Calculus I Chapter 3.

Differentiation Rules

3.5. Implicit Differentiation

When a function is expressed in the form y = f (x), we may refer to this as explicit description.

But sometimes, we may encounter functions in the form that may not be explicit such as

x2 + y 2 = 9 y ≥ 0 and x3 + 6xy + y 3 = 4.

√ √
The first could be solved for y as y = ± 9 − x2 and since we are given y ≥ 0, we get y = f (x) = 9 − x2 .

The second equation is not that simple to solve for y in terms of x.

In this section, we assume that an equation in x and y defines y as function of x in some interval, and find the derivative of y
in terms of x. The method we develop for such problems is referred to as implicit differentiation.

We will do several example to show how this method works.

dy
Examples. 1) In the following equations, assume that the given equations define y as a function of x. Find dx .

a) x2 + y 3 = 4x − 1

b) x3 y 2 − 6xy = 9
 x y
c) y + x = 2y

d) xey = 2xy − y 2 .

2) The curve x3 + y 3 = 6xy is called the Folium of Descartes. Find an equation of the tangent line at the point (3, 3).

2

3) The lemniscate curve x2 + y 2 = 4 x2 − y 2 was discovered by Jacob Bernoulli in 1694. Draw the graph and find the
points where it has horizontal tangents.
 3.6. Derivatives of Logarithmic and Inverse Trigonometric Functions

This is our last section where we develop methods to find derivatives of common functions.

We will use the method of implicit differentiation of the previous section to find derivative of an inverse function.

We then use this result to find the formula for the derivatives of logarithmic functions and inverse trig functions.

Derivatives of Logarithmic Functions

Logarithmic Differentiation

Derivatives of Inverse Trigonometric Functions

Theorem 1(Derivative of the Inverse). Assume that f (x) is differentiable and one-to-one with inverse g(x) = f −1 (x) . If
b belongs to the domain of Dom g and f 0 (g(b)) 6= 0 , then g 0 (b) exists and
1
g 0 (b) = 0
f (g(b))
.

Examples 1) Find g 0 (x) where g(x) is the inverse of f (x) = x2 + 2, x > 0.

2) Find g 0 (1) where g(x) is the inverse of f (x) = x + ex .

 Theorem 2. If b > 0 and b 6= 1, then
d x
(b ) = (ln b)bx
dx
.

Example Find the derivatives.

a) f (x) = 3−4x

3
b) g(x) = 3x .

Theorem 3. The derivative the natural logarithmic function is given by

d 1
(ln x) = .
dx x
.

Also, if g(x) is a differentiable function, then

d g 0 (x)
(ln(g(x)) = .
dx g(x)
.

Example Find the derivative of the given functions.

a) f (x) = ln(7x).

 
x−1
b) g(x) = ln x+1
 
2 
c) h(x) = ln x3 + x − 1 ex +3x−4

Example (Logarithmic Differentiation) Find the derivative of

G(x) = xx .
Derivatives of Inverse Trigonometric Functions

Theorem 4 (Derivatives of Arcsine and Arccosine.)

d 1
sin−1 x = √


dx 1 − x2
and
d −1
cos−1 x = √


.
dx 1 − x2

Theorem 5. The derivatives of the other inverse trig functions are given by

d
tan−1 x = 1


a) dx 1+x2 .

−1
d
cot−1 x =


b) dx 1+x2 .

d
sec−1 x = √1


c) dx |x| x2 −1
.

−1
d
csc−1 x =


d) dx
√
|x| x2 −1
.
Examples Find the derivative of the given functions.

a) f (x) = sin−1 (7x).

b) g(x) = cos−1 x3 − 1 .

 2
c) h(x) = cos−1 ex .

 
d) F (x) = tan−1 x−1
x+1 .
 3.7. Rates of Change in the Sciences

Physics

Chemistry

Biology

Economics

Physics

If s = f t) is the position function of a particle that is moving in a straight line, then

∆s
∆t
represents the average velocity over a time period ∆t and
ds
v=
dt
represents the instantaneous velocity (the rate of change of displacement with respect to time). The instantaneous rate of change
of velocity with respect to time is acceleration:
a(t) = v 0 (t) = s00 (t).

Now that we know the differentiation formulas, we are able to more easily solve problems involving the motion of objects.
 Example 1 The position of a particle is given by the equation
s = f (t) = t3 − 5t2 + 6t,
where t is measured in seconds and s in meters. Here the graph of s:

a) Find the velocity at time t

b) What is the velocity after 2 seconds? After 4?

c) When is the particle at rest?

d) When is the particle moving forward (that is, in the positive direction)?

e) Find the total distance traveled by the particle during the first five seconds.

f) Find the acceleration at time t and and after 4 seconds.

g) Graph the position, velocity, and acceleration functions for 0 ≤ t ≤ 5

h) When is the particle speeding up? When is it slowing down?

 Example 2. If a rod or piece of wire is homogeneous, then its linear density is uniform and is defined as the mass per
unit length (ρ = m/l) and measured in kilograms per meter. Suppose, however, that the rod is not homogeneous but that its
mass measured from its left end to a point x is m = f (x), as shown in below:

The mass of the rod between x = x1 and x = x2 is ∆m = f (x2 ) − f (x1 ) and so the average density of the rod is

∆m f (x2 (−f (x1 )

average density = =
∆x x2 − x1

If we now let ∆x → 0, (that is, x2 → x1 ),, we are computing the average density over smaller and smaller intervals. The linear
density
∆m dm
ρ(x) = lim = = f 0 (x).
∆x→0 ∆x dx

Thus, the linear density is the rate of change of mass with respect to length.

Example. Suppose the mass of a rod of length 3 is given f (x) = x ln(x + 1), where 0 ≤ x ≤ 3.

a) Find the average density of the rod over the interval [1, 2],

b) Find the density of the rod at x = 0, 1, 2, 3.

 Chemistry

Example 1. A chemical reaction results in the formation of one or more substances (called products) from one or more
starting materials (called reactants). For instance, the “equation”

2H1 + O2 =⇒ 2H2 O

indicates that two molecules of hydrogen and one molecule of oxygen form two molecules of water.

Let us consider the reaction

A + B =⇒ C,
where A and B are the reactants and C is the product. The concentration of a reactant A is the number of moles (1 mole
= 6.022 × 1023 molecules per liter and is denoted by [A]. The concentration varies during a reaction, so [A], [B], and [C] are all
functions of time t. The average rate of reaction of the product C over a time interval t1 ≤ t ≤ t2 is

∆[C] [C](t2 − [C](t1

=
∆t t2 − t1
But chemists are interested in the instantaneous rate of reaction because it gives information about the mechanism of the
chemical reaction. The instantaneous rate of reaction is obtained by taking the limit of the average rate of reaction as the time
interval ∆t approaches 0:

∆[C] d[C]
rate of reaction = lim = .
∆t→0 ∆t dt
Since the concentration of the product increases as the reaction proceeds, the derivative d[C]/dt will be positive, and so the
rate of reaction of [C] is positive. The concentrations of the reactants, however, decrease during the reaction, so, to make the
rates of reaction of A and B positive numbers, we put minus signs in front of the derivatives d[A]/dt and d[B]/dt. Since [A] and
[B] each decrease at the same rate that [C] increases, we have

d[C] d[A] d[B]

rate of reaction = =− =− .
dt dt dt

More generally, it turns out that for a reaction of the form

aA + bB =⇒ cC + dD

we have

1 d[C] 1 d[B] 1 d[C] 1 d[D]

− =− = = .
a dt b dt c dt d dt

The rate of reaction can be determined from data and graphical methods. In some cases there are explicit formulas for the
concentrations as functions of time that enable us to compute the rate of reaction.
 Example 2. One of the quantities of interest in thermodynamics is compressibility. If a given substance is kept at a constant
temperature, then its volume V depends on its pressure P . We can consider the rate of change of volume with respect to pressure,
that is , the derivative
dV
.
dP
As P increases, V decreases, so
dV
< 0.
dP

. The isothermal compressibility is defined by introducing a minus sign and dividing this derivative by the volume
1 dV
sothermal compressibility = β = − .
V dP
Thus β (read as beta) measures how fast, per unit volume, the volume of a substance decreases as the pressure on it increases
at constant temperature.
 Biology

Example Let n = f (x) be the number of individuals in an animal or plant population at time t. The change in the population
size between the times t = t1 and t = t2 is ∆n = f (t2 ) − f (t1 ), and so the average rate of growth during the time period t1 ≤ t ≤ t2
is

∆ f (t2 ) − f (t1 )
average rate of growth = =
∆t t2 − t1
The instantaneous rate of growth is obtained from this average rate of growth by letting the time period ∆t approach 0:
∆n dn
growth rate = lim = .
∆t→0 ∆t dt
Remark: Strictly speaking, this is not quite accurate because the actual graph of a population function n = f (x) would
be a step function that is discontinuous whenever a birth or death occurs and therefore not differentiable. However, for a large
animal or plant population, we can replace the graph by a smooth approximating curve as in the figure below

For more on this and another example from biology, read Example 6 and 7 from page 230 to 232 of your text.
 Economics

Marginal Cost in Economics Many products can be manufactured more cheaply if produced in large quantities. The num-
ber of units manufactured is called the production level. To study the relation between costs and production level, economists
define a rate of change called marginal cost.

Let C(x) denote the dollar cost (including labor and parts) of producing x units of a particular product. The marginal cost
at production level x0 is the cost of producing one additional unit:

Marginal Cost = C(x0 + 1) − C(x0 ) ≈ C 0 (x0 )

Example (Cost of an Air Flight) Company data suggest that the total dollar cost of a 1,200-mile flight is approximately
C(x) = 0.0005x3 − 0.38x2 + 120x,
where x is the number of passengers.

(a) Estimate the marginal cost of an additional passenger if the flight already has 150 passengers.

(b) Compare your estimate with the actual cost.

(c) Is it more expensive to add a passenger when x = 150 or 200?

 Linear Motion

A linear motion refers to the motion of an object along a straight line (such as motion of a vehicle along a straight highway
or the vertical motion of a falling object).

Let s(t) denote the position or distance from the origin at time t. Velocity is the ROC of position with respect to time:

ds
Velocity = v(t) = .
dt

Speed is defined as the absolute value of velocity

Speed = |v(t)|.

The sign of v(t) indicates the direction of motion.

For example, if s(t) is the height above ground, then v(t) > 0 indicates that the object is rising.

Example. A large truck enters the off-ramp of a freeway at t = 0. Its position after t seconds is s(t) = 84t − t3 ft for
0 ≤ t ≤ 5.

(a) How fast is the truck going at the moment it enters the off-ramp?

(b) Is the truck speeding up or slowing down?

 3.8. Exponential Growth and Decay

Population Growth

Radioactive Decay

Newton’s Law of Cooling

Continuously Compounded Interest

Population Growth Suppose y(t) is the value of a quantity at time t and the rate of change of y with respect to t is
proportional to its size y(t) at any time, then

y 0 (t) = ky

where k is a constant. This equation is sometimes called the law of natural growth (if k > 0 and the law of natural decay
(if k < 0. It is called a differential equation because it involves an unknown function y and its derivative y 0 (t).

From the fact that

d kt

e = kekt ,
dt

we can conclude that

y(t) = Cekt
is a solution of y(t) = ky(t), for any constant C. If we know the initial population is y0 , then y(0) = y0 and it follows that

y(0) = Cek(0) = y0 and hence C = y0 = initial population.

 Example. Use the fact that the world population was 2560 million in 1950 and 3040 million in 1960 to model the population
of the world in the second half of the 20th century. (Assume that the growth rate is proportional to the population size.) What is
the relative growth rate? Use the model to estimate the world population in 1993 and to predict the population in the year 2025.

Solution If P (t) is the world population at time t, then our model is

P (t) = P0 ekt ,

where P( 0) is the initial population and k is the ”growth constant”. If we take time t = 0 to correspond to the year 1950, then
P0 = 2560 (million). So, our model now becomes
P (t) = 2560ekt .
To find k, we use the second information about the population. The year 1960 corresponds to t = 10 and so P (10) = 3040
(million). Hence
P (10) = 3040 = 2560ek×10 .
We now solve for t. First divide 3040 = 2560ek×10 by 2560 to
3040
= e10k
2560
and take the natural log of both sides to get  
1 3040
k= ln ≈ 0.017185.
10 2560
Thus the population at any time is
P (t) = 2560e0.017185t
and, since 2025 corresponds to t = 75, we estimate the population to be

P (75) = 2560e0.017185 ≈ 9289 million

 Radioactive Decay Radioactive substances decay by spontaneously emitting radiation. If m(t) is the mass remaining from
an initial mass m0 of the substance after time t, then the relative decay rate
dm/dt
−
m
has been found experimentally to be constant. (Since dm/dt is negative, the relative decay rate is positive.) It follows that
dm
= km
dt
where k < 0 is the decay constant. As in the previous example, we see that

m(t) = m0 ekt , k < 0.

Example The half-life of radium-226 is 1590 years.

(a) A sample of radium- 226 has mass 100 mg. Find a formula for the mass of the sample that remains after t years.

(b) Find the mass remaining after 1000 years correct to the nearest milligram

(c) When will the mass be reduced to 30 mg?

Solution. a) Here m(0) = m0 = 100. Hence

m(t) = 100ekt .
To find k we use the information about half-life of radium-226 being 1590. This means
1
m(1590) = m(0)
2

Thus m(1590) = 1/2 × 100 = 50. This gives

 
1590k 1590k 1 1 1 ln 2
100e = 50 or e = or k= ln =− ≈ −0.0004359.
2 1590 2 1590

Hence

m(t) = 100e−0.0004359t .

You can now answer b) and c).

 Newton’s Law of Cooling states that the rate of cooling of an object is proportional to the temperature difference between
the object and its surroundings, provided that this difference is not too large. (This law also applies to warming.) If we let T (t)
be the temperature of the object at time t and Ts be the temperature of the surroundings, then we can formulate Newton’s Law
of Cooling as a differential equation:

dT
= k(T − T0 ),
dt

where k is a constant. This equation is not quite the same as y 0 = ky, so we make the change of variable

y = T − Ts .

Since Ts is a constant, we have y 0 (t) = T 0 (t) and so the equation T 0 (t) = kT (t) becomes y 0 (t) = ky(t), whose solution is y(t) = Cekt .
Since y(t) = T − Ts , we see that

T (t) = Ts + Cekt

Exercise. A bottle of iced tea at room temperature (72◦ F ) is placed in a refrigerator where the temperature is 44◦ F . After
half an hour the tea has cooled to 61◦ F .

(a) What is the temperature of the tea after another half hour?

(b) How long does it take for the tea to cool to 50◦ F ?
 3.9. Related Rates

Example 1. A 5-meter ladder leans against a wall. The bottom of the ladder is 1.5 meters from the wall at time t = 0 and
slides away from the wall at a rate of 0.8m/s. Find the velocity of the top of the ladder at time t = 1.

Example 2. Water pours into a fish tank at a rate of 0.3m3 /min. How fast is the water level rising if the base of the tank
is a rectangle of dimensions 2 × 3 meters?
 Example 3. Assume that the radius r of a sphere is expanding at a rate of 30 cm/min. The volume of a sphere is V = 43 πr3
and its surface area is S = 4πr2 .

a) Find the rate of the volume with respect to time when r = 15 cm.

b) Find the rate of the volume with respect to time at t = 2 min, assuming that r = 0 at t = 0.

c) Find the rate of the surface area with respect to time when r = 40cm.

d) Find the rate of the surface area with respect to time at t = 2min, assuming that r = 10 at t = 0.
 Example 4. A police car traveling south toward Sioux Falls, Iowa, at 160 km/h pursues a truck traveling east away from
Sioux Falls at 140 km/h. At time t = 0, the police car is 20 km north and the truck is 30 km east of Sioux Falls. Calculate the
rate at which the distance between the vehicles is changing: At time

a) t=0 b) t = 5 min later

Exercise 5. One ship is approaching a port from the north at 15 miles per hour, and is presently 3 miles north of the port.
A second ship has already left the port, traveling east at 20 miles per hour, and is presently 4 miles east of the port. At this
instant, what is the rate of change of the distance between the two ships? Are they getting closer or further apart?
 3.10. Linear Approximations and Differentials

Linearization and Approximation

Applications to Physics

Differentials

The Linear Approximation uses the slope of the tangent line (i.e., the derivative) to estimate ∆f without computing it
exactly.

From
f (a + ∆x) − f (a) ∆f
f 0 (a) = lim = lim
∆x→0 ∆x ∆→0 ∆x

we have

∆f
≈ f 0 (a) or ∆f ≈ f 0 (a)∆x.
∆x

Example 1. Use the Linear Approximation to estimate the change in f (x) = 1/x as x goes from 10 to 10.2 ; that is, to
estimate
1 1
− .
10.2 10
How accurate is the estimate?

√
3
√
3
√
3
Example 2. Approximate how much greater 8.1 is than 8 = 2, and then use the result to approximate 8.1.
The linearization of f (x) centered at x = a is defined by

L(x) = f 0 (a)(x − a) + f (a).

Notice that y = L(x) is the equation of the tangent line at x = a. For values of x close to a,

f (x) ≈ L(x) or f (x) ≈ f (a) + f 0 (a)(x − a).

√
Example. Determine the approximation formula for f (x) = xex−1 resulting from the linearization at a = 1

Example Estimate tan(π/4 + 0.02). Use the fact that

error
percentage error = %
actual value

to compute the percentage error.

Differential Form of Linear Approximation: If f is differentiable at a and dx is small, then

∆y ≈ dy = f 0 (a)dx.

The error E in the above approximation is

E = error = |∆f − f 0 (a)∆x|

It can be proved that

1
E≤ K(∆x)2 , where K = max|f 00 (x)|.
2
 3.11. Hyperbolic Functions

Hyperbolic Functions and Their Derivatives

Inverse Hyperbolic Functions and Their Derivatives

Hyperbolic Functions and Their Derivatives The hyperbolic sine and hyperbolic cosine functions are defined as

ex − e−x ex + e−x
sinh x = and cosh x =
2 2

Here is the graphs of y) = sinh x and y = cosh x

The other six hyperbolic functions are

sinh x cosh x
tanh x = , coth x =
cosh x sinh x
1 1
sech x = , csch x = .
cosh x sinh x

Here is the graph of y = tanh x.

Exercise. Draw the graphs of the other three hyperbolic function.

Theorem 1.(Derivatives of Hyperbolic Function).

(sinh x)0 = cosh x (cosh x)0 = sinh x

(tanh x)0 = sech 2 x (coth x)0 = −csh 2 x,

(sech x)0 = −sech x tanh x (csch x)0 = −csch x coth x.

Exercise. Prove each of the above formulas.

 Inverse Hyperbolic Functions and Their Derivatives

Theorem 2. ( Inverse Hyperbolic Functions) Let sinh−1 x, cosh−1 x, and tanh−1 x be the inverses of sinh x, cosh x, and
tanh x, respectively. Then
√
sinh−1 x = ln x + x2 + 1 f or all x.

√
cosh1 x = ln x +


x2 − 1 f or all x ≥ 1.
 
tanh−1 x = 1
2 ln 1+x
1−x f or all x.

Proof. We will prove the first and leave the rest for you as an exercise. In particular, make sure to explain the domains.

Let y = sinh−1 x. Then by definition of inverse, we have x = sinh y. We now use the definition of sinh y to express this as
ey − e−y
x=
2
and solve this for y. First multiply by 2ey to get
2xey = e2y − 1
or
e2y − 2xey − 1 = 0
2
. If we let z = ey , then e2y = (ey ) = z 2 and hence the above equation becomes

z 2 − 2xz − 1 = 0.
We now use quadratic formula to get
p
−(−2x) ± (−2x)2 − 4(1)(−1) p
z= = x ± x2 + 1
2(1)
Since z = ey > 0 for all y, we see that p
z = ey = x + x2 + 1.
(Why?) Taking natural of both sides we get
 p 
y = sinh−1 x = ln x + x2 + 1 .
Exercise. Find the domain and the formula the other inverse hyperbolic functions.

Theorem 2. ( The Derivative of Inverse Hyperbolic Functions)

d
sinh−1 x = √ 1


dx x2 +1
f or all x.

d
cosh1 x = √ 1


dx x2 −1
f or all x ≥ 1.

d
tanh−1 x = √ 1


dx 1−x2
f or all x.

Proof. These follow form the formulas from Theorem 1.

Exercise. Find the formula the derivatives of other inverse hyperbolic functions.