The College Mathematics Journal

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/ucmj20

Improper Integrals: An Alternative Criterion

Katiuscia Teixeira

To cite this article: Katiuscia Teixeira (2023) Improper Integrals: An Alternative Criterion, The
College Mathematics Journal, 54:3, 232-234, DOI: 10.1080/07468342.2023.2201566
To link to this article: https://doi.org/10.1080/07468342.2023.2201566

Published online: 03 May 2023.

Submit your article to this journal

Article views: 160

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=ucmj20
CLASSROOM CAPSULES
EDITORS
Ricardo Alfaro Lixing Han Kenneth Schilling
University of Michigan-Flint University of Michigan-Flint University of Michigan-Flint
ralfaro@umflint.edu lxhan@umflint.edu ksch@umflint.edu
Classroom Capsules are short (1–3 pages) notes that contain new mathematical insights on a topic
from undergraduate mathematics, preferably something that can be directly introduced into a college
classroom as an effective teaching strategy or tool. Classroom Capsules should be prepared according
to the guidelines on the inside front cover and submitted through Editorial Manager.

Improper Integrals: An Alternative Criterion

Katiuscia Teixeira (katiuscia.teixeira@ucf.edu), University of Central Florida.
The topic Improper Integrals, often introduced in the second course of Calculus, is
an important, though difficult concept for students to grasp, viz. [1–3].
In this article we discuss an alternative (geometric) criterion for an improper integral
to diverge. While the criterion is indeed efficient and easy to apply, if one believes, like
I do, that teaching Calculus is more than training students to manipulate formulas, then
the opportunity to present and discuss the reasoning leading to such a result should be
thought as more valuable than the criterion, per se.
Let us start off with a classical example of divergent integral:
 1
1
dx = +∞. (1)
0 x

This can be obtained by—and it is often presented as—a direct application of the
Fundamental Theorem of Calculus, i.e.,:
 1  1
1 1
dx := lim dx = lim [ln 1 − ln t] = +∞.
0 x t→0+ t x t→0+

While certainly powerful, such a solution hides a beautiful geometric interpretation

rooted in the very motivation of integrals, namely the calculus of areas.
Indeed, if one looks that the area of the degenerating rectangles of base [0, t] and
height 1t , one immediately sees that they all have constant area 1, no matter how small
t > 0 is. The fact that one can find rectangles with arbitrarily small base and constant
area explains geometrically why such an improper integral must diverge.

Figure 1. The geometric idea: since all such rectangles have area 1, the improper integral
diverges.
doi.org/10.1080/07468342.2023.2201566

232 © THE MATHEMATICAL ASSOCIATION OF AMERICA

The key point to note is that whether an improper integral converges or diverges de-
pends upon the (sum of the) areas of those degenerating rectangles, near the vertical
asymptote. If they all contributes with a constant amount, then the improper integral
diverges.
Such a result can certainly be analyzed through Riemann sums, but we present here
an alternative approach which may be easier to grasp:
 1  2t  1 
1 1 1
dx := lim dx + dx
0 x t→0+
t 2t x 2t 1x 
1 1
≥ lim dx + dx
t→0+ 2t
t 
1 2t x
1
≥ 12 + lim dx
+
 1
t→0 2t x
1 1
= + dx,
2 0 x

1
where in the second line we have used that the function is decreasing and thus in
 1 x
1
[t, 2t], one has x ≥ 2t . Clearly the inequality above implies
1 1
dx cannot be a real
0 x
number, and hence such an improper integral diverges.
The reasoning above could possibly be better explained as a “contradiction argu-
 1
1
ment,” that is: assuming the improper integral dx converges, say to a limit num-
0 x
ber L, one would end with the inequality: L ≥ 2 + L, leading to a contradiction.
1

While rather simple, this argument yields an efficient, more direct and didactical
criterion for divergent integrals, namely:
Theorem 1. Let f : (a, b] → R be a continuous function defined on a bounded in-
terval. Assume for some a < c < b, the function f (x) is nonincreasing in (a, c). Then
 b
lim (x − a)f (x) = μ > 0 =⇒ f (x) = +∞.
x→a + a

Proof. The proof is just a mere generalization of the argument explained above. In-
deed, one can write
 b  a+2t  b 
f (x)dx := lim f (x)dx + f (x)dx
t→0+
a a+t
a+2t
a+2t 
b 
≥ lim f (a + 2t)dx + f (x)dx
t→0+ a+t  b a+2t 
≥ lim tf (a + 2t) + lim f (x)dx
t→0+ t→0
 b a+2t
(x − a)
= lim f (x) + f (x)dx
x→0+
 b2 a
μ
= + f (x)dx.
2 a

In the second line we have used that for t small enough, a + 2t < c, along with
the assumption that f is nonincreasing in (a, c). In the third line, we computed
lim (x−a)
2
f (x) by the change of variables, a + 2t = x.
x→0+

VOL. 54, NO. 3, MAY 2023 THE COLLEGE MATHEMATICS JOURNAL 233
  b
μ
Arguing as before, since 2
is positive, the above inequality implies that f (x)dx
a
cannot be a real number, and thus the improper integral must diverge. 
Remark. It’s worth noticing that one could use the “comparison theorem” to confirm
the thesis of Theorem 1. Indeed, if one assumes lim (x − a)f (x) = μ > 0, then, for
x→a +
 a+δ 1
some δ > 0, we have f (x) > 2 · x−a , for all a < x < a + δ. Since a x−a
μ 1
dx =
+∞, the conclusion follows by comparison. The argument presented above, though,
is arguably more pedagogical and its visual representation more appealing.
Here are few examples elucidating the applicability of such a criterion.
1
Example 1. The improper integral 0 sin1 x dx diverges. Indeed, one simply calculates
lim t · sin1 t = 1.
t→0+
1 1
Example 2. The improper integral 0 x 1−x dx diverges. To verify that, one computes
lim t · t 1−t
1
= lim t t = 1.
t→0+ t→0+
Example 3. Easily one generalizes the previous Example to check that the improper
1 1 1
integral 0 x 1−1√x dx diverges, and in fact 0 x 1−x α dx diverges, for all α > 0. That’s
tα
because lim t · t 1−t α = lim t = 1.
1
t→0+ t→0+
The very same argument can be easily employed as a criterion for improper integrals
of functions blowing-up at the right-end point of the interval. That is, if f : (a, b) → R
is a continuous function on a bounded interval with lim f (x) = +∞. Then, if for
x→b−
some a < c < b, the function f (x) is nondecreasing in (c, b), one has
 b
lim (b − x)f (x) = μ > 0 =⇒ f (x) = +∞.
x→b−
 π/2 a
Example 4. The improper integral 0 tan xdx diverges. Indeed,
π
lim ( − t) · tan t = 1.
t→ π2 − 2
Similar analysis can also be done for nonnegative continuous functions f : [a, ∞) →
R. If the degenerating rectangles
 ∞ of base [a, x] and height [0, f (x)] have uniform pos-
itive area as x → ∞, then a f (x)dx must diverge. The argument is exactly the
same. We invite the readers to state the result and modify the proof of Theorem 1 to
verify it.
Acknowledgment. The author thanks the anonymous referee for several insightful comments
and remarks that benefited the final outcome of this capsule.
ORCID
Katiuscia Teixeira https://orcid.org/0000-0002-1904-7592

References

[1] Sofronas, K. S., DeFranco, T. C., Vinsonhaler, C., Gorgievski, N., Schroeder, L., Hamelin, C. (2011). What
does it mean for a student to understand the first-year calculus? Perspectives of 24 experts. J. Math. Behav.
30(2): 131–148.
[2] González-Martı́n, A. S., Camacho, M. (2004). What is first-year Mathematics students’ ac-
tual knowledge about improper integrals? Int. J. Math. Educ. Sci. Technol. 35(1): 73–89.
doi.org/10.1080/00207390310001615615
[3] Mojica, D., Vasquez, A. R., Jesus, F. R. M. V. (2020). Understanding of University students about improper
integral. Int. J. Educ. Res. Methodol. 11: 36–54. doi.org/10.24297/ijrem.v11i.8681

234 © THE MATHEMATICAL ASSOCIATION OF AMERICA