SECTION 7.

8 Improper Integrals 567

43. The intensity of light with wavelength ! traveling through which the Trapezoidal Rule with n − 2 is more accurate than
a diffraction grating with N slits at an angle " is given by the Midpoint Rule.
Is"d − N 2 sin 2kyk 2, where k − s#Nd sin "dy! and d is 46. Sketch the graph of a continuous function on f0, 2g for which
the distance between adjacent slits. A helium-neon laser the right endpoint approximation with n − 2 is more accurate
with wavelength ! − 632.8 3 1029 m is emitting a nar- than Simpson’s Rule.
row band of light, given by 21026 , " , 1026, through
a grating with 10,000 slits spaced 1024 m apart. Use the 47. If f is a positive function and f 0sxd , 0 for a < x < b, show that
Midpoint Rule with n − 10 to estimate the total light Tn , y f sxd dx , Mn
b

intensity y210
10 26
Is"d d" emerging from the grating.
26
a

48. Show that if f is a polynomial of degree 3 or lower, then

44. Use the Trapezoidal Rule with n − 10 to approximate
Simpson’s Rule gives the exact value of ya f sxd dx.
b
y 020 coss# xd dx. Compare your result to the actual value.
Can you explain the discrepancy? 49. Show that 12 sTn 1 Mn d − T2n.
45. Sketch the graph of a continuous function on f0, 2g for 50. Show that 13 Tn 1 23 Mn − S2n.

In defining a definite integral yab f sxd dx we dealt with a function f defined on a finite
interval fa, bg and we assumed that f does not have an infinite discontinuity (see Sec-
tion 4.2). In this section we extend the concept of a definite integral to the case where the
interval is infinite and also to the case where f has an infinite discontinuity in fa, bg. In
either case the integral is called an improper integral. One of the most important applica-
tions of this idea, probability distributions, will be studied in Section 8.5.

Type 1: Infinite Intervals

y Consider the infinite region S that lies under the curve y − 1yx 2, above the x-axis, and
1
y= to the right of the line x − 1. You might think that, since S is infinite in extent, its area
≈
must be infinite, but let’s take a closer look. The area of the part of S that lies to the left
1 of the line x − t (shaded in Figure 1) is
=1
area=1-
t

G
x=1 t
1 1 1
Astd − y
t
−12
0 1 t x 2 dx − 2 t
1 x x 1

FIGURE 1 Notice that Astd , 1 no matter how large t is chosen.

We also observe that

lim Astd − lim 1 2

tl` tl`
1
t
−1 S D
The area of the shaded region approaches 1 as t l ` (see Figure 2), so we say that the
area of the infinite region S is equal to 1 and we write
1 t 1
y y
`
2 dx − tlim dx − 1
1 x l` 1 x2

y y y y

1 2 4
area= 2 area= 3 area= 5 area=1

0 1 2 x 0 1 3 x 0 1 5 x 0 1 x

FIGURE 2

Copyright 2016 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
568 CHAPTER 7 Techniques of Integration

Using this example as a guide, we define the integral of f (not necessarily a positive
function) over an infinite interval as the limit of integrals over finite intervals.

1 Definition of an Improper Integral of Type 1

(a) If yat f sxd dx exists for every number t > a, then

y y
` t
f sxd dx − lim f sxd dx
a tl` a

provided this limit exists (as a finite number).

(b) If ytb f sxd dx exists for every number t < b, then

y y
b b
f sxd dx − lim f sxd dx
2` t l2` t

provided this limit exists (as a finite number).

The improper integrals ya` f sxd dx and y2`
b
f sxd dx are called convergent if the
corresponding limit exists and divergent if the limit does not exist.
(c) If both ya f sxd dx and y2` f sxd dx are convergent, then we define
` a

y f sxd dx − y f sxd dx 1 y f sxd dx

` a `

2` 2` a

In part (c) any real number a can be used (see Exercise 76).

Any of the improper integrals in Definition 1 can be interpreted as an area provided that
f is a positive function. For instance, in case (a) if f sxd > 0 and the integral ya` f sxd dx
is convergent, then we define the area of the region S − hsx, yd x > a, 0 < y < f sxdj |
in Figure 3 to be
AsSd − y f sxd dx
`

This is appropriate because ya f sxd dx is the limit as t l ` of the area under the graph
`

of f from a to t.

y=ƒ

FIGURE 3 0 a x

EXAMPLE 1 Determine whether the integral y1` s1yxd dx is convergent or divergent.

SOLUTION According to part (a) of Definition 1, we have

1 t 1
y dx − lim y | |g
` t
dx − lim ln x 1
1 x tl` 1 x tl`

− lim sln t 2 ln 1d − lim ln t − `

tl` tl`

Copyright 2016 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 SECTION 7.8 Improper Integrals 569

The limit does not exist as a finite number and so the improper integral y1 s1yxd dx is
`

divergent. n

Let’s compare the result of Example 1 with the example given at the beginning of this
section:
` 1 ` 1
y1 x 2 dx converges y1 x dx diverges
Geometrically, this says that although the curves y − 1yx 2 and y − 1yx look very similar
for x . 0, the region under y − 1yx 2 to the right of x − 1 (the shaded region in Figure 4)
has finite area whereas the corresponding region under y − 1yx (in Figure 5) has infinite
area. Note that both 1yx 2 and 1yx approach 0 as x l ` but 1yx 2 approaches 0 faster than
1yx. The values of 1yx don’t decrease fast enough for its integral to have a finite value.

y y

1 1
y= y=
≈ x

infinite area
finite area

0 1 x 0 1 x

FIGURE 4 FIGURE 5
y1 s1yx 2 d dx converges
`
y1 s1yxd dx diverges
`

EXAMPLE 2 Evaluate y xe x dx.

0

2`

SOLUTION Using part (b) of Definition 1, we have

y y
0 0
xe x dx − lim xe x dx
2` t l2` t

We integrate by parts with u − x, dv − e x dx so that du − dx, v − e x:

y g 2 y e x dx
0 0 0
xe x dx − xe x t
t t

− 2te t 2 1 1 e t

TEC In Module 7.8 you can investi- We know that e t l 0 as t l 2`, and by l’Hospital’s Rule we have
gate visually and numerically whether
several improper integrals are conver- t 1
lim te t − lim − lim
gent or divergent. t l2` t l2` e2t t l2` 2e2t

− lim s2e t d − 0
t l2`

Therefore

y
0
xe x dx − lim s2te t 2 1 1 e t d
2` t l2`

− 20 2 1 1 0 − 21 n

Copyright 2016 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
570 CHAPTER 7 Techniques of Integration

1
EXAMPLE 3 Evaluate y
`
dx.
2` 1 1 x2
SOLUTION It’s convenient to choose a − 0 in Definition 1(c):

1 1 1
y 2 dx − y 2 dx 1 y
` 0 `
dx
2` 11x 2` 11x 0 1 1 x2

We must now evaluate the integrals on the right side separately:

1 dx
y dx − lim y g
` t t
− lim tan21x
0 1 1 x2 tl` 0 1 1 x2 tl` 0

#
− lim stan 21 t 2 tan21 0d − lim tan21 t −
tl` tl` 2
1 dx
y lim y g
0 0 0
2 dx − t l − lim tan21x
2` 11x 2` t 1 1 x 2 t l 2` t

− lim stan 21 0 2 tan 21 td − 0 2 2

t l 2`
S D #
2
−
#
2

Since both of these integrals are convergent, the given integral is convergent and
y
1 # #
1
y
`
y= area=π 2 dx − 1 −#
1+≈ 2` 11x 2 2

0 x Since 1ys1 1 x 2 d . 0, the given improper integral can be interpreted as the area of
the infinite region that lies under the curve y − 1ys1 1 x 2 d and above the x-axis (see
FIGURE 6 Figure 6). n

EXAMPLE 4 For what values of p is the integral

1
y
`
dx
1 xp
convergent?
SOLUTION We know from Example 1 that if p − 1, then the integral is divergent, so
let’s assume that p ± 1. Then

G
x−t
1 x2p11
y y
` t 2p
p dx − tlim x dx − lim
1 x l` 1 t l ` 2p 1 1
x−1

− lim
tl`
1
12p t
1
p21 2 1F G
If p . 1, then p 2 1 . 0, so as t l `, t p21 l ` and 1yt p21 l 0. Therefore

1 1
y
`
p dx − if p . 1
1 x p21

and so the integral converges. But if p , 1, then p 2 1 , 0 and so

1
− t 12p l ` as t l `
t p21

and the integral diverges. n

Copyright 2016 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 SECTION 7.8 Improper Integrals 571

We summarize the result of Example 4 for future reference:

1
y
`
2 dx is convergent if p . 1 and divergent if p < 1.
1 xp

Type 2: Discontinuous Integrands

y
Suppose that f is a positive continuous function defined on a finite interval fa, bd but
has a vertical asymptote at b. Let S be the unbounded region under the graph of f and
y=ƒ x=b above the x-axis between a and b. (For Type 1 integrals, the regions extended indefinitely
in a horizontal direction. Here the region is infinite in a vertical direction.) The area of the
part of S between a and t (the shaded region in Figure 7) is
0 a t b x
Astd − y f sxd dx
t

a
FIGURE 7
If it happens that Astd approaches a definite number A as t l b2, then we say that
the area of the region S is A and we write

y y
b t
f sxd dx − lim2 f sxd dx
a tlb a

We use this equation to define an improper integral of Type 2 even when f is not a posi-
tive function, no matter what type of discontinuity f has at b.

Parts (b) and (c) of Definition 3 are

illustrated in Figures 8 and 9 for the 3 Definition of an Improper Integral of Type 2
case where f sxd > 0 and f has vertical (a) If f is continuous on fa, bd and is discontinuous at b, then
asymptotes at a and c, respectively.
y y
b t
y f sxd dx − lim2 f sxd dx
a tlb a

if this limit exists (as a finite number).

(b) If f is continuous on sa, bg and is discontinuous at a, then

y y
b b
f sxd dx − lim1 f sxd dx
a tla t

0 a t b x if this limit exists (as a finite number).

The improper integral yab f sxd dx is called convergent if the corresponding limit
FIGURE 8
exists and divergent if the limit does not exist.
y c
(c) If f has a discontinuity at c, where a , c , b, and both ya f sxd dx and
ycb f sxd dx are convergent, then we define

y f sxd dx − y f sxd dx 1 y f sxd dx

b c b

a a c

0 a c b x 1
EXAMPLE 5 Find y
5
dx.
2 sx 2 2
FIGURE 9
SOLUTION We note first that the given integral is improper because f sxd − 1ysx 2 2
has the vertical asymptote x − 2. Since the infinite discontinuity occurs at the left

Copyright 2016 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
572 CHAPTER 7 Techniques of Integration

y endpoint of f2, 5g, we use part (b) of Definition 3:

1
y=
œ„„„„
x-2 dx dx
y − lim1 y g
5 5 5
− lim1 2sx 2 2
2 sx 2 2 t l 2 t sx 2 2 t l2 t

3
area=2œ„ − lim1 2(s3 2 st 2 2 ) − 2s3
t l2
0 1 2 3 4 5 x
Thus the given improper integral is convergent and, since the integrand is positive, we
FIGURE 10 can interpret the value of the integral as the area of the shaded region in Figure 10. n

EXAMPLE 6 Determine whether y

#y2
sec x dx converges or diverges.
0
SOLUTION Note that the given integral is improper because lim x ls#y2d2 sec x − `.
Using part (a) of Definition 3 and Formula 14 from the Table of Integrals, we have

y y |g
t
|
#y2 t
sec x dx − lim sec x dx − lim ln sec x 1 tan x 0
0 t l s#y2d2 0 t l s#y2d 2

− lim flnssec t 1 tan td 2 ln 1g − `

t l s#y2d2

because sec t l ` and tan t l ` as t l s#y2d2. Thus the given improper integral is
divergent. n

dx
EXAMPLE 7 Evaluate y
3
if possible.
0 x 2 1

SOLUTION Observe that the line x − 1 is a vertical asymptote of the integrand. Since
it occurs in the middle of the interval f0, 3g, we must use part (c) of Definition 3 with
c − 1:
dx dx dx
y0 x 2 1 − y0 x 2 1 1 y1 x 2 1
3 1 3

dx dx
y − lim2 y |g
t
|
1 t
where − lim2 ln x 2 1
0 x21 t l1 0 x 2 1 t l1 0

|
− lim2 (ln t 2 1 2 ln 21
t l1
| | |) − tlim
l1 2
lns1 2 td − 2`

because 1 2 t l 0 1 as t l 12. Thus y01 dxysx 2 1d is divergent. This implies that

y03 dxysx 2 1d is divergent. [We do not need to evaluate y13 dxysx 2 1d.] n

WARNING If we had not noticed the asymptote x − 1 in Example 7 and had instead
confused the integral with an ordinary integral, then we might have made the following
erroneous calculation:

dx
y |g
3
|
3
− ln x 2 1 0
− ln 2 2 ln 1 − ln 2
0 x21

This is wrong because the integral is improper and must be calculated in terms of limits.
From now on, whenever you meet the symbol yab f sxd dx you must decide, by looking at
the function f on fa, bg, whether it is an ordinary definite integral or an improper integral.

y
1
EXAMPLE 8 ln x dx.
0

SOLUTION We know that the function f sxd − ln x has a vertical asymptote at 0 since

Copyright 2016 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 SECTION 7.8 Improper Integrals 573

lim x l 01 ln x − 2`. Thus the given integral is improper and we have

y ln x dx − lim1 y ln x dx
1 1

0 t l0 t

Now we integrate by parts with u − ln x, dv − dx, du − dxyx, and v − x:

y ln x dx − x ln x t 2 y dx g
1 1 1

t t
y
− 1 ln 1 2 t ln t 2 s1 2 td − 2t ln t 2 1 1 t
To find the limit of the first term we use l’Hospital’s Rule:
0 1 x
ln t 1yt
area=1 lim t ln t − lim1 − lim1 − lim1 s2td − 0
t l 01 t l0 1yt t l 0 21yt 2 t l0

y
1
Therefore ln x dx − lim1 s2t ln t 2 1 1 td − 20 2 1 1 0 − 21
0 t l0
y=ln x
Figure 11 shows the geometric interpretation of this result. The area of the shaded
FIGURE 11 region above y − ln x and below the x-axis is 1. n

A Comparison Test for Improper Integrals

Sometimes it is impossible to find the exact value of an improper integral and yet it is
important to know whether it is convergent or divergent. In such cases the following
theorem is useful. Although we state it for Type 1 integrals, a similar theorem is true for
Type 2 integrals.
y
Comparison Theorem Suppose that f and t are continuous functions with
f f sxd > tsxd > 0 for x > a.
g (a) If ya` f sxd dx is convergent, then ya` tsxd dx is convergent.
(b) If ya` tsxd dx is divergent, then ya` f sxd dx is divergent.
0 a x
We omit the proof of the Comparison Theorem, but Figure 12 makes it seem plau-
FIGURE 12 sible. If the area under the top curve y − f sxd is finite, then so is the area under the
bottom curve y − tsxd. And if the area under y − tsxd is infinite, then so is the area
under y − f sxd. [Note that the reverse is not necessarily true: If ya` tsxd dx is convergent,
ya` f sxd dx may or may not be convergent, and if ya f sxd dx is divergent, ya` tsxd dx may
`

or may not be divergent.]

EXAMPLE 9 Show that y e2x dx is convergent.

` 2

0
2

y SOLUTION We can’t evaluate the integral directly because the antiderivative of e2x is
2
not an elementary function (as explained in Section 7.5). We write
y=e _x
y e2x dx − y e2x dx 1 y e2x dx
` 2 1 2 ` 2

0 0 1
y=e _x
and observe that the first integral on the right-hand side is just an ordinary definite inte-
gral. In the second integral we use the fact that for x > 1 we have x 2 > x, so 2x 2 < 2x
2
0 1 x and therefore e2x < e2x. (See Figure 13.) The integral of e2x is easy to evaluate:

y y
` t
FIGURE1313
FIGURE e2x dx − lim e2x dx − lim se21 2 e2t d − e21
1 tl` 1 tl`

7et070813
10/13/09
Copyright 2016 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
574 CHAPTER 7 Techniques of Integration

Therefore, taking f sxd − e2x and tsxd − e2x in the Comparison Theorem, we see that
2

y1` e2x dx is convergent. It follows that y 0 e2x dx is convergent.

2 ` 2
n
Table 1
y0 e2x dx
t 2
t In Example 9 we showed that y 0 e2x dx is convergent without computing its value. In
` 2

1 0.7468241328 Exercise 72 we indicate how to show that its value is approximately 0.8862. In probability
2 0.8820813908 theory it is important to know the exact value of this improper integral, as we will see in
3 0.8862073483 Section 8.5; using the methods of multivariable calculus it can be shown that the exact
4 0.8862269118 value is s# y2. Table 1 illustrates the definition of an improper integral by showing how
the (computer-generated) values of y0t e2x dx approach s# y2 as t becomes large. In fact,
2
5 0.8862269255
2
6 0.8862269255 these values converge quite quickly because e2x l 0 very rapidly as x l `.
1 1 e2x
EXAMPLE 10 The integral y
`
Table 2 dx is divergent by the Comparison Theorem
1 x
t y1t fs1 1 e2x dyxg dx
because
1 1 e2x 1
2 0.8636306042 .
5 1.8276735512 x x
10 2.5219648704
and y1` s1yxd dx is divergent by Example 1 [or by (2) with p − 1]. n
100 4.8245541204
1000 7.1271392134
Table 2 illustrates the divergence of the integral in Example 10. It appears that the
10000 9.4297243064
values are not approaching any fixed number.

1. Explain why each of the following integrals is improper.

y y
` 0
9. e 25p dp 10. 2 r dr
x 1 2
(a) y (b) y
2 ` 2`
dx 3 dx
1 x 2 1 0 1 1 x x2
y y
` `
11. dx 12. s y 3 2 3y 2 d dy
y y s1 1 x
` 2 2x 2 #y4 0 3 2`
(c) x e dx (d) cot x dx
2` 0
e21yx
y y
` `
2. Which of the following integrals are improper? Why? 13.
2
xe2x dx 14. dx
2` 1 x2
y y
#y4 #
(a) tan x dx (b) tan x dx
y y
` `
0 0 15. sin 2 $ d$ 16. sin " e cos" d"
0 0
dx
y y
1 ` 3
(c) (d) e2x dx
2 1 dv
x 2x22
y y
21 0 ` `
17. dx 18.
3. Find the area under the curve y − 1yx from x − 1 to x − t 3 1 x2 1 x 2 v 2 1 2v 2 3
and evaluate it for t − 10, 100, and 1000. Then find the total
y y
0 `
area under this curve for x > 1. 19. ze 2z dz 20. ye 23y dy
2` 2

; 4. (a) Graph the functions f sxd − 1yx and tsxd − 1yx in ln x ln x

1.1 0.9
y y
` `
21. dx 22. dx
the viewing rectangles f0, 10g by f0, 1g and f0, 100g 1x 1 x2
by f0, 1g. z 1
23. y y
0 `
(b) Find the areas under the graphs of f and t from x − 1 dz 24. dx
2` z 4 1 4 e xsln xd2
to x − t and evaluate for t − 10, 100, 10 4, 10 6, 10 10,
dx
and 10 20.
y y
` 2sy `
25. e dy 26.
(c) Find the total area under each curve for x > 1, if it exists. 0 sx 1 x sx
1

1 1
y 28. y
1 5
5–40 Determine whether each integral is convergent or divergent. 27. dx dx
Evaluate those that are convergent. 0 x s
0 3 5 2 x

dx x
y
1
y
1
y 30. y
` ` 14 2
5. dx 6. dx 29. dx
21 sx 1 1d 2
3 sx 2 2d3y2 0
s1 1 x
4 22
s
4
x12
1 1 1 dx
7. y y y y
0 ` 3 1
dx 8. dx 31. dx 32.
2` 3 2 4x 1 s2x 1 1d3 22 x4 0
s1 2 x 2

Copyright 2016 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.