1

Chapter 4 Differentiation

In this chapter, we will learn the definition of derivative. We will study the
basic rules of differentiation and compute the derivatives of trigonometric
functions, inverse trigonometric functions, the exponential function and
logarithmic functions. Moreover, some techniques of differentiation will be
discussed.
 2

I. Definition
II. Basic Properties
III. Chain Rule
IV. Derivatives of Trigonometric Functions
V. Derivatives of Exponential and Logarithmic Functions
VI. Implicit Differentiation
VII. Higher Derivatives
 3

I. Definition
A.First Principle
The derivative of the function f(x) with respect to the variable x at x = x0 is
defined by
′
d𝑓𝑓 𝑓𝑓(𝑥𝑥0 + ℎ) − 𝑓𝑓(𝑥𝑥0 )
𝑓𝑓 (𝑥𝑥0 ) = � = lim
d𝑥𝑥 𝑥𝑥=𝑥𝑥0 ℎ→0 ℎ
provided the limit exists. If this limit exists at x0, we say that f is differentiable
at x0. This definition is called the ‘First Principle’.

To denote the derivative of the function y = f(x) at x = x0, we may use one of
the following notations:
𝑓𝑓 ′ (𝑥𝑥0 ) or 𝑦𝑦′(𝑥𝑥0 ) (Newton’s Notation)
d𝑓𝑓 d𝑦𝑦
or � or � (Leibniz’s Notation)
d𝑥𝑥 𝑥𝑥=𝑥𝑥0 d𝑥𝑥 𝑥𝑥=𝑥𝑥0
 4

Geometrically, the derivative gives the slope of the curve y = f(x).

The tangent line to the curve 𝑦𝑦 = 𝑓𝑓(𝑥𝑥) at 𝑥𝑥0 is defined as the straight line
through 𝑥𝑥0 with slope 𝑓𝑓 ′ (𝑥𝑥0 ). (red line in the above picture)
 5

Let’s illustrate the limiting process with the example 𝑓𝑓(𝑥𝑥) = 𝑥𝑥 2 at x = 0:

• The graph of 𝑦𝑦 = 𝑓𝑓(𝑥𝑥) is given by the blue line.
• We draw a straight line (red) joining (0, 𝑓𝑓 (0)) and (1, 𝑓𝑓(1)) and
compute its slope to be 1.
• Next, we choose a point closer to 0, the green line joining (0, 𝑓𝑓(0)) and
(0.5, 𝑓𝑓(0.5)) has slope 0.5.

x
 6

• If we continue the process of choosing a point closer to the origin, we get a

straight line with smaller slope closer to 0.
• The limiting process is exactly the idea of differentiation. We compute the
derivative at x = 0:
𝑓𝑓 ( 0 + ℎ ) − 𝑓𝑓(0) ℎ 2−0
𝑓𝑓 ′ (0) = lim = lim = lim ℎ = 0
ℎ→0 ℎ ℎ→0 ℎ ℎ→0

• We say that the slope of 𝑓𝑓 at x = 0 is 0. And the tangent line is a horizontal

line.
 7

We can consider the derivative at any x, provided that the limit exists. So we
can in turn define another function called the derivative of 𝑓𝑓(𝑥𝑥):
𝑓𝑓(𝑥𝑥 + ℎ) − 𝑓𝑓(𝑥𝑥)
𝑓𝑓 ′ (𝑥𝑥)
= lim
ℎ→0 ℎ
• The domain is the set of points where the function f is differentiable (note
that it can be different from the domain of the original function f).
• The following notations are equivalent in representing the derivative of
the function 𝑦𝑦 = 𝑓𝑓(𝑥𝑥):
d𝑓𝑓 d𝑦𝑦 d
𝑓𝑓 ′ (𝑥𝑥), 𝑦𝑦 ′ , , , 𝑓𝑓(𝑥𝑥)
d𝑥𝑥 d𝑥𝑥 d𝑥𝑥

The process of finding the derivative is called differentiation / differentiating

f (with respect to x).
 8

Example 1.1
Evaluate the derivative of the function (by First Principle) 𝑓𝑓(𝑥𝑥) = √𝑥𝑥 for x > 0.
Solution:
′( )
𝑓𝑓 (𝑥𝑥 + ℎ) − 𝑓𝑓(𝑥𝑥)
𝑓𝑓 𝑥𝑥 = lim
ℎ→0 ℎ
√𝑥𝑥 + ℎ − √𝑥𝑥
= lim
ℎ→0 ℎ
�√𝑥𝑥 + ℎ − √𝑥𝑥��√𝑥𝑥 + ℎ + √𝑥𝑥�
= lim
ℎ→0 ℎ�√𝑥𝑥 + ℎ + √𝑥𝑥�
𝑥𝑥 + ℎ − 𝑥𝑥
= lim
ℎ→0 ℎ�√𝑥𝑥 + ℎ + √𝑥𝑥�

1 1
= lim =
ℎ→0 �√𝑥𝑥 + ℎ + √𝑥𝑥� 2√𝑥𝑥
 9

B. One-sided Derivative
Right-hand derivative:
𝑓𝑓(𝑥𝑥 + ℎ) − 𝑓𝑓(𝑥𝑥)
𝑅𝑅𝑅𝑅′(𝑥𝑥) = lim+
ℎ→0 ℎ
Left-hand derivative:
𝑓𝑓(𝑥𝑥 + ℎ) − 𝑓𝑓(𝑥𝑥)
𝐿𝐿𝐿𝐿′(𝑥𝑥) = lim−
ℎ→0 ℎ
A function has a derivative at a point if and only if it has left-hand and right-
hand derivatives, and these one-sided derivatives are equal.
 10

Example 1.2
Given
𝑥𝑥 2 + 3𝑥𝑥 + 2 𝑖𝑖𝑖𝑖 𝑥𝑥 < 1
𝑓𝑓(𝑥𝑥) = �
5𝑥𝑥 + 1 𝑖𝑖𝑖𝑖 𝑥𝑥 ≥ 1
Is f(x) differentiable at x = 1? If so, find f’(1).
Solution:
We first compute the left-hand derivative,
′(
𝑓𝑓 (1 + ℎ) − 𝑓𝑓(1) (1 + ℎ)2 + 3(1 + ℎ) + 2 − 6 ℎ2 + 5ℎ
𝐿𝐿𝑓𝑓 1) = lim− = lim− = lim− =5
ℎ→0 ℎ ℎ→0 ℎ ℎ→0 ℎ
Next, we compute the right-hand derivative,
′(
𝑓𝑓 (1 + ℎ) − 𝑓𝑓 (1) 5(1 + ℎ) + 1 − 6 5ℎ
𝑅𝑅𝑓𝑓 1) = lim+ = lim+ = lim+ =5
ℎ→0 ℎ ℎ→0 ℎ ℎ→0 ℎ

Since 𝐿𝐿𝑓𝑓 ′ (1) = 𝑅𝑅𝑓𝑓 ′ (1), thus f(x) is differentiable at x = 1 and f’(1) = 5.
 11

C. Examples (of functions which are not differentiable)

1.Corner: one-sided derivatives differ
𝑓𝑓(𝑥𝑥) = |𝑥𝑥| is not differentiable at x = 0.

( )
(a) Right-hand derivative: lim+ 𝑓𝑓 0+ℎℎ−𝑓𝑓(0) = lim+ ℎ−0
ℎ
=1
ℎ→0 ℎ→0

( )
(b) Left-hand derivative: lim− 𝑓𝑓 0+ℎℎ−𝑓𝑓(0) = lim− −ℎ−0
ℎ
= −1
ℎ→0 ℎ→0

𝑓𝑓(0 + ℎ) − 𝑓𝑓(0) 𝑓𝑓 (0 + ℎ) − 𝑓𝑓(0)

lim ≠ lim+
ℎ→0− ℎ ℎ→0 ℎ
 12

2.Vertical tangent: slope approaches ∞

𝑓𝑓(𝑥𝑥) = √𝑥𝑥 is not differentiable at x = 0.

1
𝑓𝑓′(0) = lim+ →∞
ℎ→0 �√ℎ + 0�
 13

3.Cusp: slope approaches ∞ from one side and −∞ from the other side.

−𝑥𝑥 if 𝑥𝑥 < 0
𝑓𝑓(𝑥𝑥) = �√
√𝑥𝑥 if 𝑥𝑥 ≥ 0

√ℎ
(a) Right-hand derivative: lim →∞
ℎ→0+ ℎ
√−ℎ
(b) Left-hand derivative: lim− → −∞
ℎ→0 ℎ
 14

4.Discontinuity
Heaviside function is not differentiable at x = 0.
0 if 𝑥𝑥 < 0
𝐻𝐻(𝑥𝑥) = �0.5 if 𝑥𝑥 = 0
1 if 𝑥𝑥 > 0

Right-hand derivative:
𝐻𝐻 (0 + ℎ) − 𝐻𝐻(0) 1 − 0.5 0.5
lim+ = lim+ = lim+ →∞
ℎ→0 ℎ ℎ→0 ℎ ℎ→0 ℎ
 15

D.Differentiability implies Continuity

If f is differentiable at x = c, then f is continuous at x = c.

Caution: Converse does not hold!

 16

Example 1.3
𝑎𝑎𝑎𝑎 if 𝑥𝑥 < 0
𝑓𝑓(𝑥𝑥) = � 0 if 𝑥𝑥 = 0
sin 𝑥𝑥 + 𝑏𝑏 if 𝑥𝑥 > 0
If f is differentiable at x = 0, find the values of a and b.
Solution:
i. Since f is differentiable at x = 0, f is continuous at x = 0.
𝑓𝑓(0) = lim+ 𝑓𝑓(𝑥𝑥 )
𝑥𝑥→0
0 = lim+ (sin 𝑥𝑥 + 𝑏𝑏)
𝑥𝑥→0
𝑏𝑏 = 0
ii. Since f is differentiable at x = 0
𝑓𝑓(0 + ℎ) − 𝑓𝑓(0) 𝑓𝑓 (0 + ℎ) − 𝑓𝑓(0)
lim = lim+
ℎ→0− ℎ ℎ→0 ℎ
𝑎𝑎ℎ sin ℎ + 𝑏𝑏
lim = lim+
ℎ→0− ℎ ℎ→0 ℎ
𝑎𝑎 = 1
 17

Example 1.4
Suppose that a function f satisfies the following conditions for all values of x and y:
𝑓𝑓(𝑥𝑥 + 𝑦𝑦) = 𝑓𝑓 (𝑥𝑥 )𝑓𝑓(𝑦𝑦)
𝑓𝑓(𝑥𝑥 ) = 1 + 𝑥𝑥𝑥𝑥(𝑥𝑥)
where g is a function with lim 𝑔𝑔(𝑥𝑥 ) = 1. Show that f is differentiable and 𝑓𝑓 ′ (𝑥𝑥) = 𝑓𝑓 (𝑥𝑥 ).
𝑥𝑥→0
Solution:
′
𝑓𝑓(𝑥𝑥 + ℎ) − 𝑓𝑓(𝑥𝑥 ) 𝑓𝑓(𝑥𝑥 )𝑓𝑓(ℎ) − 𝑓𝑓 (𝑥𝑥 ) 𝑓𝑓(𝑥𝑥 )[𝑓𝑓(ℎ) − 1]
𝑓𝑓 (𝑥𝑥) = lim = lim = lim
ℎ→0 ℎ ℎ→0 ℎ ℎ→0 ℎ
By the second condition, we have
[𝑓𝑓(ℎ) − 1]
= 𝑔𝑔(ℎ)
ℎ
Thus,
𝑓𝑓(𝑥𝑥 )[𝑓𝑓(ℎ) − 1]
lim = lim 𝑓𝑓 (𝑥𝑥 )𝑔𝑔(ℎ) = 𝑓𝑓(𝑥𝑥) lim 𝑔𝑔(ℎ) = 𝑓𝑓(𝑥𝑥 ) ⋅ 1 = 𝑓𝑓(𝑥𝑥)
ℎ→0 ℎ ℎ→0 ℎ→0

Hence, f is differentiable and 𝑓𝑓 ′ (𝑥𝑥 ) = 𝑓𝑓 (𝑥𝑥 ).

 18

II. Basic Properties

Let u and v be differentiable functions of x.
A.Sum Rule
d d𝑢𝑢 d𝑣𝑣
(𝑢𝑢 + 𝑣𝑣 ) = +
d𝑥𝑥 d𝑥𝑥 d𝑥𝑥

B. Product Rule
d d𝑣𝑣 d𝑢𝑢
(𝑢𝑢𝑢𝑢 ) = 𝑢𝑢 + 𝑣𝑣
d𝑥𝑥 d𝑥𝑥 d𝑥𝑥
In particular, for any differentiable function f and any constant c,
d d𝑓𝑓
(𝑐𝑐𝑐𝑐 ) = 𝑐𝑐
d𝑥𝑥 d𝑥𝑥
 19

C. Quotient Rule
𝑢𝑢
If 𝑣𝑣 (𝑥𝑥) ≠ 0, then is differentiable at x and
𝑣𝑣

d𝑢𝑢 d𝑣𝑣
d 𝑢𝑢 𝑣𝑣 − 𝑢𝑢
� �= d𝑥𝑥 d𝑥𝑥
d𝑥𝑥 𝑣𝑣 𝑣𝑣 2

D.Power Rule
For any real number a,
d 𝑎𝑎
𝑥𝑥 = 𝑎𝑎𝑎𝑎 𝑎𝑎−1
d𝑥𝑥
 20

For example,
d 5
𝑥𝑥 = 5𝑥𝑥 4
d𝑥𝑥
d 1 d −2 −3
2
2
= 𝑥𝑥 = −2𝑥𝑥 = − 3
d𝑥𝑥 𝑥𝑥 d𝑥𝑥 𝑥𝑥
d d 1/2 1 −1 1
√𝑥𝑥 = 𝑥𝑥 = 𝑥𝑥 2 =
d𝑥𝑥 d𝑥𝑥 2 2√𝑥𝑥
 21

Example 2.1
Differentiate the following functions with respect to x:
(a) 𝑦𝑦 = (𝑥𝑥 2 + 1)(𝑥𝑥 3 + 3)
𝑦𝑦 ′ = (𝑥𝑥 2 + 1)′ (𝑥𝑥 3 + 3) + (𝑥𝑥 2 + 1)(𝑥𝑥 3 + 3)′
= 2𝑥𝑥(𝑥𝑥 3 + 3) + 3𝑥𝑥 2 (𝑥𝑥 2 + 1)
= 5𝑥𝑥 4 + 3𝑥𝑥 2 + 6𝑥𝑥
Alternatively, note that 𝑦𝑦 = 𝑥𝑥 5 + 𝑥𝑥 3 + 3𝑥𝑥 2 + 3. Then,
𝑦𝑦 ′ = 5𝑥𝑥 4 + 3𝑥𝑥 2 + 6𝑥𝑥
 22

(b)
𝑥𝑥 2 − 1
𝑦𝑦 = 2
𝑥𝑥 + 1
dy (𝑥𝑥 2 + 1)(2𝑥𝑥) − (𝑥𝑥 2 − 1)(2𝑥𝑥) 4𝑥𝑥
= 2 2
= 2
d𝑥𝑥 (𝑥𝑥 + 1) (𝑥𝑥 + 1)2
 23

(c)
(𝑥𝑥 − 1)(𝑥𝑥 2 − 2𝑥𝑥)
𝑦𝑦 =
𝑥𝑥 4

Firstly, note that

1 3 2
𝑦𝑦 = − 2 + 3
𝑥𝑥 𝑥𝑥 𝑥𝑥

Then, we differentiate y
d𝑦𝑦 1 6 6
=− 2+ 3− 4
d𝑥𝑥 𝑥𝑥 𝑥𝑥 𝑥𝑥
 24

Example 2.2
d𝑦𝑦 d𝑦𝑦
(a) Calculate � and � for
d𝑥𝑥 𝑥𝑥=0 d𝑥𝑥 𝑥𝑥=−1

𝑦𝑦 = 𝑥𝑥 7 − 5𝑥𝑥 2 − 100
Solution:
(a) 𝑦𝑦′ = 7𝑥𝑥 6 − 10𝑥𝑥
d𝑦𝑦
� = 7(0)6 − 10(0) = 0
d𝑥𝑥 𝑥𝑥=0
d𝑦𝑦
� = 7(−1)6 − 10(−1) = 17
d𝑥𝑥 𝑥𝑥=−1
 25

(b) Find the tangent lines for the curve

𝑦𝑦 = 𝑥𝑥 4 − 2𝑥𝑥 2 + 2
at (i) 𝑥𝑥 = 0, (ii) 𝑥𝑥 = 1, (iii) 𝑥𝑥 = −1, (iv) 𝑥𝑥 = 2.
Solution:
𝑦𝑦′ = 4𝑥𝑥 3 − 4𝑥𝑥
(i) At 𝑥𝑥 = 0, 𝑦𝑦′(0) = 0.
The tangent line at 𝑥𝑥 = 0 is given by
𝑦𝑦 = 2
(ii) At 𝑥𝑥 = 1, 𝑦𝑦′(1) = 0.
The tangent line at 𝑥𝑥 = 1 is given by
𝑦𝑦 = 1
 26

(iii) At 𝑥𝑥 = −1, 𝑦𝑦′(−1) = 0.

The tangent line at 𝑥𝑥 = −1 is given by
𝑦𝑦 = 1

(iv) At 𝑥𝑥 = 2, 𝑦𝑦′(2) = 24.

The tangent line at 𝑥𝑥 = 2 is given by
𝑦𝑦 = 24𝑥𝑥 − 38
 27

III. Chain Rule

If f(u) is differentiable at the point u = g(x) and g(x) is differentiable at x, then

the composite function (𝑓𝑓 ∘ 𝑔𝑔)(𝑥𝑥) = 𝑓𝑓(𝑔𝑔(𝑥𝑥)) is differentiable at x, and
(𝑓𝑓 ∘ 𝑔𝑔)′(𝑥𝑥) = 𝑓𝑓′(𝑔𝑔(𝑥𝑥)) ∙ 𝑔𝑔′(𝑥𝑥).

In Leibniz’s notation, if y = f(u) and u = g(x), then

d𝑦𝑦 d𝑦𝑦 d𝑢𝑢
= ∙
d𝑥𝑥 d𝑢𝑢 d𝑥𝑥
d𝑦𝑦
where is evaluated at u = g(x).
d𝑢𝑢
 28

Example 3.1
Differentiate the following functions with respect to x:
(a) 𝑦𝑦(𝑥𝑥) = (3𝑥𝑥 2 + 1)2

We may define 𝑢𝑢 = 3𝑥𝑥 2 + 1, then 𝑦𝑦 = 𝑢𝑢2 . By Chain Rule, we have

𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑𝑢𝑢2 𝑑𝑑(3𝑥𝑥 2 + 1)
= = = (2𝑢𝑢)(6𝑥𝑥) = 12𝑥𝑥(3𝑥𝑥 2 + 1)
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑

Alternatively, we may simply write

𝑑𝑑𝑑𝑑 𝑑𝑑 (3𝑥𝑥 2 + 1)2 𝑑𝑑 (3𝑥𝑥 2 + 1)2 𝑑𝑑 (3𝑥𝑥 2 + 1)
= =
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑 (3𝑥𝑥 2 + 1) 𝑑𝑑𝑑𝑑
= [2(3𝑥𝑥 2 + 1)](6𝑥𝑥) = 12𝑥𝑥(3𝑥𝑥 2 + 1)
 29

(b)
𝑓𝑓(𝑥𝑥) = (5𝑥𝑥 3 − 𝑥𝑥 4 )7
𝑓𝑓′(𝑥𝑥) = 7(5𝑥𝑥 3 − 𝑥𝑥 4 )6 (15𝑥𝑥 2 − 4𝑥𝑥 3 )
(c)
1
𝑦𝑦(𝑥𝑥) =
3𝑥𝑥 − 2
′
1 3
𝑦𝑦 (𝑥𝑥) = − 2
(3) = −
(3𝑥𝑥 − 2) (3𝑥𝑥 − 2)2
(d)
𝑓𝑓(𝑥𝑥) = (𝑥𝑥 3 + 2)7 (𝑥𝑥 2 − 5)2
𝑓𝑓 ′ (𝑥𝑥) = (𝑥𝑥 3 + 2)7 [2(𝑥𝑥 2 − 5)(2𝑥𝑥)] + (𝑥𝑥 2 − 5)2 [7(𝑥𝑥 3 + 2)6 (3𝑥𝑥 2 )]
= 4𝑥𝑥(𝑥𝑥 3 + 2)7 (𝑥𝑥 2 − 5) + 21𝑥𝑥 2 (𝑥𝑥 3 + 2)6 (𝑥𝑥 2 − 5)2
 30

𝑥𝑥
(e) 𝑦𝑦 = for −1 < 𝑥𝑥 < 1
√1−𝑥𝑥 2
1
√1 − − 𝑥𝑥 𝑥𝑥 2 (−2𝑥𝑥)
𝑑𝑑𝑑𝑑 2√1 − 𝑥𝑥 2
=
𝑑𝑑𝑑𝑑 1 − 𝑥𝑥 2
𝑥𝑥 2
√1 − 𝑥𝑥 2 +
√1 − 𝑥𝑥 2
=
1 − 𝑥𝑥 2
1
=
(1 − 𝑥𝑥 2 )3/2
 31

(f)

𝑓𝑓(𝑥𝑥) = 4𝑥𝑥 �𝑥𝑥 + √𝑥𝑥

for x > 0.
1 1 6𝑥𝑥 + 5√𝑥𝑥
𝑓𝑓 ′ (𝑥𝑥) = 4�𝑥𝑥 + √𝑥𝑥 + 4𝑥𝑥 �1 + �=
2�𝑥𝑥 + √𝑥𝑥 2√𝑥𝑥 �𝑥𝑥 + √𝑥𝑥
 32

Example 3.2
(a) Suppose that f is a differentiable odd function.
Show that 𝑓𝑓 ′ (−𝑥𝑥) = 𝑓𝑓 ′ (𝑥𝑥) for any x.
(b) Suppose that f is a differentiable even function.
Show that 𝑓𝑓 ′ (−𝑥𝑥) = −𝑓𝑓 ′ (𝑥𝑥) for any x.
Solution:
(a) Since f is an odd function, 𝑓𝑓(−𝑥𝑥) = −𝑓𝑓(𝑥𝑥) for all x. We differentiate
both sides of the above equation with respect to x. Applying Chain Rule
to the left hand side, we get
𝑓𝑓′(−𝑥𝑥)(−1) = −𝑓𝑓 ′ (𝑥𝑥)
Hence, 𝑓𝑓 ′ (−𝑥𝑥) = 𝑓𝑓 ′ (𝑥𝑥).
(b) Since f is an even function, 𝑓𝑓(−𝑥𝑥) = 𝑓𝑓(𝑥𝑥) for all x. We differentiate
both sides of the above equation with respect to x,
𝑓𝑓′(−𝑥𝑥)(−1) = 𝑓𝑓 ′ (𝑥𝑥) ⇒ 𝑓𝑓 ′ (−𝑥𝑥) = −𝑓𝑓 ′ (𝑥𝑥)
 33

Derivative Rule for Inverses

If f is a function with an interval domain I and the derivative 𝑓𝑓′ exists and is
non-zero on I. Then the inverse function 𝑓𝑓 −1 is differentiable and the
derivative is given by:
−1 ′
1
(𝑓𝑓 ) (𝑏𝑏) =
𝑓𝑓′(𝑓𝑓 −1 (𝑏𝑏))
That is, (in Leibniz’s notation),
d𝑓𝑓 −1 1
� =
d𝑥𝑥 𝑥𝑥=𝑏𝑏 d𝑓𝑓
�
d𝑥𝑥 𝑥𝑥=𝑓𝑓−1 (𝑏𝑏)
 34

IV. Derivatives of Trigonometric Functions

A.Derivative of sine and cosine
d
(sin 𝑥𝑥) = cos 𝑥𝑥
d𝑥𝑥
d
(cos 𝑥𝑥 ) = −sin 𝑥𝑥
d𝑥𝑥
 35

B. Derivatives of the Other Trigonometric Functions

d
(tan 𝑥𝑥) = sec 2 𝑥𝑥
d𝑥𝑥
d
(sec 𝑥𝑥 ) = sec 𝑥𝑥 tan 𝑥𝑥
d𝑥𝑥
d
(csc 𝑥𝑥 ) = −csc 𝑥𝑥 cot 𝑥𝑥
d𝑥𝑥
d
(cot 𝑥𝑥 ) = −csc 2 𝑥𝑥
d𝑥𝑥
 36

C. Derivatives of the Inverse Trigonometric Functions

d −1
d 1
(sin 𝑥𝑥 ) = arcsin 𝑥𝑥 =
d𝑥𝑥 d𝑥𝑥 √1 − 𝑥𝑥 2
d −1
d 1
(cos 𝑥𝑥 ) = arccos 𝑥𝑥 = −
d𝑥𝑥 d𝑥𝑥 √1 − 𝑥𝑥 2
d d 1
(tan−1 𝑥𝑥 ) = arctan 𝑥𝑥 =
d𝑥𝑥 d𝑥𝑥 1 + 𝑥𝑥 2
d −1
1
(𝑠𝑠𝑠𝑠𝑠𝑠 𝑥𝑥 ) =
d𝑥𝑥 |𝑥𝑥|√𝑥𝑥 2 − 1
d −1
1
(𝑐𝑐𝑐𝑐𝑐𝑐 𝑥𝑥 ) = −
d𝑥𝑥 |𝑥𝑥|√𝑥𝑥 2 − 1
d −1
1
(cot 𝑥𝑥 ) = −
d𝑥𝑥 1 + 𝑥𝑥 2
 37

Remark: These are the INVERSE functions, which are different from the
reciprocals.

Example 4.1
Differentiate the following functions with respect to x:
(a)
𝑦𝑦(𝑥𝑥) = 𝑥𝑥 3 − 8𝑥𝑥 + cos 𝑥𝑥
𝑦𝑦′(𝑥𝑥) = 3𝑥𝑥 2 − 8 − sin 𝑥𝑥
(b)
4
𝑦𝑦(𝑥𝑥) = + 3 sin 𝑥𝑥
𝑥𝑥
4
𝑦𝑦′(𝑥𝑥) = − 2 + 3cos 𝑥𝑥
𝑥𝑥
 38

(c)
𝑦𝑦(𝑥𝑥) = 𝑥𝑥 2 tan 𝑥𝑥
𝑦𝑦′(𝑥𝑥) = 2𝑥𝑥 tan 𝑥𝑥 + 𝑥𝑥 2 sec 2 𝑥𝑥
(d)
𝑦𝑦(𝑥𝑥) = sin(𝑥𝑥 2 )
Then, 𝑦𝑦 = sin 𝑢𝑢 where 𝑢𝑢 = 𝑥𝑥 2 . By Chain Rule, we have
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑 sin 𝑢𝑢 𝑑𝑑(𝑥𝑥 2 )
= = = cos 𝑢𝑢 (2𝑥𝑥) = 2𝑥𝑥 cos(𝑥𝑥 2 )
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑
Alternatively, we may write
𝑑𝑑𝑑𝑑 𝑑𝑑 sin(𝑥𝑥 2 ) 𝑑𝑑 sin(𝑥𝑥 2 ) 𝑑𝑑 (𝑥𝑥 2 )
= =
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑 (𝑥𝑥 2 ) 𝑑𝑑𝑑𝑑
= [cos(𝑥𝑥 2 )](2𝑥𝑥) = 2𝑥𝑥 cos(𝑥𝑥 2 )
 39

Example 4.2
Differentiate the following functions with respect to x:
(a) 𝑦𝑦(𝑥𝑥) = sin5 𝑥𝑥 + sin(𝑥𝑥 2 + 𝑥𝑥)
𝑦𝑦′(𝑥𝑥) = 5sin4 𝑥𝑥 cos 𝑥𝑥 + cos(𝑥𝑥 2 + 𝑥𝑥) (2𝑥𝑥 + 1)

(b) 𝑦𝑦(𝑥𝑥) = tan(5 − sin(2𝑥𝑥))

𝑦𝑦 ′ (𝑥𝑥) = sec 2 (5 − sin(2𝑥𝑥)) [− cos(2𝑥𝑥)(2)]
= −2 cos(2𝑥𝑥) sec 2 (5 − sin(2𝑥𝑥))
 40

(c)
sin 𝑥𝑥
𝑦𝑦(𝑥𝑥) =
𝑥𝑥
𝑥𝑥cos 𝑥𝑥 − sin 𝑥𝑥
𝑦𝑦′(𝑥𝑥) =
𝑥𝑥 2

(d)
𝑦𝑦 = tan √1 − 𝑥𝑥
−1 −�sec 2 √1 − 𝑥𝑥�
𝑦𝑦 ′ = �sec 2 √1 − 𝑥𝑥� =
2√1 − 𝑥𝑥 2√1 − 𝑥𝑥
 41

(e)
𝑦𝑦 = csc 5 (1 − 𝑥𝑥 + 3𝑥𝑥 2 )

𝑦𝑦 ′
= [5csc 4 (1 − 𝑥𝑥 + 3𝑥𝑥 2 )][− csc(1 − 𝑥𝑥 + 3𝑥𝑥 2 ) cot(1 − 𝑥𝑥 + 3𝑥𝑥 2 )]
(−1 + 6𝑥𝑥)
= 5(1 − 6𝑥𝑥)csc 5 (1 − 𝑥𝑥 + 3𝑥𝑥 2 ) cot(1 − 𝑥𝑥 + 3𝑥𝑥 2 )
 42

Example 4.3
Differentiate the following functions with respect to x:
(a)
𝑦𝑦 = sin−1 (𝑥𝑥 2 ) + tan−1 √𝑥𝑥
(b)
𝑦𝑦 = sin−1 (sin2 𝑥𝑥)
(c)
1 + 2 cos 𝑥𝑥
𝑦𝑦 = cos −1 � �
2 + cos 𝑥𝑥
 43

Example 4.4
2
1
𝑓𝑓(𝑥𝑥) = � 𝑥𝑥 sin 𝑥𝑥 if 𝑥𝑥 ≠ 0
0 if 𝑥𝑥 = 0
Find the derivative of f(x). Is f’(x) continuous at x = 0?
Solution:
1 1
For 𝑥𝑥 ≠ 0, 𝑓𝑓 ′ (𝑥𝑥 ) = 2𝑥𝑥 sin − cos
𝑥𝑥 𝑥𝑥
1
ℎ2 sin −0 1 1
At 𝑥𝑥 = 0, 𝑓𝑓 ′ (0) = lim ℎ
= lim ℎ sin = 0 (since 0 ≤ �ℎ sin � ≤ |ℎ|, by
ℎ→0 ℎ ℎ→0 ℎ ℎ

Sandwich Theorem, we have the last equality)

f’(x) is not continuous at x = 0 because
′(
1 1
lim 𝑓𝑓 𝑥𝑥 ) = lim �2𝑥𝑥 sin − cos �
𝑥𝑥→0 𝑥𝑥→0 𝑥𝑥 𝑥𝑥
does not exist.
 44

′(
Remark: We cannot find 𝒇𝒇 𝟎𝟎) by taking the
′(
limit 𝐥𝐥𝐥𝐥𝐥𝐥𝒇𝒇 𝒙𝒙), which has assumed the
𝒙𝒙→𝟎𝟎
continuity of 𝒇𝒇′ (𝒙𝒙).
 45

V. Derivatives of Exponential and Logarithmic Functions

A.Natural Logarithm Function
Definition: ln 𝑥𝑥 is the inverse of the exponential function e𝑥𝑥
Properties of ln 𝑥𝑥:
For any numbers a > 0 and x > 0, the natural logarithm satisfies the following
rules:
1.Product Rule: ln 𝑎𝑎𝑎𝑎 = ln 𝑎𝑎 + ln 𝑥𝑥
𝑎𝑎
2.Quotient Rule: ln = ln 𝑎𝑎 − ln 𝑥𝑥
𝑥𝑥
𝑟𝑟
3.Power Rule: ln 𝑥𝑥 = 𝑟𝑟 ln 𝑥𝑥 where r is a rational number
 46

Derivative of ln 𝑥𝑥:
d 1
(ln 𝑥𝑥 ) =
d𝑥𝑥 𝑥𝑥

For any positive number a,

d 1
(log 𝑎𝑎 𝑥𝑥 ) =
d𝑥𝑥 (ln 𝑎𝑎)𝑥𝑥
 47

Example 5.1
Differentiate the following functions with respect to x:
(i) 𝑦𝑦 = ln(sin2 𝑥𝑥 + 𝑥𝑥 2 + 1)
1
𝑦𝑦′ = 2 2
[2 sin 𝑥𝑥 cos 𝑥𝑥 + 2𝑥𝑥]
sin 𝑥𝑥 + 𝑥𝑥 + 1
(ii) 𝑦𝑦 = 𝑥𝑥ln(1 − 𝑥𝑥)
′
1 𝑥𝑥
𝑦𝑦 = ln(1 − 𝑥𝑥) + 𝑥𝑥 � (−1)� = ln(1 − 𝑥𝑥) +
1 − 𝑥𝑥 𝑥𝑥 − 1
(iii)𝑦𝑦 = [ln(𝑥𝑥 2 + 1)]3
1 6𝑥𝑥
𝑦𝑦 ′ = 3[ln(𝑥𝑥 2 + 1)]2 � 2 � (2𝑥𝑥) = 2 [ln(𝑥𝑥 2 + 1)]2
𝑥𝑥 + 1 𝑥𝑥 + 1
 48

B. Exponential Function
The exponential function:
e𝑥𝑥 = exp 𝑥𝑥 = ln−1 𝑥𝑥
Derivative of e𝑥𝑥 :
d 𝑥𝑥
(e ) = e𝑥𝑥
d𝑥𝑥
For any positive number a,
d 𝑥𝑥
(𝑎𝑎 ) = ln 𝑎𝑎 𝑎𝑎 𝑥𝑥
d𝑥𝑥
49
 50

Example 5.2
Differentiate the following functions with respect to x:
(a)
tan 𝑥𝑥 + 𝑥𝑥 3
𝑦𝑦 =
𝑒𝑒 𝑥𝑥
𝑒𝑒 𝑥𝑥 (sec 2 𝑥𝑥 + 3𝑥𝑥 2 ) − (tan 𝑥𝑥 + 𝑥𝑥 3 )𝑒𝑒 𝑥𝑥
𝑦𝑦′ =
𝑒𝑒 2𝑥𝑥
sec 2 𝑥𝑥 − tan 𝑥𝑥 − 𝑥𝑥 3 + 3𝑥𝑥 2
=
𝑒𝑒 𝑥𝑥
 51

(b)
𝑒𝑒 𝑥𝑥 tan 𝑥𝑥
𝑦𝑦 =
𝑥𝑥

( 𝑥𝑥 sec 2 𝑥𝑥 + tan 𝑥𝑥 𝑒𝑒 𝑥𝑥 ) − 𝑒𝑒 𝑥𝑥 tan 𝑥𝑥

𝑥𝑥 𝑒𝑒
𝑦𝑦 ′ =
𝑥𝑥 2
𝑒𝑒 𝑥𝑥 (𝑥𝑥 sec 2 𝑥𝑥 + 𝑥𝑥 tan 𝑥𝑥 − tan 𝑥𝑥 )
=
𝑥𝑥 2
 52

(c)
1
sin𝑥𝑥
𝑦𝑦 = 𝑒𝑒
1
sin𝑥𝑥 1 1
𝑦𝑦′ = 𝑒𝑒 cos �− 2 �
𝑥𝑥 𝑥𝑥

(d)
𝑥𝑥
𝑦𝑦 = 4ln 𝑥𝑥

1
𝑥𝑥 ln 𝑥𝑥 − 𝑥𝑥 � � 𝑥𝑥 ln 𝑥𝑥 − 1
′
𝑦𝑦 = (ln 4)4ln 𝑥𝑥 𝑥𝑥 = (ln 4)4ln 𝑥𝑥
(ln 𝑥𝑥) 2 (ln 𝑥𝑥 )2
 53

(e)
𝑓𝑓(𝑥𝑥) = ln �tan( 𝑒𝑒 √𝑥𝑥 )�
1 1
𝑓𝑓 ′ (𝑥𝑥) = sec 2
�𝑒𝑒 √𝑥𝑥 � 𝑒𝑒 √𝑥𝑥
tan( 𝑒𝑒 √𝑥𝑥 ) 2√𝑥𝑥
𝑒𝑒 √𝑥𝑥
=
2√𝑥𝑥 sin(𝑒𝑒 √𝑥𝑥 ) cos(𝑒𝑒 √𝑥𝑥 )
𝑒𝑒 √𝑥𝑥
=
√𝑥𝑥 sin(2𝑒𝑒 √𝑥𝑥 )

(d)
(𝑒𝑒 𝑥𝑥 )
𝑦𝑦 = 𝑒𝑒
(𝑒𝑒 𝑥𝑥 ) 𝑥𝑥
𝑦𝑦′ = 𝑒𝑒 𝑒𝑒
 54

C. Logarithmic Differentiation
1. For functions involving a lot of products, quotients or powers of positive
functions, instead of the tedious calculation involving product rule or quotient
rules, we may apply logarithmic differentiation.
Steps
(i) Taking logarithm on both sides of the equation 𝑦𝑦 = 𝑓𝑓(𝑥𝑥) (>0). Applying
logarithm laws, the expressions can be simplified to a sum of logarithms.
(ii) Differentiate both sides of the equation with respect to x, treating y as a
1 d𝑦𝑦
differentiable function of x: =⋯
𝑦𝑦 d𝑥𝑥
(iii)Hence, expressing y in terms of x,
d𝑦𝑦
= 𝑓𝑓(𝑥𝑥)(…)
d𝑥𝑥
 55

Example 5.3
d𝑦𝑦
Find the derivative if
d𝑥𝑥
(𝑥𝑥 2 + 1)(𝑥𝑥 + 3)1/2
𝑦𝑦 =
𝑥𝑥 − 1
for x > 1.
Solution:
We take the logarithm of both sides to get
2
1
)
ln 𝑦𝑦 = ln(𝑥𝑥 + 1 + ln(𝑥𝑥 + 3) − ln(𝑥𝑥 − 1)
2
Differentiate both sides of the equation,
1 ′ 2𝑥𝑥 1 1
𝑦𝑦 = 2 + −
𝑦𝑦 𝑥𝑥 + 1 2(𝑥𝑥 + 3) 𝑥𝑥 − 1
( 2 )( )1/2
′
𝑥𝑥 + 1 𝑥𝑥 + 3 2𝑥𝑥 1 1
𝑦𝑦 = � 2 + − �
𝑥𝑥 − 1 𝑥𝑥 + 1 2(𝑥𝑥 + 3) 𝑥𝑥 − 1
 56

Example 5.4
d𝑦𝑦
Find the derivative if
d𝑥𝑥
3
𝑦𝑦 = 𝑥𝑥 3 (2 sin 𝑥𝑥 − 4
1) � √𝑥𝑥 + 1�
for π/6 < x < π/2.
 57

2. Logarithmic differentiation can also be applied to functions where both

the base and radical are in the variable x.
Example 5.5
d𝑦𝑦
(i) Find the derivative if
d𝑥𝑥
𝑦𝑦 = 𝑥𝑥 𝑥𝑥
for x > 0.
Solution:
We take the natural logarithm of both sides of the equation,
ln 𝑦𝑦 = 𝑥𝑥 ln 𝑥𝑥
1 ′ 1
𝑦𝑦 = ln 𝑥𝑥 + 𝑥𝑥
𝑦𝑦 𝑥𝑥
𝑦𝑦 ′ = 𝑦𝑦(ln 𝑥𝑥 + 1) = 𝑥𝑥 𝑥𝑥 (ln 𝑥𝑥 + 1)
 58

d𝑦𝑦
(ii) Find the derivative if
d𝑥𝑥
𝑦𝑦 = 𝑥𝑥 1/𝑥𝑥
for x > 0.

Supp. Ex. 5 Q11

 59

VI. Implicit Differentiation

Sometimes, the function y may not be defined explicitly as ‘y = f(x)’.
Instead, we may have an implicit relation between the variables x and y.
It is effective to apply ‘implicit differentiation’ for these cases:

Steps
1. Differentiate both sides of the equation with respect to x, treating y as a
differentiable function of x.
d𝑦𝑦
2. Collect terms and rewrite the equation as (…) =⋯
d𝑥𝑥
d𝑦𝑦
3. Solve for ‘algebraically’.
d𝑥𝑥
 60

Example 6.1
Find the slope of the circle 𝑥𝑥 2 + 𝑦𝑦 2 = 25 at the point (3, −4).
a. Method 1: explicit b. Method 2: implicit
Hence, find the equation of the tangent line to the circle at the point (3, −4).
Solution:
a. 𝑦𝑦 = −√25 − 𝑥𝑥 2 (exercise)
b. Differentiate both sides of the equation
𝑥𝑥 2 + 𝑦𝑦 2 = 25
2𝑥𝑥 + 2𝑦𝑦𝑦𝑦′ = 0
𝑥𝑥
𝑦𝑦 ′ = −
𝑦𝑦
𝑥𝑥 3 3
Thus, the slope of the circle at (3, −4) is − = − =
𝑦𝑦 −4 4
3 25
Hence, the equation of the tangent line is 𝑦𝑦 = 𝑥𝑥 −
4 4
 61

Example 6.2
d𝑦𝑦
Find if 𝑦𝑦 2 = 𝑥𝑥 2 + sin 𝑥𝑥𝑥𝑥.
d𝑥𝑥

Solution:
We differentiate both sides of the equation 𝑦𝑦 2 = 𝑥𝑥 2 + sin 𝑥𝑥𝑥𝑥,
we get 2𝑦𝑦𝑦𝑦 ′ = 2𝑥𝑥 + (cos 𝑥𝑥𝑥𝑥) (𝑥𝑥𝑦𝑦 ′ + 𝑦𝑦)
(2𝑦𝑦 − 𝑥𝑥 cos 𝑥𝑥𝑥𝑥)𝑦𝑦 ′ = 2𝑥𝑥 + 𝑦𝑦 cos 𝑥𝑥𝑥𝑥
2𝑥𝑥 + 𝑦𝑦 cos 𝑥𝑥𝑥𝑥
𝑦𝑦′ =
2𝑦𝑦 − 𝑥𝑥 cos 𝑥𝑥𝑥𝑥
 62

Example 6.3
d𝑦𝑦 𝑥𝑥−1
Find if 𝑦𝑦 = tan−1 � �
d𝑥𝑥 𝑥𝑥+1

Supp. Ex. 5 Q12-13

 63

VII. Higher Derivatives

If f(x) is a differentiable function, then the derivative f’(x) is also a function.
We can define derivative of the function f’(x).

Consider an example from physics:

- Displacement
- Velocity = derivative of displacement
- Acceleration = derivative of velocity, second derivative of displacement
 64

Second Derivative
Given 𝑦𝑦 = 𝑓𝑓(𝑥𝑥) , the second derivative is defined as the derivative of the first
derivative, that is
d d𝑦𝑦
� �
d𝑥𝑥 d𝑥𝑥
d2 𝑦𝑦
Notation: 𝑓𝑓"(𝑥𝑥), , 𝑦𝑦"
d𝑥𝑥 2

In general, for any positive integer n, the nth derivative of y with respect to x is
denoted by,
𝑛𝑛
(𝑛𝑛)
d 𝑦𝑦
𝑦𝑦 , 𝑛𝑛
d𝑥𝑥
and it is defined by
d
𝑦𝑦 (𝑛𝑛) = 𝑦𝑦 (𝑛𝑛−1) .
d𝑥𝑥
 65

Example 7.1
Let
𝑦𝑦 = 𝑥𝑥 3 − 3𝑥𝑥 2 + 2
Find 𝑦𝑦 (4) .

Solution:
𝑦𝑦′ = 3𝑥𝑥 2 − 6𝑥𝑥
𝑦𝑦′′ = 6𝑥𝑥 − 6
𝑦𝑦 (3) = 6
𝑦𝑦 (4) = 0
 66

Example 7.2 - Simple Harmonic Motion

A body hanging from a spring is stretched 5 units beyond its rest position and
released at time t = 0 to bob up and down. Its position at any later time t is
𝑠𝑠 = 5 cos 𝑡𝑡 .
What are its velocity and acceleration at time t?
 67

Example 7.3
d2 𝑦𝑦
Find if
d𝑥𝑥 2

2𝑥𝑥 3 − 3𝑦𝑦 2 = 8
Solution:
Differentiate both sides of the equation 2𝑥𝑥 3 − 3𝑦𝑦 2 = 8, we have
6𝑥𝑥 2 − 6𝑦𝑦𝑦𝑦′ = 0
𝑥𝑥 2
𝑦𝑦′ =
𝑦𝑦
for 𝑦𝑦 ≠ 0.
Differentiate a second time to obtain the second derivative
2
2 𝑥𝑥
𝑦𝑦 (2𝑥𝑥 ) − 𝑥𝑥 2 ′
𝑦𝑦 2𝑥𝑥𝑥𝑥 − 𝑥𝑥 2𝑥𝑥 𝑥𝑥 4
𝑦𝑦
𝑦𝑦 ′′ = = = −
𝑦𝑦 2 𝑦𝑦 2 𝑦𝑦 𝑦𝑦 3
for 𝑦𝑦 ≠ 0.
 68

Example 7.4
Suppose the equation
𝑦𝑦 3 + cos(𝑥𝑥𝑥𝑥) = 𝑥𝑥 2
Implicitly determined a function 𝑦𝑦 = 𝑦𝑦(𝑥𝑥) near the point (0, −1).
d𝑦𝑦
(a) By implicit differentiation, find at (0, −1).
d𝑥𝑥
Solution:
Differentiate both sides of the equation with respect to x
3𝑦𝑦 2 𝑦𝑦 ′ − sin(𝑥𝑥𝑥𝑥) (𝑦𝑦 + 𝑥𝑥𝑥𝑥′) = 2𝑥𝑥
2𝑥𝑥 + 𝑦𝑦 sin(𝑥𝑥𝑥𝑥)
𝑦𝑦 ′ = 2
3𝑦𝑦 − 𝑥𝑥 sin(𝑥𝑥𝑥𝑥)
d𝑦𝑦
� =0
d𝑥𝑥 (0,−1)
 69

d2 𝑦𝑦
(b) By implicit differentiation, find at (0, −1).
d𝑥𝑥 2
Solution:
Differentiate both sides of the following equation with respect to x
(3𝑦𝑦 2 − 𝑥𝑥 sin(𝑥𝑥𝑥𝑥))𝑦𝑦 ′ = 2𝑥𝑥 + 𝑦𝑦 sin(𝑥𝑥𝑥𝑥)
we have
d 2
d𝑦𝑦 2
d2 𝑦𝑦 d𝑦𝑦 d𝑦𝑦
(3𝑦𝑦 − 𝑥𝑥 sin(𝑥𝑥𝑥𝑥)) + (3𝑦𝑦 − 𝑥𝑥 sin(𝑥𝑥𝑥𝑥)) 2 = 2 + sin(𝑥𝑥𝑥𝑥) + 𝑦𝑦 cos(𝑥𝑥𝑥𝑥)(𝑦𝑦 + 𝑥𝑥 )
d𝑥𝑥 d𝑥𝑥 d𝑥𝑥 d𝑥𝑥 d𝑥𝑥
d𝑦𝑦
With � = 0,
d𝑥𝑥 (0,−1)

d2 𝑦𝑦
3 2
� = 2 + (−1)2 cos(0)
d𝑥𝑥 (0,−1)

d2 𝑦𝑦
2
� =1
d𝑥𝑥 (0,−1)
 70

Example 7.5
For 𝑦𝑦 = sin(𝑥𝑥), find 𝑦𝑦 (𝑛𝑛) .

Solution:
sin 𝑥𝑥 𝑖𝑖𝑖𝑖 𝑛𝑛 = 4𝑘𝑘
cos 𝑥𝑥 𝑖𝑖𝑖𝑖 𝑛𝑛 = 4𝑘𝑘 + 1
𝑦𝑦 (𝑛𝑛) =�
−sin 𝑥𝑥 𝑖𝑖𝑖𝑖 𝑛𝑛 = 4𝑘𝑘 + 2
−cos 𝑥𝑥 𝑖𝑖𝑖𝑖 𝑛𝑛 = 4𝑘𝑘 + 3
𝑛𝑛𝑛𝑛
= sin �𝑥𝑥 + �
2
 71

Example 7.6
For 𝑦𝑦 = 𝑒𝑒 𝑎𝑎𝑎𝑎 ,
𝑦𝑦 (𝑛𝑛) = 𝑎𝑎𝑛𝑛 𝑒𝑒 𝑎𝑎𝑎𝑎
 72

Example 7.7
For 𝑦𝑦 = 𝑥𝑥 𝑚𝑚 , find 𝑦𝑦 (𝑛𝑛) .
 73

Example 7.8
𝑥𝑥
Let 𝑓𝑓(𝑥𝑥) = , find 𝑓𝑓 (𝑛𝑛) (0).
1−𝑥𝑥 2
Solution:
We first note that we can rewrite f as (by partial fraction decomposition)
𝑥𝑥 1 1 1
2
= �− + �
1 − 𝑥𝑥 2 1 + 𝑥𝑥 1 − 𝑥𝑥
Furthermore,
d𝑛𝑛 1 (−1)𝑛𝑛 𝑛𝑛!
� �=
d𝑥𝑥 1 + 𝑥𝑥
𝑛𝑛 (1 + 𝑥𝑥)𝑛𝑛+1
d𝑛𝑛 1 𝑛𝑛!
� �=
d𝑥𝑥 1 − 𝑥𝑥
𝑛𝑛 (1 − 𝑥𝑥)𝑛𝑛+1
 74

Hence,
1 ( −1 ) 𝑛𝑛+1 𝑛𝑛! 𝑛𝑛!
(𝑛𝑛) ( )
𝑓𝑓 𝑥𝑥 = � 𝑛𝑛+1
+ 𝑛𝑛+1
�
(
2 1 + 𝑥𝑥 ) ( 1 − 𝑥𝑥 )
and
(𝑛𝑛) ( 1
𝑓𝑓 0) = ((−1)𝑛𝑛+1 𝑛𝑛! + 𝑛𝑛!)
2
𝑛𝑛! if 𝑛𝑛 is odd
=�
0 if 𝑛𝑛 is even
 75

Example 7.9 (Leibniz’s Formula)

The Leibniz’s Formula is an extended version of the product rule to the
higher derivatives of the product of two functions:
𝑛𝑛
(𝑢𝑢𝑢𝑢 )(𝑛𝑛) = � C𝑘𝑘𝑛𝑛 𝑢𝑢(𝑘𝑘) 𝑣𝑣 (𝑛𝑛−𝑘𝑘)
𝑘𝑘=0
where we denote u by 𝑢𝑢(0) and similarly for v. The formula can be proved
by mathematical induction.
 76

For example, consider differentiating 𝑢𝑢(𝑥𝑥)𝑣𝑣(𝑥𝑥) = (𝑥𝑥 2 + 3) cos 𝑥𝑥.

With n = 1, we have the usual product rule
1
[𝑢𝑢(𝑥𝑥)𝑣𝑣 (𝑥𝑥)]′ = � C𝑘𝑘1 𝑢𝑢(𝑘𝑘) 𝑣𝑣 (1−𝑘𝑘)
𝑘𝑘=0
= 𝐶𝐶01 𝑢𝑢𝑣𝑣 ′ + 𝐶𝐶11 𝑢𝑢′ 𝑣𝑣
= (𝑥𝑥 2 + 3)(− sin 𝑥𝑥) + (2𝑥𝑥) cos 𝑥𝑥
Let’s apply the formula to n = 2,
2
[𝑢𝑢(𝑥𝑥)𝑣𝑣 (𝑥𝑥)](2) = � C𝑘𝑘2 𝑢𝑢(𝑘𝑘) 𝑣𝑣 (2−𝑘𝑘)
𝑘𝑘=0
= 𝐶𝐶02 𝑢𝑢𝑣𝑣 ′′ + 𝐶𝐶12 𝑢𝑢′𝑣𝑣 ′ + 𝐶𝐶22 𝑢𝑢′′ 𝑣𝑣
= (𝑥𝑥 2 + 3)(− cos 𝑥𝑥 ) + 2(2𝑥𝑥)(− sin 𝑥𝑥 ) + 2 cos 𝑥𝑥
 77

We go further to try n = 3,
3
[𝑢𝑢(𝑥𝑥)𝑣𝑣 (𝑥𝑥)](3) = � C𝑘𝑘3 𝑢𝑢(𝑘𝑘) 𝑣𝑣 (3−𝑘𝑘)
𝑘𝑘=0
= 𝐶𝐶03 𝑢𝑢(0) 𝑣𝑣 (3) + 𝐶𝐶13 𝑢𝑢(1) 𝑣𝑣 (2) + 𝐶𝐶23 𝑢𝑢(2) 𝑣𝑣 (1) + 𝐶𝐶33 𝑢𝑢(3) 𝑣𝑣 (0)
= (𝑥𝑥 2 + 3)(sin 𝑥𝑥) + 3(2𝑥𝑥)(− cos 𝑥𝑥) + 3(2)(− sin 𝑥𝑥 )
+ (0)cos 𝑥𝑥
= (𝑥𝑥 2 − 3)(sin 𝑥𝑥) − 6𝑥𝑥 cos 𝑥𝑥
• As we noted in the previous example, 𝑢𝑢(𝑘𝑘) = 0 for k > 2.
• Due to this fact, we can easily continue to differentiate the product (say,
[𝑢𝑢(𝑥𝑥)𝑣𝑣 (𝑥𝑥)](10) , …) and only a few terms in the sum are non-zero!
 78

10
[𝑢𝑢(𝑥𝑥 )𝑣𝑣(𝑥𝑥 )](10) = � C𝑘𝑘10 𝑢𝑢(𝑘𝑘) 𝑣𝑣 (10−𝑘𝑘)
𝑘𝑘=0

= 𝐶𝐶010 𝑢𝑢(0) 𝑣𝑣 (10) + 𝐶𝐶110 𝑢𝑢(1) 𝑣𝑣 (9) + 𝐶𝐶210 𝑢𝑢(2) 𝑣𝑣 (8) + 0

= (𝑥𝑥 2 + 3)(− cos 𝑥𝑥 ) + 10(2𝑥𝑥 )(−sin 𝑥𝑥 ) + (45)(2)(cos 𝑥𝑥 )

Note that we have applied the technique in Example 7.5 to compute 𝑣𝑣 (8) , 𝑣𝑣 (9) , 𝑣𝑣 (10) .
 79

As another example, consider 𝑦𝑦 = 𝑥𝑥𝑥𝑥 𝑎𝑎𝑎𝑎 , then

𝑛𝑛

𝑦𝑦 (𝑛𝑛) = � C𝑛𝑛𝑘𝑘 𝑥𝑥(𝑘𝑘) (𝑒𝑒𝑎𝑎𝑎𝑎 )(𝑛𝑛−𝑘𝑘)

𝑘𝑘=0

= C𝑛𝑛0 𝑥𝑥(𝑒𝑒𝑎𝑎𝑎𝑎 )(𝑛𝑛) + C𝑛𝑛1 (1)(𝑒𝑒𝑎𝑎𝑎𝑎 )(𝑛𝑛−1)

= 𝑥𝑥𝑎𝑎𝑛𝑛 𝑒𝑒𝑎𝑎𝑎𝑎 + 𝑛𝑛𝑎𝑎𝑛𝑛−1 𝑒𝑒𝑎𝑎𝑎𝑎
= 𝑎𝑎𝑛𝑛−1 𝑒𝑒𝑎𝑎𝑎𝑎 (𝑥𝑥𝑥𝑥 + 𝑛𝑛)
 80

Example 7.10
For 𝑦𝑦(𝑥𝑥) = (cos −1 𝑥𝑥)2 , show that (1 − 𝑥𝑥 2 )𝑦𝑦 ′′ − 𝑥𝑥𝑦𝑦 ′ − 2 = 0. Hence,
show that (1 − 𝑥𝑥 2 )𝑦𝑦 (𝑛𝑛+2) − (2𝑛𝑛 + 1)𝑥𝑥𝑦𝑦 (𝑛𝑛+1) − 𝑛𝑛2 𝑦𝑦 (𝑛𝑛) = 0.
Solution:
−1
By direct computation, 𝑦𝑦 ′ = 2 cos −1 𝑥𝑥 � �
√1−𝑥𝑥 2
which implies √1 − 𝑥𝑥 2 𝑦𝑦 ′ + 2 cos −1 𝑥𝑥 = 0.
Implicitly differentiate both sides of the equation, we have
′′
1 ′
2
�1 − 𝑥𝑥 𝑦𝑦 +
2 (−2𝑥𝑥)𝑦𝑦 − =0
2√1 − 𝑥𝑥 2 √1 − 𝑥𝑥 2

(1 − 𝑥𝑥 2 )𝑦𝑦 ′′ − 𝑥𝑥𝑦𝑦 ′ − 2 = 0 (*)

 81

To show the second identity, we differentiate both sides of (*) n times.

[(1 − 𝑥𝑥 2 )𝑦𝑦 ′′ ](𝑛𝑛) − (𝑥𝑥𝑦𝑦 ′ )(𝑛𝑛) = 0
By the Leibniz’s Formula,
𝑛𝑛 𝑛𝑛

� C𝑘𝑘𝑛𝑛 (1 − 𝑥𝑥 2 )(𝑘𝑘) 𝑦𝑦 (𝑛𝑛−𝑘𝑘+2) − � C𝑘𝑘𝑛𝑛 𝑥𝑥 (𝑘𝑘) 𝑦𝑦 (𝑛𝑛−𝑘𝑘+1) = 0

𝑘𝑘=0 𝑘𝑘=0

2 ) (𝑛𝑛+2) (𝑛𝑛+1)
𝑛𝑛(𝑛𝑛 − 1)
(1 − 𝑥𝑥 𝑦𝑦 − 2𝑛𝑛𝑛𝑛𝑦𝑦 + (−2)𝑦𝑦 (𝑛𝑛) − 𝑥𝑥𝑦𝑦 (𝑛𝑛+1) − 𝑛𝑛𝑦𝑦 (𝑛𝑛) = 0
2
(1 − 𝑥𝑥 2 )𝑦𝑦 (𝑛𝑛+2) − (2𝑛𝑛 + 1)𝑥𝑥𝑦𝑦 (𝑛𝑛+1) − 𝑛𝑛2 𝑦𝑦 (𝑛𝑛) = 0
 82

SUMMARY (Chapter 4)
A.Definition of derivative by the first principle

B. Differentiability of functions (for example, piecewise functions)

C. Basic rules of differentiation:

1.sum rule
2.product rule
3.quotient rule
4.power rule
5.chain rule

D.Derivatives of trigonometric functions

 83

E. Derivatives of exponential and logarithmic functions

F. Derivative of inverse function

G.Skills:
1.Logarithmic differentiation
2.Implicit differentiation

H.Higher derivatives

Reference: Thomas’ Calculus Chapter 3, 7.

Supplementary Exercise 5 Q1-13.