CHAPTER 40

Integration of Rational
Functions; The Method
of Partial Fractions

This chapter will give a general method for evaluating indefinite integrals of the type

N(x)
dx
D(x)

where N(x) and D(x) are polynomials. That is to say, we shall show how to find the antiderivative of any rational
function f (x) = N(x)/D(x) (see Section 9.3). Two assumptions will be made, neither of which is really restrictive:
(i) the leading coefficient (the coefficient of the highest power of x) in D(x) is +1; (ii) N(x) is of lower degree than
D(x) [that is, f (x) is a proper rational function].

EXAMPLES
8x 4 −7 8x 4 −56x 4
(a) = = 10
− 17 x 10 + 3x − 11 −7 − x + 3x − 11
1 10 x − 21x + 77
7

x 4 + 7x
(b) Consider the improper rational function f (x) = . Long division (see Fig. 40-1) yields
x2 − 1

7x + 1
f (x) = x 2 + 1 + 2
x −1

Consequently,
   
7x + 1 x3 7x + 1
f (x)dx = (x 2 + 1) dx + dx = +x+ dx
x2 − 1 3 x2 − 1
and the problem reduces to finding the antiderivative of a proper rational function.

The theorems that follow hold for polynomials with arbitrary real coefficients. However, for simplicity we shall
illustrate them only with polynomials whose coefficients are integers.

339
Copyright © 2008, 1997, 1985 by The McGraw-Hill Companies, Inc. Click here for terms of use.
340 THE METHOD OF PARTIAL FRACTIONS [CHAP. 40

x2 + 1
x 2 − 1 x 4 + 7x
x4 − x2
x 2 + 7x
x2 − 1
7x + 1
Fig. 40-1

Theorem 40.1: Any polynomial D(x) with leading coefficient 1 can be expressed as the product of linear factors, of
the form x − a, and irreducible quadratic factors (that cannot be factored further), of the form x 2 +
bx + c, repetition of factors being allowed.
As explained in Section 7.4, the real roots of D(x) determine its linear factors.

EXAMPLES
(a) x 2 − 1 = (x − 1)(x + 1)
Here, the polynomial has two real roots (±1) and, therefore, is a product of two linear factors.
(b) x 3 + 2x 2 − 8x − 21 = (x − 3)(x 2 + 5x + 7)
The root x = 3, which generates the linear factor x − 3, was found by testing the divisors of 21. Division of D(x) by x − 3
yielded the polynomial x 2 + 5x + 7. This polynomial is irreducible, since, by the quadratic formula, its roots are
 √
−b ± b2 − 4c −5 ± −3
x= =
2 2
which are not real numbers.

Theorem 40.2 (Partial Fractions Representation): Any (proper) rational function f (x) = N(x)/D(x) may be written
as a sum of simpler, proper rational functions. Each summand has as denominator one of the linear or
quadratic factors of D(x), raised to some power.

By Theorem 40.2, f (x)dx is given as a sum of simpler antiderivatives—antiderivatives which, in fact, can be
found by the techniques already known to us.
It will now be shown how to construct the partial fractions representation and to integrate it term by term.
Case 1: D(x) is a product of nonrepeated linear factors.
The partial fractions representation of f (x) is
N(x) A1 A2 An
= + + ··· +
(x − a1 )(x − a2 ) · · · (x − an ) x − a1 x − a2 x − an
The constant numerators A1 , . . . , An are evaluated as in the following example.

2x + 1 A1 A2
EXAMPLE = +
(x + 1)(x − 1) x+1 x−1
Clear the denominators by multiplying both sides by (x + 1)(x − 1),

2x + 1 A A
(x + 1)(x − 1) = (x + 1)(x − 1) 1 + (x + 1)(x − 1) 2
(x + 1)(x − 1) x+1 x−1
2x + 1 = A1 (x − 1) + A2 (x + 1) (1)

In (1), substitute individually the roots of D(x). With x = −1,

1
−1 = A1 (−2) + 0 or A1 =
2
CHAP. 40] THE METHOD OF PARTIAL FRACTIONS 341

and with x = 1,
3
3 = 0 + A2 (2) or A2 =
2
With all constants known, the antiderivative of f (x) will be the sum of terms of the form

A
dx = A ln |x − a|
x−a

Case 2: D(x) is a product of linear factors, at least one of which is repeated.

This is treated in the same manner as in Case 1, except that a repeated factor (x − a)k gives rise to a sum of the
form
A1 A2 Ak
+ + ··· +
x − a (x − a) 2 (x − a)k

3x + 1 A1 A2 A3
EXAMPLE = + +
(x − 1)2 (x − 2) x − 1 (x − 1)2 x−2
Multiply by (x − l)2 (x − 2),
3x + 1 = A1 (x − 1)(x − 2) + A2 (x − 2) + A3 (x − 1)2 (2)
Letting x = 1,
4 = 0 + A2 (−1) + 0 or A2 = −4
Letting x = 2,
7 = 0 + 0 + A3 (1) or A3 = 7
The remaining numerator, A1 , is determined by the condition that the coefficient of x 2 on the right side of (2) be zero (since it is
zero on the left side). Thus,
A 1 + A3 = 0 or A1 = −A3 = −7
[More generally, we use all the roots of D(x) to determine some of the A’s, and then compare coefficients—of as many powers of
x as necessary—to find the remaining A’s.]

Now the antiderivatives of f (x) will consist of terms of the form ln |x − a| plus at least one term of the form

A B
dx = [ j ≥ 2]
(x − a) j (x − a) j−1

Case 3: D(x) has irreducible quadratic factors, but none is repeated.

In this case, each quadratic factor x 2 + bx + c contributes a term

Ax + B
x2 + bx + c

to the partial fractions representation.

x2 − 1 A1 A x + A3
EXAMPLE = + 22
(x 2 + 1)(x + 2) x+2 x +1
Multiply by (x 2 + l)(x + 2),

x 2 − 1 = A1 (x 2 + 1) + (A2 x + A3 )(x + 2) (3)

Let x = −2,
3
3 = A1 (5) + 0 or A1 =
5
342 THE METHOD OF PARTIAL FRACTIONS [CHAP. 40

Comparing coefficients of x 0 (the constant terms),

 
1 1 8 4
−1 = A1 + 2A3 or A3 = − (1 + A1 ) = − =−
2 2 5 5

Comparing coefficients of x 2 ,
2
1 = A1 + A2 or A2 = 1 − A1 =
5

The sum for f (x) dx will now include, besides terms arising from any linear factors, at least one term of the form
   
b AB
  A x+ +C c≡B−
Ax + B 2  2 
dx =  2 dx  
x + bx + c
2  b 2 
b 2
x+ +δ 2 δ ≡c− >0
2 4
 
Au + C b
= du let u ≡ x +
u2 + δ 2 2
 
u du du
=A +C
u2 + δ 2 u2 + δ 2
A C u
= ln (u2 + δ 2 ) + tan−1
2 δ δ
(For a guarantee that δ is a real number, see Problem 40.7.)

Case 4: D(x) has at least one repeated irreducible quadratic factor.

A repeated quadratic factor (x 2 + bx + c)k contributes to the partial fractions representation the expression
A1 x + A2 A3 x + A 4 A2k−1 x + A2k
+ + ··· +
ax + bx + c (ax + bx + c)
2 2 2 (ax 2 + bx + c)k
The computations in this case may be long and tedious.

x3 + 1 A x + A2 A x + A4
EXAMPLE = 12 + 32
(x 2 + 1)2 x +1 (x + 1)2
Multiply by (x 2 + 1)2 ,
x 3 + 1 = (A1 x + A2 )(x 2 + 1) + A3 x + A4 (4)
Compare coefficients of x 3 ,
1 = A1
Compare coefficients of x 2 ,
0 = A2
Compare coefficients of x,
0 = A1 + A3 or A3 = −A1 = −1
Compare coefficients of x 0 ,
1 = A2 + A4 or A4 = 1 − A 2 = 1

The new contribution to f (x) dx will consist of one or more terms of the form
  
Ax + B u du du
dx = A + C [as in Case 3]
(x 2 + bx + c) j (u2 + δ 2 ) j (u2 + δ 2 ) j

E
= 2 + F cos2( j−1) θ dθ [let u = δ tan θ]
(u + δ 2 ) j−1
and we know how to evaluate the trigonometric integral [see Problem 38.12(a) or example (b) of Section 39.1].
CHAP. 40] THE METHOD OF PARTIAL FRACTIONS 343

Solved Problems

2x 3 + x 2 − 6x + 7
40.1 Evaluate dx.
x2 + x − 6
The numerator has greater degree than the denominator. Therefore, divide the numerator by the denominator,

2x – 1
x 2 + x – 6 2x3 + x2 – 6x + 7
2x3 + 2x2 – 12x
–x2 + 6x + 7
–x – x + 6
2

7x +1
2x 3 + x 2 − 6x + 7 7x + 1
Thus, = 2x − 1 + 2
x2 + x − 6 x +x−6
Next, factor the denominator, x 2 + x − 6 = (x + 3)(x − 2). The partial fractions decomposition has the form (Case 1)

7x + 1 A1 A2
= +
(x + 3)(x − 2) x+3 x−2

Multiply by the denominator (x + 3)(x − 2),

7x + 1 = A1 (x − 2) + A2 (x + 3)
Let x = 2, 15 = 0 + 5A2 or A2 = 3
Let x = −3, − 20 = −5A1 + 0 or A1 = 4
7x + 1 4 3
Thus, = +
(x + 3)(x − 2) x+3 x−2
   
2x 3 + x 2 − 6x + 7 4 3
and dx = (2x − 1)dx + dx + dx
x2 + x − 6 x+3 x−2
= x 2 − x + 4 ln |x + 3| + 3 ln |x − 2| + C

x 2 dx
40.2 Find .
x 3 − 3x 2 − 9x + 27
Testing the factors of 27, we find that 3 is a root of D(x). Dividing D(x) by x − 3 yields

x 3 − 3x 2 − 9x + 27 = (x − 3)(x 2 − 9) = (x − 3)(x − 3)(x + 3) = (x − 3)2 (x + 3)

and so the partial fractions representation is (Case 2)

x2 A1 A2 A3
= + +
(x − 3) (x + 3)
2 x − 3 (x − 3)2 x+3

Multiply by (x − 3)2 (x + 3),

x 2 = A1 (x − 3)(x + 3) + A2 (x + 3) + A3 (x − 3)2
3
Let x = 3, 9 = 0 + 6A2 + 0 or A2 =
2
1
Let x = −3, 9 = 0 + 0 + A3 (−6)2 or A3 =
4

Compare coefficients of x 2 ,
3
1 = A1 + A3 or A1 = 1 − A3 =
4
344 THE METHOD OF PARTIAL FRACTIONS [CHAP. 40

x2 3 1 3 1 1 1
Thus, = + +
x − 3x − 9x + 27
3 2 4 x − 3 2 (x − 3)2 4 x+3

x 2 dx 3 3 1 1
and = ln |x − 3| − + ln |x + 3| + C
x 3 − 3x 2 − 9x + 27 4 2x−3 4

x+1
40.3 Find dx.
x(x 2 + 2)
This is Case 3,
x+1 A A x + A3
= 1 + 22
x(x 2 + 2) x x +2
Multiply by x(x 2 + 2),
x + 1 = A1 (x 2 + 2) + x(A2 x + A3 )

1
Let x = 0, 1 = 2A1 + 0 or A1 =
2
Compare coefficients of x 2 ,
1
0 = A1 + A2 or A2 = −A1 = −
2
Compare coefficients of x,
1 = A3
 
x+1 1 1 (− 12 )x + 1
Thus, = +
x(x 2 + 2) 2 x x2 + 2
   
x+1 1 1 1 x dx dx
and dx = dx − +
x(x + 2)
2 2 x 2 x +2
2 x +2
2

Because the quadratic factor x 2 + 2 is a complete square, we can perform the integrations on the right without a change of
variable,

x+1 1 1 1 x
dx = ln |x| − ln (x 2 + 2) + √ tan−1 √ + C
x(x 2 + 2) 2 4 2 2

1
40.4 Evaluate dx.
1 − sin x + cos x
Observe that the integrand is a rational function of sin x and cos x. Any rational function of the six trigonometric
functions reduces to a function of this type, and the method we shall use to solve this particular problem will work for any
such function.
Make the change of variable z = tan (x/2); that is, x = 2 tan−1 z. Then,
2
dx = dz
1 + z2
and, by Theorem 26.8,
x x tan (x/2)
sin x = 2 sin cos = 2 2
2 2 sec (x/2)
tan (x/2) 2z
=2 =
1 + tan2 (x/2) 1 + z2
x tan2 (x/2)
cos x = 1 − 2 sin2 =1−2 2
2 sec (x/2)
tan2 (x/2) 2z2 1 − z2
=1−2 =1− =
1 + tan (x/2)
2 1+z 2 1 + z2
CHAP. 40] THE METHOD OF PARTIAL FRACTIONS 345

When these substitutions are made, the resulting integrand will be a rational function of z (because compositions and
products of rational functions are rational functions). The method of partial fractions can then be applied,
   
2z 1 − z2 −1 2
(1 − sin x + cos x)−1 dx = 1− + dz
1 + z2 1 + z2 1 + z2
  −1
(1 + z2 ) − 2z + (1 − z2 ) 2
= dz
1 + z2 1 + z2
   
2 − 2z −1 2 1 + z2 2
= dz = dz
1+z 2 1+z 2 2 − 2z 1 + z2

1
= dz = − ln |1 − z| + C
1−z
 x 

= − ln 1 − tan  + C
2

x dx
40.5 Find .
(x + 1)(x 2 + 2x + 2)2
This is Case 4 for D(x), and so
x A1 A x + A3 A x + A5
= + 22 + 2 4
(x + 1)(x 2 + 2x + 2)2 x + 1 x + 2x + 2 (x + 2x + 2)2

Multiply by (x + 1)(x 2 + 2x + 2)2 ,

x = A1 (x 2 + 2x + 2)2 + (A2 x + A3 )(x + 1)(x 2 + 2x + 2) + (A4 x + A5 )(x + 1)

or, partially expanding the right-hand side,

x = A1 (x 4 + 4x 3 + 8x 2 + 8x + 4) + (A2 x + A3 )(x 3 + 3x 2 + 4x + 2) + (A4 x + A5 )(x + 1) (1)

In (1), let x = −1, −1 = A1 (1) = A1

Compare coefficients of x 4 ,
0 = A1 + A2 or A2 = −A1 = 1
Compare coefficients of x 3 ,
0 = 4A1 + 3A2 + A3 or A3 = 4A1 − 3A2 = 1
Compare coefficients of x 2 ,

0 = 8A1 + 4A2 + 3A3 + A4 or A4 = −8A1 − 4A2 − 3A3 = 1

Compare coefficients of x 0 ,

0 = 4A1 + 2A3 + A5 or A5 = −4A1 − 2A3 = 2

Therefore,
   
x dx dx (x + 1) dx (x + 2) dx
= −1 + +
(x + 1)(x 2 + 2x + 2)2 x+1 x + 2x + 2
2 (x + 2x + 2)2
2
  
u du u du du
= − ln |x + 1| + + +
u2 + 1 (u2 + 1)2 (u2 + 1)2
[by Quick Formulas II and I]
  
1 1 1
= − ln |x + 1| + ln (u2 + 1) − + cos2 θ dθ [Case 4: let u = tan θ]
2 2 u2 + 1
 
1 1 1 θ sin 2θ
= −ln |x + 1| + ln (x 2 + 2x + 2) − + + +C
2 2 x 2 + 2x + 2 2 4
Now θ = tan−1 u = tan−1 (x + 1)
346 THE METHOD OF PARTIAL FRACTIONS [CHAP. 40

and (see Problem 40.4)

2 tan θ 2(x + 1)
sin 2θ = = 2
1 + tan2 θ x + 2x + 2
so that we have, finally,
  
x dx 1 2 + 2x + 2) − 1 1
= −ln |x + 1| + In (x
(x + 1)(x 2 + 2x + 2)2 2 2 x 2 + 2x + 2
 
1 1 x+1
+ tan−1 (x + 1) + +C
2 2 x 2 + 2x + 2
 
1 x
= ln (x 2 + 2x + 2) + 2 + tan−1 (x + 1) − ln |x + 1| + C
2 x + 2x + 2

Supplementary Problems
40.6 Find the following antiderivatives:
  
dx x dx x 4 − 4x 2 + x + 1
(a) (b) (c) dx
x −9
2 (x + 2)(x + 3) x2 − 4
  
2x 2 + 1 x2 − 4 x3 + 1
(d) dx (e) dx (f ) dx
(x − 1)(x − 2)(x − 3) x 3 − 3x 2 − x + 3 x(x + 3)(x + 2)(x − 1)
  
x dx x−5 2x dx
(g) (h) dx (i)
x 4 − 13x 2 + 36 x 2 (x + 1) (x − 2)2 (x + 2)
  
x+4 x 4 dx dx
( j) dx (k) (l)
x 3 + 6x 2 + 9x x − 2x 2 − 7x − 4
3 x(x 2 + 5)
   4
x 2 dx dx x +1
(m) (n) (o) dx
(x − 1)(x 2 + 4x + 5) (x 2 + 1)(x 2 + 4) x 3 + 9x
  
dx x 2 dx x3 + 1
(p) (p) (r) dx
x(x + 1)2
2 (x − 1)(x 2 + 4)2 x(x + x + 1)2
2
  
x−1 x2 + 2 dx
(s) dx (t) dx (u)
x + 2x 2 − x − 2
3 x(x + 5x + 6)
2 1 + ex

40.7 Show that p(x) = x 2 + bx + c is irreducible if and only if c − (b2 /4) > 0. [Hint: A quadratic polynomial is irreducible if
and only if it has no linear factor; that is (by Theorem 7.2), if and only if it has no real root.]

40.8 (a) Find the area of the region in the first quadrant under the curve y = l/(x 3 + 27) and to the left of the line x = 3.
(b) Find the volume of the solid generated by revolving the region of part (a) around the y-axis.

dx
40.9 Find . [Hint: See Problem 40.4.]
1 − sin x

cos x dx
40.10 Find . (a) Use the method of Problem 40.4. (b) Use the Quick Formula II. (c) Verify that your answers are
sin x − 1
equivalent.

40.11 Evaluate the following integrals involving fractional powers:

   
dx dx dx dx
(a) √ (b) √ (c) √ (d) √ √
3
x−x 1+ 4 x−1 x 1 + 3x 3
x+ x
   2/3
dx √ x dx
(e) √ (f ) 1 + ex dx (g)
x+1 x+1
[Hints: Part (a) let x = z3 ; part (b) let x − 1 = z4 ; part (c) let 1 + 3x = z2 ; part (d) let x = z6 .]
ANS.] ANSWERS TO SUPPLEMENTARY PROBLEMS 391


√ 1 √  √ e2 + 1 − 1
39.16 17 + ln ( 17 + 4). 39.17 e2 + 1 − (1 + 2) + ln √ .
4 2−1

39.18 Same answer as to Problem 39.17 (because the two arcs are mirror images in the line y = x).
√ √ √
39.19 ln (2 + 3). 39.20 π 9 4 = 6π.

CHAPTER 40
 
1  x − 3  x3 1 3
40.6 (a) ln   + C; (b) 3 ln |x + 3| − 2 ln |x + 2| C; (c) + ln |x + 2| + ln |x − 2| + C;
6 x+3 3 4 4
3 19 3 3 5
(d) ln |x − 1| − 9 ln |x − 2| + ln |x − 3| + C; (e) ln |x − 1| − ln |x + 1| + ln |x − 3| + C;
2 2 4 8   8  
 x2 − 9   x  5
1 13 7 1 1     + + C;
( f ) − ln |x| + ln |x + 3| − ln |x + 2| + ln |x − 1| + C; (g) ln   + C; (h) 6 ln 
6 6 6 6 10  x 2 − 4  x + 1 x
   
1  x − 2  1 4  x  1 1
(i) ln  − + C; ( j) ln + + C;
4 x + 2 x − 2 9 x + 3 3 x + 3  
 
x2 1 5 1 x2
(k) + 2x + 256 ln |x − 4| + 19 ln |x + 1| + + C; (l) ln 2 + C;
2 25 x+1 10 x +5
1 9 13 1 1 x
(m) ln |x − 1| + ln (x 2 + 4x + 5) − tan−1 (x + 2) + C; (n) tan−1 x − tan−1 + C;
10 20 10 3 6 2
x2 1 1 1 1
(o) + [ln |x| − 41 ln (x 2 + 9)] + C; (p) ln |x| − ln (x 2 + 1) + + C;
2 9 2 2 x2 + 1
1 1 3 x 1 x−4
(q) ln |x − 1| − ln (x 2 + 4) + tan−1 + + C;
 x +4
25 50 100 2 10 2
√   
1 3 2x + 1 2 x−1 x + 1
(r) ln |x| − ln (x 2 + x + 1) − tan−1 − +  
2 9
√
3 3 x2 + x + 1
C; (s) ln  x + 2  + C;
1 11
(t) ln |x| + ln |x + 3| − 3 ln |x + 2| + C; (u) x − ln (1 + ex ) + C.
3 3

1 √ 2π √ 2
40.8 (a) (3 ln 2 + π 3); (b) (π 3 − 3 ln 2). 40.9 x + C.
81 27 1 − tan
2

40.10 (a) ln (1 − sin x) + C; (b) ln |sin x − 1| + C; (c) sin x ≤ 1.

) √ √  √ 
3  2/3

 ( 4 x − 1)3 x−1 √ √  1 + 3x − 1 
 
40.11 (a) − ln 1 − x  + C; (b) 4 − + x − 1 − ln (1 + x − 1) + C; (c) ln  √
4 4
 + C;
2 3 2  1 + 3x + 1 
√ √ √ √ 4 √ √
(d) 2 x − 3 3 x + 6 6 x − ln (1 + 6 x) + C; (e) ( x − 2) x + 1 + C;
√  3
√ 1 + ex − 1 √ √
( f ) 2 1 + ex + ln √ + C = 2 1 + ex + 2 ln ( 1 + ex − 1) − x + C;
1+e +1
x
 
  1  x + 1  √ 2x 1/3 − 1
3 2/3  1/3    −1
(g) x + ln x + 1 − ln  1/3 − 3 tan √ + C.
2 2 x + 1 3