
(ii) zz α12 = z α1 −α2
α

(iii) (z α )n = z nα

Principal Value of a Complex Power If α is a complex number and z 6= 0, then the

function defined by:
z α = eαLnz (20)
is called the principal value of the complex power z α .

Analyticity In general, the principal value of a complex power z α = eαLnz is not a continuous
function on the complex plane because the function Lnz is not continuous on the complex plane.
However, since the function eαz is continuous on the entire complex plane, and since the function
Lnz is continuous on the domain |z| > 0, −π < arg(z) < π, it follows that z α is continuous on
the domain |z| > 0, −π < arg(z) < π. Using polar coordinates r = |z| and θ = arg(z) we have
found that the function defined by:
f1 (z) = eα(ln r+iθ) (21)
−π < θ < π, is a branch of the multiple-valued function F (z) = z α = eα ln z . This particular
branch is called the principal branch of the complex power z α ; its branch cut is the
nonpositive real axis, and z = 0 is a branch point.
The branch f1 agrees with the principal value z α on the domain |z| > 0, −π < arg(z) < π.
Consequently, the derivative of f1 can be found using the chain rule:
d αLnz d α α
f1′ (z) = e = eαLnz (αLnz) = eαLnz = z α = αz α−1
dz dz z z
d zα
e = αz α−1 (22)
dz

Remarks: .
(i) (z α1 )α2 6= z α1 α2 unless α2 is an integer.
(ii) Some properties that do hold for complex powers do not hold for principal values of complex
powers. For example, we can prove that (z1 z2 )α = z1α z2α for any nonzero complex numbers z1
and z2 . However, this property does not hold for principal values of these complex powers.

Trigonometric and Hyperbolic Functions

Complex Sine and Cosine Functions The complex sine and cosine functions are defined
by:
eiz − e−iz
sin z = (23)
2i
eiz + e−iz
cos z = (24)
2
Analogous to real trigonometric functions, we next define the complex tangent, cotan-
gent, secant, and cosecant functions using the complex sine and cosine:
sin z cos z
tan z = cot z = (25)
cos z sin z
1 1
sec z = csc z = (26)
cos z sin z

6
Trigonometric identities
sin(−z) = − sin z cos(−z) = cos z (27)
cos z + sin2 z
2
= 1 (28)
sin(z1 ± z2 ) = sin z1 cos z2 ± cos z1 sin z2 (29)
cos(z1 ± z2 ) = cos z1 cos z2 ∓ sin z1 sin z2 (30)

Periodicity The complex sine and cosine are periodic functions with a real period of 2π
sin(z + 2π) = sin z (31)
cos(z + 2π) = cos z (32)

Example: Find all solutions to the equation sin z = 5.

Solution: from
eiz − e−iz
sin z = =5 (33)
2i
we build a quadratic equation for eiz
e2iz − 10ieiz − 1 = 0 (34)
which gives
√
eiz = (5 ± 2 6)i (35)
then
√ (4n + 1)π √
z = −i ln(5 + 2 6)i = − i ln(5 + 2 6) (36)
2
√ (4n + 1)π √
z = −i ln(5 − 2 6)i = − i ln(5 − 2 6) (37)
2

Modulus The modulus of a complex trigonometric function can also be helpful in solving
trigonometric equations. To find a formula in terms of x and y for the modulus of the sine and
cosine functions, we first express these functions in terms of their real and imaginary parts.
eiz − e−iz ei(x+iy) − e−i(x+iy)
sin z = = (38)
2i 2i
e−y (cos x + i sin x) − ey (cos x − i sin x)
= (39)
2i
ey + e−y ey − e−y
= sin x + i cos x (40)
2 2
sin z = sin x cosh y + i cos x sinh y (41)
and
cos z = cos x cosh y − i sin x sinh y (42)
Then,
q
| sin z| = sin2 x cosh2 y + cos2 x sinh2 y (43)
q
= sin2 x + sinh2 y (44)
q
| cos z| = cos2 x + sinh2 y (45)
where we have used sin2 + cos2 = 1.

7
Zeros It is a natural question to ask whether the complex sine and cosine functions have
any additional zeros in the complex plane. One way to find the zero is by recognizing that
a complex number is equal to 0 if and only if its modulus is 0. Thus, solving the equation
sin z = 0 is equivalent to solving the equation | sin z| = 0, then

| sin z|2 = sin2 x + sinh2 y = 0 (46)

Since sin2 x and sinh2 y are both nonnegative real numbers, this equation is satisfied if and only
if sin x = 0 and sinh y = 0, that is, when x = nπ, n = 0, ±1, ±2, · · · , y = 0. That is, the zeros
of the complex sine function are the same as the zeros of the real sine functions.
The only zeros of the complex cosine function are the real numbers z = (2n + 1)π/2,
n = 0, ±1, · · · .

Analyticity The derivatives of the complex sine and cosine functions are found using the
chain rule, we get
d eiz − e−iz eiz + e−iz
 
d
sin z = = = cos z (47)
dz dz 2i 2
d
cos z = − sin z (48)
dz
Since this derivative is defined for all complex z, sin z and cos z are an entire functions.
The derivatives of sin z and cos z can then be used to show that derivatives of all of the
complex trigonometric functions are the same as derivatives of the real trigonometric functions:
d
sin z = cos z (49)
dz
d
cos z = − sin z (50)
dz
d
tan z = sec2 z (51)
dz
d
cot z = − csc2 z (52)
dz
d
sec z = sec z tan z (53)
dz
d
csc z = − csc z cot z (54)
dz
The sine and cosine functions are entire, but the tangent and secant functions have singu-
larities at z = (2n + 1)π/2 for n = 0, ±1, · · · , whereas the cotangent and cosecant functions
have singularities at z = nπ for n = 0, ±1, · · · .

Complex Hyperbolic Functions

Complex Hyperbolic Sine and Cosine The complex hyperbolic sine and hyperbolic cosine
functions are defined by:
ez − e−z
sinh z = (55)
2
ez + e−z
cosh z = (56)
2
The complex hyperbolic functions are periodic and have infinitely many zeros.

8
 The complex hyperbolic tangent, cotangent, secant, and cosecant are defined in terms of
sinh z and cosh z:
sinh z cosh z
tanh z = coth z = (57)
cosh z sinh z
1 1
sechz = cschz = (58)
cosh z sinh z
The hyperbolic sine and cosine functions are entire because the functions ez and e−z are entire.
From the chain rule we have:
d
sinh z = cosh z (59)
dz
d
cosh z = sinh z (60)
dz
d
tanh z = sech2 z (61)
dz
d
coth z = −csch2 z (62)
dz
d
sechz = −sechz tanh z (63)
dz
d
cschz = −cschz coth z (64)
dz

Relation to sine and cosine By replacing z with iz in the definition of sinh z we have
eiz − e−iz
sinh(iz) = = i sin z ⇒ sin(z) = −i sinh(iz) (65)
2
In a similar manner can be obtained others identities,

sin(z) = −i sinh(iz) (66)

cos(z) = cosh(iz) (67)
sinh(z) = −i sin(iz) (68)
cosh(z) = cos(iz) (69)

Relations between the other trigonometric and hyperbolic functions can now be derived
from the above ones, for example,

tan(iz) = i tanh(z) (70)

We can also use above relations to derive hyperbolic identities from trigonometric identities,

sinh(−z) = − sinh z cosh(−z) = cosh z (71)

cosh z − sinh2 z
2
= 1 (72)
sinh(z1 ± z2 ) = sinh z1 cosh z2 ± cosh z1 sinh z2 (73)
cosh(z1 ± z2 ) = cosh z1 cosh z2 ± sinh z1 sinh z2 (74)

Inverse Trigonometric and Hyperbolic Functions

The complex sine function is periodic with a real period of 2π. We also known that the sine
function maps the complex plane onto the complex plane. These two properties imply that for

9
any complex number z there exists infinitely many solutions w to the equation sin w = z. Let
us find an explicit formula for w
eiw − e−iw
sin w = z = ⇒ ei2w − 2izeiw − 1 = 0 (75)
2i
then

eiw = iz + (1 − z 2 )1/2 (76)

where (1 − z 2 )1/2 represents the two square roots of 1 − z 2 .

Finally, we solve for w using the complex logarithm:

w = −i ln iz + (1 − z 2 )1/2
 
(77)

Each value of w obtained from the above equation satisfies the equation sin w = z. Therefore,
we call this multiple-valued function the inverse sine:

Inverse Sine The multiple-valued function sin−1 z defined by:

sin−1 z = arcsinz = −i ln iz + (1 − z 2 )1/2

 
(78)

is called the inverse sine. The inverse sine is multiple-valued since it is defined in terms of the
complex logarithm ln z. It is also worth repeating that the expression (1 − z 2 )1/2 represents the
two square roots of 1 − z 2 .
√
Example Find all values √ of sin−1
5.
Solution: By setting z = 5 we get,
√ √ √ 1/2
  
sin −1
5 = −i ln i 5 + 1 − ( 5)2 (79)
h√ i
= −i ln i 5 + (−4)1/2 (80)

The two square roots (−4)1/2 of –4 are found to be ±2i, then

√ h√ i h √ i
sin−1 5 = −i ln i 5 ± 2i = −i ln i 5±2 (81)

Besides,
h √ i √   h √ i 
ln i 5±2 = ln | 5 ± 2 | + i Arg i 5 ± 2 + 2nπ (82)
√  π 
= ln 5±2 +i + 2nπ (83)
2
Let us noticing the following identity,
√
" √ #
 √ 5+2
ln 5−2 = ln ( 5 − 2) √ (84)
5+2
 
5−4
= ln √ (85)
5+2
 
1
= ln √ (86)
5+2
√ 
= − ln 5+2 (87)

10
which implies,
h √ i √  π 
ln i 5±2 = ± ln 5+2 +i + 2nπ (88)
2
Then,

−1
√ h √ i
sin 5 = −i ln i 5±2 (89)
h √  π i
= (−i) ± ln 5+2 +i + 2nπ (90)
√  π 2
= ∓i ln 5+2 + + 2nπ (91)
2
1 + 4n  √ 
= π ∓ i ln 5+2 (92)
2

Inverse cosine and inverse tangent

cos−1 z = −i ln z + i(1 − z 2 )1/2

 
(93)
 
−1 i i+z
tan z = ln (94)
2 i−z

Both the inverse cosine and inverse tangent are multiple-valued functions since they are defined
in terms of the complex logarithm ln z.

Branches and Analyticity The inverse sine and inverse cosine are multiple-valued functions
that can be made single-valued by specifying a single value of the square root to use for the
expression (1 − z 2 )1/2 and a single value of the complex logarithm to use. The inverse tangent,
on the other hand, can be made single-valued by just specifying a single value of ln z to use.

Example: We can define a function f that gives a value of the inverse sine by using √ the
principal square root and the principal
√ value of the complex
√ logarithm. If,
√ say, z = 5, then
the principal square root of 1 − ( 5)2 = −4 is 2i, and Ln(i 5 + 2i) = ln( 5 + 2) + iπ/2, then
√ π √
f ( 5) = − i ln( 5 + 2) (95)
2
√ √
Thus, we see that the value of the function f at z = 5 is the value of sin−1 5 associated to
n = 0 and the square root 2i in the example above.
A branch of a multiple-valued inverse trigonometric function may be obtained by choosing
a branch of the square root function and a branch of the complex logarithm. Determining the
domain of a branch defined in this manner can be quite involved.

Derivatives of Branches sin−1 z, cos−1 z, and tan−1 z The following formulas for the deriva-
tives hold only on the domains of these branches,
d 1
sin−1 z = (96)
dz (1 − z 2 )1/2
d −1
cos−1 z = (97)
dz (1 − z 2 )1/2
d 1
tan−1 z = (98)
dz 1 + z2

11
Inverse Hyperbolic Functions The foregoing discussion of inverse trigonometric functions
can be repeated for hyperbolic functions. This leads to the definition of the inverse hyperbolic
functions stated below. Once again these inverses are defined in terms of the complex logarithm
because the hyperbolic functions are defined in terms of the complex exponential.

Inverse Hyperbolic Sine, Cosine, and Tangent The multiple-valued functions sinh−1 z,
cosh−1 z, and tanh−1 z, defined by:

sinh−1 z = ln z + (z 2 + 1)1/2
 
(99)
2 1/2
−1
 
cosh z = ln z + (z − 1) (100)
 
1 1+z
tanh−1 z = ln (101)
2 1−z

These expressions allow us to solve equations involving the complex hyperbolic functions. In
particular, if w = sinh−1 z, then sinh w = z.
Branches of the inverse hyperbolic functions are defined by choosing branches of the square
root and complex logarithm, or, in the case of the inverse hyperbolic tangent, just choosing
a branch of the complex logarithm. The derivative of a branch can be found using implicit
differentiation. The following result gives formulas for the derivatives of branches of the inverse
hyperbolic functions. In these formulas, the symbols sinh−1 z, cosh−1 z, and tanh−1 z represent
branches of the corresponding inverse hyperbolic multiple-valued functions.
d 1
sinh−1 z = (102)
dz (z + 1)1/2
2

d 1
cosh−1 z = (103)
dz (z − 1)1/2
2

d −1
tanh−1 z = 2 (104)
dz z −1

12