Sequence Convergence & Limits — Sequence Convergence and Limits - Calculus 2

SECTION 1

Historical Context & Motivation

The question of whether an infinite process can yield a finite result has captivated mathematicians for millennia. Ancient Greek philosophers, most notably Zeno of Elea, posed paradoxes — such as Achilles and the Tortoise — that implicitly demanded a rigorous theory of limits. These paradoxes revealed that intuition alone could not resolve the behavior of infinite sequences; a formal framework was needed. The intellectual journey from Zeno's puzzles to the modern ε-N definition of convergence spans over two thousand years and represents one of the most profound achievements in mathematical analysis.

~450 BCE • Zeno's Paradoxes

Zeno of Elea formulates paradoxes involving infinite subdivision of motion, implicitly raising questions about the convergence of infinite sums and sequences that would remain unresolved for centuries.

1660s • Newton & Leibniz

Newton's method of fluxions and Leibniz's differential calculus rely heavily on intuitive notions of limits and infinitesimals. Both use sequences implicitly in series expansions, but neither provides a rigorous definition of convergence.

1817 • Bolzano's Contributions

Bernard Bolzano publishes work requiring sequences to satisfy conditions now recognized as the Cauchy criterion. He begins to articulate the need for arithmetic, rather than geometric, foundations for limits.

1821 • Cauchy's Cours d'Analyse

Augustin-Louis Cauchy systematically defines limits of sequences and functions, introduces the notion of a Cauchy sequence, and provides the first semi-rigorous treatment of convergence that moves calculus toward analytic precision.

1850s–1870s • Weierstrass Formalizes ε-δ and ε-N

Karl Weierstrass and his students complete the arithmetization of analysis by formalizing the ε-N definition of sequence convergence and the ε-δ definition of function limits, establishing the rigorous standards used in modern mathematics.

The central question that drove this centuries-long development was deceptively simple: what does it mean, precisely, for the terms of an infinite sequence to "approach" a number? Without a formal answer, mathematicians could not reliably determine whether series converge, whether function limits exist, or whether the fundamental theorem of calculus rests on solid ground. The ε-N definition resolved this foundational crisis and serves as the gateway to all of modern real analysis, making it the essential starting point for Calculus 2 students studying sequences and series.

SECTION 2

Core Principles & Definitions

A sequence is a function from the positive integers ℕ to the real numbers ℝ, typically written as {a_n} or (a₁, a₂, a₃, …). Unlike finite lists, sequences extend indefinitely, and their long-run behavior — whether they settle down, oscillate, or blow up — is the subject of convergence theory. The following foundational ideas structure our entire study.

Convergence

A sequence {a_n} converges to a limit L if, for every ε > 0, there exists an integer N such that |a_n − L| < ε whenever n ≥ N. The terms eventually stay arbitrarily close to L.

Divergence

A sequence diverges if no real number L satisfies the convergence definition. Divergence can manifest as unbounded growth (divergence to ±∞) or persistent oscillation without settling.

Uniqueness of Limits

If a sequence converges, its limit is unique. This follows directly from the triangle inequality: assuming two distinct limits leads to a contradiction for sufficiently small ε.

Boundedness

Every convergent sequence is bounded — there exists M > 0 such that |a_n| ≤ M for all n. The converse is false: {(−1)ⁿ} is bounded but divergent.

Monotone Convergence Theorem

A sequence that is both monotone (non-increasing or non-decreasing) and bounded must converge. This powerful theorem guarantees existence of limits without requiring us to know the limit in advance.

✦

KEY TAKEAWAY

Think of convergence like a spacecraft docking with a space station. Early in its approach, the spacecraft may be far away and maneuvering broadly (analogous to the early terms of the sequence). But past a certain point — once it enters the docking corridor — the spacecraft must remain within ever-tighter tolerances of the target. The ε-N definition formalizes exactly this: no matter how narrow the corridor (ε), there is a point in time (N) after which the sequence never leaves that corridor around L.

SECTION 3

Visual Explanation — The ε-N Definition

The ε-N definition of convergence has an elegant geometric interpretation. When we say lim a_n = L, we mean that for any horizontal ε-band centered on L (the strip from L − ε to L + ε), there exists an index N beyond which all terms of the sequence lie inside the band. The following diagram illustrates this for the sequence a_n = 1/n converging to L = 0.

The shaded cyan band represents the interval (L − ε, L + ε). The dashed amber line marks the threshold index N. For every n ≥ N, the green dots (sequence terms) remain within the ε-band, satisfying |a_n − L| < ε. Notice that shrinking ε forces N to increase — a smaller band demands more terms before the sequence "locks in."

The critical insight from this diagram is the interplay between ε and N. A skeptic can always propose a smaller ε — say ε = 0.001 instead of ε = 0.1 — and the convergence definition guarantees we can find a correspondingly larger N (in this case, N = 1001) so that all subsequent terms remain within the tighter band. This "challenge-response" structure is why the definition begins with "for every ε > 0": convergence must hold no matter how demanding the tolerance.

SECTION 4

Mathematical Framework

The formal apparatus of sequence convergence rests on a single definition and several powerful theorems that flow from it. Mastering these tools allows you to prove convergence rigorously, compute limits algebraically, and identify divergence efficiently.

ε-N DEFINITION OF CONVERGENCE

lim(n→∞) aₙ = L ⟺ ∀ε > 0, ∃N ∈ ℕ such that n ≥ N ⟹ |aₙ − L| < ε

Here ε (epsilon) is an arbitrary positive real number representing the tolerance, N is the threshold index (depending on ε), and L is the limit. The absolute value |aₙ − L| measures the distance from the n-th term to L.

LIMIT LAWS FOR SEQUENCES

lim(aₙ ± bₙ) = lim aₙ ± lim bₙ ; lim(aₙ · bₙ) = lim aₙ · lim bₙ ; lim(aₙ / bₙ) = lim aₙ / lim bₙ (lim bₙ ≠ 0)

These algebraic limit laws hold provided each individual limit exists (is finite). They enable us to decompose complicated sequences into simpler parts whose limits are known.

SQUEEZE THEOREM (SANDWICH THEOREM)

If aₙ ≤ cₙ ≤ bₙ for all n ≥ N₀ and lim aₙ = lim bₙ = L, then lim cₙ = L

The Squeeze Theorem is invaluable when a direct ε-N proof is difficult. By bounding cₙ between two sequences that share a common limit, we force cₙ to converge to that same limit.

MONOTONE CONVERGENCE THEOREM

If {aₙ} is monotone increasing and bounded above, then lim(n→∞) aₙ = sup{aₙ : n ∈ ℕ}

Analogously, a monotone decreasing sequence bounded below converges to its infimum. The key power of this theorem lies in establishing existence of a limit even when its exact value is unknown, leveraging the completeness of ℝ.

💡

Common Limit Results to Know

Several standard limits appear repeatedly: lim(1/n^p) = 0 for p > 0; lim(rⁿ) = 0 for |r| < 1; lim(n^1/n) = 1; and lim((1 + 1/n)ⁿ) = e ≈ 2.71828. Memorizing these benchmarks accelerates limit evaluation in more complex sequences.

SECTION 5

Types of Sequence Behavior

Not all sequences converge, and understanding the taxonomy of sequence behavior sharpens both intuition and proof strategy. Sequences can be classified along two independent axes: monotonicity (whether terms consistently increase or decrease) and boundedness (whether terms remain within a finite interval). The interplay of these properties determines convergence, as codified by the Monotone Convergence Theorem.

Four representative sequences illustrating convergent versus divergent behavior. The top-left (green) shows monotone convergence; the top-right (violet) shows oscillating convergence; the bottom-left (red) diverges to infinity; and the bottom-right (amber) oscillates without converging, despite being bounded.

Property	Bounded?	Monotone?	Convergent?
aₙ = 1/n	Yes	Yes (decreasing)	Yes → 0
aₙ = (−1)ⁿ/n	Yes	No	Yes → 0
aₙ = (−1)ⁿ	Yes	No	No (oscillates)
aₙ = n²	No	Yes (increasing)	No (→ ∞)
aₙ = (1 + 1/n)ⁿ	Yes	Yes (increasing)	Yes → e

Summary of boundedness, monotonicity, and convergence for common sequences

SECTION 6

Worked Example — ε-N Proof

Let us prove rigorously that the sequence a_n = (3n + 1)/(n + 2) converges to L = 3 using the ε-N definition. This example demonstrates the standard proof technique: we start from |a_n − L| < ε, simplify, and solve for n in terms of ε to find N.

Prove that lim(n→∞) (3n + 1)/(n + 2) = 3

1
Step 1 — Compute |aₙ − L|

We compute the absolute difference between a_n and the proposed limit L = 3. We have |a_n − 3| = |(3n + 1)/(n + 2) − 3| = |(3n + 1 − 3(n + 2))/(n + 2)| = |(3n + 1 − 3n − 6)/(n + 2)| = |−5/(n + 2)|.

|a_n − 3| = 5/(n + 2)
2
Step 2 — Set up the inequality

We need 5/(n + 2) < ε. Since n ≥ 1 implies n + 2 > 0, the absolute value bars drop. Solving for n: n + 2 > 5/ε, so n > 5/ε − 2.

n > 5/ε − 2 guarantees |a_n − 3| < ε
3
Step 3 — Choose N

We choose N to be any natural number greater than 5/ε − 2. Formally, let N = ⌈5/ε − 2⌉ + 1, where ⌈·⌉ denotes the ceiling function. This ensures N ∈ ℕ and N > 5/ε − 2.

N = ⌈5/ε − 2⌉ + 1
4
Step 4 — Write the formal proof

Let ε > 0 be given. Choose N = ⌈5/ε − 2⌉ + 1. Then for all n ≥ N, we have n > 5/ε − 2, which gives n + 2 > 5/ε, so 5/(n + 2) < ε. Therefore |a_n − 3| = 5/(n + 2) < ε. Since ε was arbitrary, lim(n→∞) (3n + 1)/(n + 2) = 3. □

lim(n→∞) (3n + 1)/(n + 2) = 3 ■
5
Step 5 — Numerical verification

As a sanity check, take ε = 0.01. Then N = ⌈5/0.01 − 2⌉ + 1 = ⌈498⌉ + 1 = 499. At n = 499: a₄₉₉ = (3 × 499 + 1)/(499 + 2) = 1498/501 ≈ 2.99002, and |2.99002 − 3| = 0.00998 < 0.01. ✓

a₄₉₉ ≈ 2.99002, confirming |a₄₉₉ − 3| < 0.01

SECTION 7

Convergence Techniques — Strengths & Limitations

In practice, writing a full ε-N proof for every sequence is neither necessary nor efficient. Several techniques — each with characteristic strengths and limitations — allow us to determine convergence more swiftly. Knowing which tool to reach for in a given situation is itself a key skill in analysis and Calculus 2.

Technique	Strengths	Limitations
Direct ε-N Proof	Definitive; proves convergence from first principles. Required in formal analysis courses.	Can be algebraically complex. Requires guessing L in advance.
Limit Laws (Algebraic)	Fast for rational expressions. Decompose complex sequences into simpler parts.	Only valid when component limits exist individually. Cannot handle indeterminate forms directly.
Squeeze Theorem	Ideal for oscillating sequences (e.g., sin(n)/n). Avoids direct limit computation.	Requires finding tight upper and lower bounding sequences with the same limit.
Monotone Convergence Theorem	Guarantees existence without knowing L. Natural for recursively defined sequences.	Must prove monotonicity and boundedness separately. Does not directly give the limit's value.
Continuous Function Theorem	If aₙ → L and f is continuous at L, then f(aₙ) → f(L). Handles compositions elegantly.	Requires the inner sequence to already converge and f to be continuous at the limit point.

Comparison of convergence techniques for sequences

🧭

STRATEGY GUIDE

Selecting a convergence technique is like choosing the right tool in a workshop. The ε-N definition is your precision caliper — universally applicable but labor-intensive. Limit laws are your power drill — fast and effective for standard jobs. The Squeeze Theorem is your clamp — essential when you need to pin down a sequence that wiggles unpredictably. And the Monotone Convergence Theorem is your level — it tells you the limit exists (the shelf is flat) even when you haven't measured its exact height yet.

SECTION 8

Connection to Series & Advanced Topics

Sequence convergence is not an isolated topic — it forms the foundation upon which the entire theory of infinite series, power series, and function approximation rests. A series ∑a_n converges if and only if its sequence of partial sums S_N = a₁ + a₂ + ⋯ + a_N converges as a sequence. Thus every convergence test for series — the comparison test, ratio test, root test, integral test — ultimately reduces to analyzing the convergence of a particular sequence.

Concept	Sequence Connection	Where You'll See It
Infinite Series	Convergence of ∑aₙ ⟺ convergence of the partial sum sequence {Sₙ}	Calculus 2: series convergence tests
Power Series	Radius of convergence determined by lim\|aₙ₊₁/aₙ\| or lim\|aₙ\|^(1/n)	Calculus 2: Taylor/Maclaurin series
Cauchy Sequences	{aₙ} is Cauchy if terms get mutually close — equivalent to convergence in ℝ	Real Analysis (Math 300+)
Metric Space Convergence	Generalize \|aₙ − L\| < ε to d(aₙ, L) < ε for abstract distance functions d	Topology / Functional Analysis

How sequence convergence connects to advanced mathematical topics

One of the most important consequences of sequence convergence theory is the Divergence Test (also called the n-th Term Test): if lim a_n ≠ 0 (or does not exist), then the series ∑a_n diverges. This is typically the first test applied to any series, and it relies entirely on the sequence convergence concepts developed in this lesson. As you progress through Calculus 2 and into real analysis, you will see that the ε-N language introduced here becomes the universal grammar for discussing limits in every mathematical context.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

Explain why the sequence a_n = (−1)ⁿ diverges, even though it is bounded. In particular, explain what goes wrong when you attempt to apply the ε-N definition with any candidate limit L.

Show answer

The sequence alternates between −1 and 1 and never settles near any single value. Suppose L were a limit and choose ε = 0.5. Then we would need |a_n − L| < 0.5 for all sufficiently large n. But for even n, a_n = 1, and for odd n, a_n = −1. If L is within 0.5 of 1, it is at least 1.5 away from −1 (which exceeds ε). No L simultaneously satisfies both requirements, so no N can be found. Boundedness is necessary but not sufficient for convergence — monotonicity (or some other structural property) is also needed.

Question 2: BASIC CALCULATION

Find lim(n→∞) (5n² − 3n)/(2n² + 7n + 1) using limit laws.

Show answer

Divide numerator and denominator by n² (the highest power of n): (5 − 3/n)/(2 + 7/n + 1/n²). As n → ∞, 3/n → 0, 7/n → 0, and 1/n² → 0 by standard limit results. By the limit laws (sum, quotient), the expression converges to (5 − 0)/(2 + 0 + 0) = 5/2. Therefore the limit is 5/2. This technique of dividing by the dominant power is the standard approach for rational sequences.

Question 3: INTERMEDIATE

Use the Squeeze Theorem to prove that lim(n→∞) sin(n)/√n = 0.

Show answer

Since −1 ≤ sin(n) ≤ 1 for all n, we can bound: −1/√n ≤ sin(n)/√n ≤ 1/√n. Now lim(n→∞) (−1/√n) = 0 and lim(n→∞) (1/√n) = 0 (since 1/√n = 1/

n^{1/2}

and p = 1/2 > 0). By the Squeeze Theorem, since the lower and upper bounding sequences both converge to 0, we conclude lim(n→∞) sin(n)/√n = 0. This is a prototypical application: the oscillating factor sin(n) is controlled by dividing by a growing quantity √n.

Question 4: APPLIED

A drug is administered so that the concentration after n doses is C_n = 50(1 − 0.4ⁿ) mg/L. Determine whether {C_n} converges, find its limit if so, and find the smallest n for which C_n is within 1 mg/L of the steady-state concentration.

Show answer

Since |0.4| < 1, the geometric sequence 0.4ⁿ → 0. By the continuous function theorem (f(x) = 50(1 − x) is continuous), C_n → 50(1 − 0) = 50 mg/L. To find when |C_n − 50| < 1: |50 × 0.4ⁿ| < 1 ⟹ 0.4ⁿ < 0.02 ⟹ n > ln(0.02)/ln(0.4) ≈ 3.912/0.916 ≈ 4.27. So n = 5 is the smallest integer. Check: C₅ = 50(1 − 0.4⁵) = 50(1 − 0.01024) = 49.488, and |49.488 − 50| = 0.512 < 1. ✓ This models pharmacokinetic steady-state accumulation.

Question 5: CRITICAL THINKING

Define a₁ = 2 and a_n+1 = (1/2)(a_n + 3/a_n). Prove that {a_n} converges and find its limit. (Hint: show that a_n ≥ √3 for n ≥ 2, and that the sequence is eventually decreasing.)

Show answer

First, by AM-GM inequality, a_n+1 = (a_n + 3/a_n)/2 ≥ √(a_n × 3/a_n) = √3, so {a_n} is bounded below by √3 for n ≥ 2. Next, a_n+1 − a_n = (3/a_n − a_n)/2 = (3 − a_n²)/(2a_n). Since a_n ≥ √3, we have a_n² ≥ 3, making the numerator ≤ 0. So the sequence is decreasing for n ≥ 2. By the Monotone Convergence Theorem, {a_n} converges. Let L = lim a_n. Taking limits of both sides of the recurrence: L = (L + 3/L)/2 ⟹ 2L = L + 3/L ⟹ L = 3/L ⟹ L² = 3 ⟹ L = √3 (positive root). This is the Babylonian method for computing square roots — a beautiful intersection of numerical methods and convergence theory.

SUMMARY

Lesson Summary

Summary

A sequence {aₙ} converges to a limit L if, for every ε > 0, there exists a threshold index N beyond which all terms satisfy |aₙ − L| < ε. This ε-N definition, formalized by Weierstrass in the 19th century, transforms the intuitive notion of "approaching a value" into a precise, provable statement. Key supporting results include the limit laws (which allow algebraic decomposition), the Squeeze Theorem (for bounding oscillatory sequences), and the Monotone Convergence Theorem (which guarantees existence of limits for bounded monotone sequences).

Every convergent sequence is bounded, and its limit is unique. However, boundedness alone does not guarantee convergence — oscillating sequences like (−1)ⁿ are bounded but divergent. Sequence convergence serves as the gateway to the theory of infinite series, power series, and Cauchy sequences in real analysis, making it one of the most foundational concepts in all of calculus.