Study material of B.Sc.

(Semester - I)
US01CMTH02
(Higher Order Derivatives)
Prepared by Nilesh Y. Patel
Head,Mathematics Department,V.P.and R.P.T.P.Science College

UNIT-1

1. Higher order derivatives of standard functions

1.1. Definition. Let f : (a, b) → R be differentiable on (a, b). If f ′ is differentiable, then
we say that f is twice differentiable and f ′′ is called the second derivative of f . f ′′ is also
2
denoted by ddxf2 . Further, suppose the (n − 1)th derivative f (n−1) of f exists. If f (n−1) is
differentiable on (a, b), then its derivative is called the nth derivative of f and is denoted
n
by f (n) or ddxnf . If the function f is expressed in terms of y = f (x), then the successive
dy d2 y dn y
derivatives of y are denoted by y1 , y2 , . . . , yn or dx , dx2 , . . . , dx n.

The following proposition describes the formulae for nth order derivatives of some
standard functions.
1.2. Proposition. Let a, b, c ∈ R.Then prove the following
(1) For an integer m if y = (ax + b)m , then
yn = m(m − 1) · · · (m − n + 1)an (ax + b)m−n .
(2) If y = (ax + b)m , with m ∈ N, then the above reduces to
m!
yn = an (ax + b)m−n .
(m − n)!
(3) If y = (ax + b)−1 , then
(−1)n n!an
yn = .
(ax + b)n+1
(4) For y = log(ax + b),
(−1)n−1 (n − 1)!an
yn = .
(ax + b)n
(5) For y = amx , yn = mn (log a)n amx .
(6) For y = emx , yn = mn emx .
( )
(7) For y = cos(ax + b), yn = an cos ax + b + nπ .
n
( nπ
2
)
(8) For y = sin(ax + b), yn = a sin ax + b + 2 .
√
(9) For y = eax( cos(bx
) + c), yn = rn eax cos(bx + c + nφ), where r = a2 + b2 ,
φ = tan−1 ab .
√
(10) For y = eax( sin(bx
) + c), yn = rn eax sin(bx + c + nφ), where r = a2 + b2 ,
φ = tan−1 ab .
1
2

Proof. (1) We prove this by induction. Indeed, y1 = ma(ax + b)m−1 .

Assume that yn = m(m − 1)(m − 2) · · · (m − n + 1)an (ax + b)m−n for a fixed n ∈ N. Then
dyn d
yn+1 = = (m(m − 1) · · · (m − n + 1)an (ax + b)m−n )
dx dx
d
= m(m − 1) · · · (m − n + 1)an ((ax + b)m−n )
dx
= m(m − 1) · · · (m − n + 1)a (m − n)(ax + b)(m−n−1) a
n

= m(m − 1) · · · (m − (n + 1) + 1)an+1 (ax + b)(m−(n+1)) .

(2) and (3) are particular cases of (1).
1 (−1)0 (0)! a1 (−1)1−1 (1−1)! a1
(4) y = log(ax + b). So, y1 = ax+b
a = (ax+b)1
= (ax+b)1
.
(−1)n−1 (n−1)! an
Now assume that yn = (ax+b)n
. Then
( )
dyn d (−1)n−1 (n − 1)!an
yn+1 = =
dx dx (ax + b)n
( )
n d 1
= (−1) (n − 1)!a
n−1
dx (ax + b)n
d
= (−1)n−1 (n − 1)!an (ax + b)−n
dx
= (−1) (n − 1)!a (−n)(ax + b)−n−1
n−1 n+1

(−1)n n!an+1
= (−1)n n!an+1 (ax + b)−(n+1) = .
(ax + b)(n+1)
This proves (4).
(5) y = amx . Then y1 = amx (log a)m = m(log a)amx . Similarly we get, y2 = m2 (log a)2 amx ,
y3 = m3 (log a)3 amx . In general,
yn = mn (log a)n amx .
(6) follows by taking a = e in (5).
(7) y = cos(ax + b). Then,
( π)
y1 = −a sin(ax + b) = a cos ax + b +
( 2 )
π 2π
y2 = −a sin(ax + b + ) = a cos ax + b +
2 2
2 2
( )
2π 3π
y3 = −a sin(ax + b +
3 3
) = a cos ax + b +
2 2
... ... ... ... ...
(n − 1)π ( nπ )
yn = −a sin(ax + b +
n n
) = a cos ax + b + .
2 2
(8) can be proved similarly.
(9) y = eax cos(bx + c).
y1 = aeax cos(bx + c) − beax sin(bx + c)
1. Standard functions 3

= eax (a cos(bx + c) − b sin(bx + c)). (1.2.1)

b
√
Let a = r cos
( )φ and b = r sin φ. Then a2 + b2 = r2 and a
= tan φ. That is, r = a2 + b2 ,
φ = tan−1 ab . Now by (1.2.1),
y1 = eax (r cos φ cos(bx + c) − r sin φ sin(bx + c))
= reax cos(bx + c + φ).
Similarly,
y2 = r2 eax cos(bx + c + 2φ).
Continuing in this manner, in general we get,
yn = rn eax cos(bx + c + nφ),
√ ( )
where r = a2 + b2 , φ = tan−1 ab .
(10) can be proved similarly. 

1.3. Example. If y = cos mx − sin mx, then prove that

1
yn = mn (1 − (−1)n sin 2mx) 2 .

Solution. Here y = cos mx − sin mx. Hence

[ ( nπ ) ( nπ )]
yn = mn cos mx + − sin mx +
2 2
[[ ( ) ( )]2 ] 12
nπ nπ
= mn cos mx + − sin mx +
2 2
[ ( nπ ) ( nπ )] 12
= mn 1 − 2 cos mx + sin mx +
2 2
1
= mn [1 − sin(2mx + nπ)] 2
1
= mn [1 − sin 2mx cos nπ − cos 2mx sin nπ] 2
1
= mn [1 − (−1)n sin 2mx] 2 .


1.4. Example. Find yn for y = e2x cos x sin2 2x.

Solution. Here,
y = e2x cos x sin2 2x
( )
2x 1 − cos4x
= e cos x
2
1
= e2x (cos x − cos 4x cos x)
2
1 1
= e2x (cos x − (cos 5x + cos 3x))
2 2
1 2x 1 2x 1
= e cos x − e cos 5x − e2x cos 3x
2 4 4
4

5n/2 2x ( ) 29n/2 2x ( )
⇒ yn = e cos x + n tan−1 12 − e cos 5x + n tan−1 25
2 4
13n/2 ( )
− e2x cos 3x + n tan−1 23 .
4


2. Leibniz’s rule
2.1. Theorem (Leibniz’s rule). State and prove Leibniz’s theorem.
Statement : Let u, v : E → R be sufficiently many times differentiable functions. Then for
any n ∈ N,
(uv)n = un v + nC1 un−1 v1 + nC2 un−2 v2 + · · · + uvn .
Proof. We prove this theorem by mathematical induction on n. Note that the result for
n = 1 is (uv)1 = u1 v + uv1 , which is obviously true. Suppose that the result holds for
n = k. That is,
(uv)k = uk v + kC1 uk−1 v1 + kC2 uk−2 v2 + · · · + uvk .
Differentiating this we get,
(uv)k+1 = uk+1 v + uk v1 + kC1 (uk v1 + uk−1 v2 ) + kC2 (uk−1 v2 + uk−2 v3 )
+ · · · + u1 vk + uvk+1
= uk+1 v + (1 + kC1 )uk v1 + ( kC1 + kC2 )uk−1 v2
+ ( kC2 + kC3 )uk−2 v3 + · · · + uvk+1
= uk+1 v + k+1
C1 uk v1 + k+1
C2 uk−1 v2 + C3 uk−2 v3 + · · · + uvk+1 ,
k+1

as kCr−1 + kCr = k+1

Cr . Thus the result is true for n = k + 1. Hence the result holds for
all n ∈ N. 
Given a product of functions, usually it is a common practice to select the function
as u whose nth derivative is known to us.
2.2. Example. If y = x log(x − 1), then find yn .
Solution. Let u = log(x − 1), then by Leibniz’s Theorem,
(uv)n = un v + nC1 un−1 v1 + nC2 un−2 v2 + · · · + uvn .
Hence,
(−1)n−1 (n − 1)! (−1)n−2 (n − 2)!
yn = x + n 1 + 0 + ···
(x − 1)n (x − 1)n−1
(−1)n−2 (−1)(n − 1)(n − 2)! x (−1)n−2 (n − 2)!(x − 1)
= + n
(x − 1)n (x − 1)n
(−1)n−2 (n − 2)! (−1)n−2 (n − 2)!
= (−x(n − 1) + n(x − 1)) = (x − n).
(x − 1)n (x − 1)n

(1 )
2.3. Example. If x = cos m
log y , then find yn (0).
2. Leibniz’s rule 5

Solution. We can write the given relation as

−1
y = em cos x
. (2.3.1)
Then
( )
m cos−1 x m √
y1 = e −√ ⇒ 1 − x2 y1 = −my. (2.3.2)
1 − x2
Hence,
(1 − x2 )y12 = m2 y 2 .

By differentiating this we get,

2(1 − x2 )y1 y2 − 2xy12 = 2m2 yy1 ⇒ (1 − x2 )y2 − xy1 = m2 y. (2.3.3)

Hence by Leibniz’s Theorem, we get,
yn+2 (1 − x2 ) + n C1 yn+1 (−2x) − 2 n
C 2 yn
− yn+1 x − n C1 yn = m2 yn
⇒(1 − x2 )yn+2 − 2nxyn+1 − n(n − 1)yn
− xyn+1 − nyn − m2 yn = 0
⇒(1 − x2 )yn+2 − (2n + 1)xyn+1 − (n2 + m2 )yn = 0. (2.3.4)
From (2.3.1),(2.3.2), (2.3.3) and (2.3.4) we have,
y(0) = emπ/2 ; y1 (0) = −memπ/2 ; y2 (0) = m2 emπ/2 ;
yn+2 (0) = (n2 + m2 )yn (0). (2.3.5)
Putting n = 1, 2, 3, . . . in (2.3.5) we get,
y3 (0) = (12 + m2 )y1 (0) = −emπ/2 m(m2 + 1);
y4 (0) = (22 + m2 )y2 (0) = emπ/2 m2 (m2 + 22 );
y5 (0) = (32 + m2 )y3 (0) = −emπ/2 m(m2 + 1)(m2 + 32 ).
In general,
{
emπ/2 m2 (m2 + 22 ) · · · (m2 + (n − 2)2 ) if n is even;
yn (0) =
−emπ/2 m(m2 + 12 ) · · · (m2 + (n − 2)2 ) n odd; n ̸= 1.

2.4. Example. If √
y = (x − 4 + x2 )m , (2.4.1)
then find yn (0).
Solution. By differentiating with respect to x, we get,
( )
√ 2x
y1 = m(x − 4 + x2 ) m−1
1− √
2 4 + x2
(√ )
√ 4 + x2−x
= m(x − 4 + x2 )m−1 √
4 + x2
6

√
m(x − 4 + x2 )m
=− √
4 + x2
my
= −√
4 + x2
√
⇒ 4 + x2 y1 = −my. (2.4.2)
Squaring both the sides,
(4 + x2 )y12 = m2 y 2 ,
which, on differentiation gives,
(4 + x2 )2y1 y2 + 2xy12 = m2 2yy1 ⇒ (4 + x2 )y2 + xy1 = m2 y. (2.4.3)
By Leibniz’s Theorem, we get,
yn+2 (4 + x2 ) + n C1 yn+1 (2x) + 2 n
C2 yn
+ yn+1 x + n C1 yn = m2 yn
⇒(4 + x2 )yn+2 + 2nxyn+1 + n(n − 1)yn
+ xyn+1 + nyn − m2 yn = 0
⇒(4 + x2 )yn+2 + (2n + 1)xyn+1 + (n2 − m2 )yn = 0. (2.4.4)
From (2.4.1),(2.4.2), (2.4.3) and (2.4.4) we have,
y(0) = (−2)m ;
y(0) m
y1 (0) = −m = − (−2)m ;
2 2
m2 y(0) m2
y2 (0) = = (−2)m ;
4 4
(m2 − n2 )yn (0)
yn+2 (0) = . (2.4.5)
4
Putting n = 1, 2, 3, . . . in (2.4.5) to get,
(m2 − 12 )y1 (0) m (m2 − 12 )
y3 (0) = = − (−2)m ;
4 2 4
(m2 − 22 )y2 (0) m2 (m2 − 22 )
y4 (0) = = (−2)m ;
4 4 4
(m2 − 32 )y3 (0) m (m2 − 12 ) (m2 − 32 )
y5 (0) = = − (−2)m .
4 2 4 4
In general,
{
m2 m2 −22 m2 −(n−2)2
(−2)m 4 4
· · · 4
if n is even;
yn (0) =
−(−2)m m2 (m 4−1 ) · · · (m −(n−2)
2 2 2 2)

4
n odd; n ̸= 1.


2.5. Example. Let y = (x2 − 2)m . Find the value of m such that
(x2 − 2)yn+2 + 2xyn+1 − n(n + 1)yn = 0.
2. Leibniz’s rule 7

Solution. Differentiating y with respect to x, we have,

y1 = m(x2 − 2)m−1 2x
⇒(x2 − 2)y1 = 2mxy
⇒(x2 − 2)y2 + 2xy1 = 2m(xy1 + y)
⇒(x2 − 2)y2 + 2(1 − m)xy1 − 2my = 0.

Hence, by applying Leibniz’s Theorem, we get,

(x2 − 2)yn+2 + n C1 2xyn+1 + n C2 2yn

+ 2(1 − m)(xyn+1 + nyn ) − 2myn = 0
⇒(x2 − 2)yn+2 + 2nxyn+1 + n(n − 1)yn
+ 2(1 − m)(xyn+1 + nyn ) − 2myn = 0
⇒(x − 2)yn+2 + (n − m + 1)2xyn+1 + (n2 − 2mn + n − 2m)yn = 0
2

⇒(x2 − 2)yn+2 + (n − m + 1)2xyn+1 + (n − 2m)(n + 1)yn = 0.

Comparing the coefficients of the last equation with the given equation, we find that
m = n. 

♣♣♣♣♣♣♣♣