Limit of sequence lecture notes

Reference: Limit and continuity by C.S. Lee 1986 Fillans Limited

Edited by Mr. Francis Hung on March 28, 2009 Last updated: 24 March 2023
1. Definition of limit of sequence.

Consider the sequence: xn n1 .

lim xn exists and equal to l if the following condition is satisfied:
n

  > 0,  N  ℕ such that  n > N, |xn – l| < .

Otherwise, we say that lim xn does not exist; or the sequence diverges or not convergent.
n

Example 1.1

1 
Consider the sequence:   .
 n  n1
1  1 1
  > 0, let N =    1  ℕ, then N >  < 
  N
1 1 1
such that  n > N, |xn – 0| = = < < .
n n N
1
 lim = 0
n  n

Example 1.2

 n  1
Consider the sequence:   .
 n n 1

1  1 1
  > 0, let N =    1  ℕ, then N >  < 
  N
n 1 1 1
such that  n > N, |xn – 1| =  1 =  < .
n n N
n 1
 lim =1
n  n

Example 1.3

1
Consider the sequence:  n  .
 2  n1

 log 1  log 1 1 1
  > 0, let N =    1  ℕ, then N > N log 2 > log 1 2N >  N < 
 log 2  log 2  2

1 1 1
such that  n > N, |xn – 0|= n
 0 = n  N < .
2 2 2
1
 lim =0
n 2 n

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 1
Limit of sequence notes Edited by Mr. Francis Hung
1. Definition of limit of sequence

Example 1.4
Consider the sequence:  n   
n 1 , where 0  .
Claim: Bernoulli inequality If x  –1, then (1 + x)n  1 + nx, n  ℕ
Proof: Induction on n. n = 1, (1 + x)1 = 1 + x, the result is obvious.
Suppose (1 + x)k  1 + kx
Multiply both sides by (1 + x), which is non-negative.
(1 + x)n+1  (1 + nx)(1 + x)
(1 + x)n+1  1 + (n + 1)x + nx2  1 + (n + 1)x

By MI, if x  –1, then (1 + x)n  1 + nx, n  ℕ

If  = 0, lim  n = 0; if  = 1, lim  n = 1
n n

If  > 1, let  = 1 + x; where x > 0

By Bernoulli inequality, (1 + x)n  1 + nx  (1 + x)n  nx  (1)

Claim lim  n does not exist for  > 1.

n

Proof: Suppose on the contrary, lim  n exists and equal to l.

n

Clearly n > 0 and l > 0

  > 0,  N  ℕ such that  n > N, |n – l| < 
– < n – l < 
– + l < n <  + l
nx  (1 + x)n <  + l

n
x
2 2
Let  = l, n   n > N, which means that n is bounded above by .
x x
That is a contradiction.

 lim  n does not exist for  > 1.

n

1
If 0 <  < 1, let  = ; where x > 0
1 x
1 1
By Bernoulli inequality, (1 + x)n  1 + nx    (2)
1  x n
nx
1 1 1
  > 0, let N =   + 1  ℕ, then N   
 x  x Nx
1 1 1
 n > N, n > |n – 0| = n =  <
x 1  x  nx
n

 lim  n = 0
n

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 2
Limit of sequence notes Edited by Mr. Francis Hung
1. Definition of limit of sequence

Example 1.5

Prove that if a > 1, then lim n a = 1.

n

Let n a = 1 + n, n > 0

By Bernoulli inequality, a = (1 + n)n  1 + nn
a 1
n 
n
 a  1 a 1 a 1
Hence, for any given  > 0, let N =   1 N > >
    N
a 1 a 1
 n > N, | n a – 1| = |1 + n – 1| = |n| = n  < <
n N

Thus, by definition, lim n a = 1.

n

Example 1.6

Prove that if 0 <  < 1, then lim n n = 0.

n

1 1
> 1; Let = 1 + h, where h > 0.
 
1 nn  1 2 nn  1 2
= (1 + h)n = 1 + nh + h +> h for n  2
 n
2 2
2 2
n <  nn <
nn  1h 2
n  1h 2
 2  2 2
For any given  > 0, let N =  2   2  N – 1 > 2   >
 h  h N  1h 2
2 2
 n > N, | nn – 0| = nn < < <
n  1h N  1h 2
2

Thus, by definition, lim n n = 0 .

n

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 3
Limit of sequence notes Edited by Mr. Francis Hung
2. Divergent sequence
2. Divergent sequences
(i) The sequence an = 32n–1 diverges to positive infinity.
(ii) The sequence bn = 1 – 2n diverges to negative infinity.
(iii) The sequence cn = (–1)n oscillates (between 1 and –1).
(iv) The sequence dn = (–1)nn oscillates divergent (to ).

Definition 2.1: The sequence {an} is said to tend to infinity (+) if given any real number M (however

large), there exists N  ℕ such that an > M for all n > N.

We write lim an   .
n

Similarly, we write lim an   if given any real number M (however small), there exists N  ℕ
n

such that an < M for all n > N.

It should be emphasized that  and – are not positive numbers and the sequences are not convergent.
Thus,

(i) lim 32 n1  

n

(ii) lim1  2n    .
n

Example 2.1

Prove by definition that (a) lim 32 n1   ; (b) lim1  2n    .

n n

 1  log M 
(a)  M ℝ, let N =    1  1 , then 32N–1 > M
 2  log 3 

 n > N, 32n–1 > 32N–1 > M

 lim 32 n1  
n

1  M 
(b)  M ℝ, let N =   1 , then 1 – 2N < M
 2 

 n > N, 1 – 2n < 1 – 2N < M

 lim1  2n   
n

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 4
Limit of sequence notes Edited by Mr. Francis Hung
2. Divergent sequence
Example 2.2

 1 1 1 
Let an =     . (sum of n + 1 terms) Prove that lim an   .
 n n 1 2n  n

1
Observe that the smallest term is .
2n
N
 M ℝ, let N = [2M2] + 1, then N > 2M2  >M
2
1 1 1 1 1 n 1 n n N
 n > N,         M
n n 1 2n 2n 2n 2n 2n 2 2
n 1 terms

 an > M  lim an  
n

Definition 2.2
If an does not tend to a limit or to  or to –, we say that an oscillates (or is an oscillating sequence).
If an oscillates and is bounded, it oscillates finitely. If an oscillates and is not bounded, it oscillates
infinitely.

Example 2.3

1. cn = (–1)n oscillates finitely (between 1 and –1).

2. dn = (–1)nn oscillates infinitely.

3. The sequence an =
 1
n
is not an oscillating sequence. It has a limit = 0.
n

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 5
Limit of sequence notes By Mr. Francis Hung
3. Uniqueness of limit
Theorem 3.1 A sequence can converge to only one limit, i.e. if a limit exists, it is unique.

Proof: Let xn n1 be the given sequence. Try show that if lim xn = a and lim xn = b, then a = b.
n n

By hypothesis, given any  > 0, we can find N1 and N2  ℕ such that


|xn – a| < whenever n > N1
2

and |xn – b| < whenever n > N2
2
then, whenever n > N = max{N1, N2}, we have
|a – b| = |a – xn + xn – b|  |a – xn| + |xn – b| (by triangle inequality)
 
= |xn – a| + |xn – b | < + =
2 2
i.e. |a – b| is less than any positive  (however small) and so must be zero. Thus a = b.

1 1
Example 3.1 If an = sin , then |an| = sin  1, therefore, an is a bounded sequence.
n n

Theorem 3.2 If lim xn = l, then the sequence is bounded.

n

Proof:   > 0,  N  ℕ such that  n > N, |xn – l| <  .

Let  = 1, | |xn| – |l| | < |xn – l| < 1
–1 < |xn| – |l| < 1
|xn| < 1 + |l|  n > N

Let M = Max{|x1|, |x2|,  , |xN|, 1 + |l|}

 xn n1 is bounded above by M.

We remark that
(1) In other words, if {an} is bounded, {an} is not convergent. For example, the sequence {an}

 1 1 1 
defined by an =     is not convergent. Since
 n n 1 2n 

1 1 1 n n N
an =        M (however large)
n n 1 2n 2n 2 2
this mean {an} is not bound, therefore {an} is not convergent.

(2) The converse of the theorem is not true in general. For example, the sequence {an} defined by
an = (–1)n , then –1  an  1, that is, {an} is bounded, but {an} is not convergent.
It is an oscillating sequence.

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 6
Limit of sequence notes Edited by Mr. Francis Hung
4. Theorems on limits
Theorem 4.1 If lim an = l1 and lim bn = l2 , then (1) lim  an  bn  = l1 + l2; (2) lim  an  bn  = l1 – l2.
n  n  n  n 

Proof: By hypothesis, for any given  > 0, we can find N1 and N2 such that
1
|an – l1| <  for all n > N1, and
2
1
|bn – l2| <  for all n > N2
2
then, for any given  > 0, we can find N = max{N1, N2} such that
|(an + bn) – (l1 + l2)| = |(an – l1) + (bn – l2)|
 |an – l1| + |bn – l2| by triangle inequality
1 1
<  +  =  for all n > N
2 2
By definition, we have lim  an  bn  = l1 + l2 .
n 

Also, (an – bn) – (l1 – l2)| = |(an – l1) + (l2 – bn)|

 |an – l1| + |l2 – bn| by triangle inequality
= |an – l1| + |bn – l2|
1 1
<  +  =  for all n > N
2 2
By definition, we have lim  an  bn  = l1 – l2 .
n 

Theorem 4.2 If lim an = l1 and lim bn = l2 , then lim  anbn  = lim an  lim bn = l1l2
n  n  n  n  n 

Proof: Since lim an = l1, it is bounded by P, i.e. |an|  P for all n ℕ for some positive constant P.
n 

By hypothesis, for any given  > 0, we can find N1 and N2, such that

|an – l1| < for all n > N1, and
2   2  1

|bn – l2| < for all n > N2
2P
Now, for any given  > 0, we can find N = max{N1, N2} such that
|anbn – l1l2| = |an(bn – l2) + l2(an – l1)|
 |an|||bn – l2| + |l2||an – l1| by triangle inequality
 P|bn – l2| + (|l2|+ 1)|an – l1|
 
< P    2  1 
2P 2   2  1
1 1
<  +  =  for all n > N
2 2
Therefore, by definition, we have lim  anbn  = lim an  lim bn = l1l2 .
n  n  n 

1
Lemma If lim bn = l2  0, then there exists a natural number N such that |bn| >  2 for all n > N .
n  2
1
Proof: By hypothesis we can find N such that |bn – l2| <  2 for all n > N
2
|l2| = |l2 – bn + bn|  |l2 – bn| + |bn|
1
= |bn – l2| + |bn| <  2 + |bn| for all n > N .
2
1
Which gives |bn| >  2 for all n > N.
2

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 7
Limit of sequence notes By Mr. Francis Hung
4. Theorems on limits
1 1 1
Theorem 4.3 If lim bn = l2  0, then lim = = .
n  n  bn lim bn  2
n 

1
Proof: By hypothesis, for any  > 0, we can find N1 such that |bn – l2| <  22 for all n > N1 .
2
1
Also, by the above lemma, we can find N2 such that |bn| >  2 for all n > N2 .
2
Thus, for any given  > 0, we can find N = max{N1, N2} such that
1 2
1 1   b  2
 = 2 n
< 2 =  for all n > N
bn  2 bn  2 1
2  2
2
1 1 1
Therefore, by definition, lim = = .
n  b
n lim bn  2
n 

an lim an 
Corollary If If lim an = l1 and lim bn = l2  0, then lim = n = 1
n  n  n  b
n lim bn  2
n 

Proof: By theorems 4.2 and 4.3, we have lim

an
n  b
n
1 
 1
 1 
= lim  an    = lim an  lim  = 1  = 1 .
n 
 bn  n  n bn  2 2

Examples and exercises 4

Example 4.1 Evaluate each of the following, using the theorems on limits:
5 5
3 3  lim
3n  5n
2
n = n  n 30 3
(a) lim 2 = lim = =
n  5n  2n  6 n  2 6 2 6 500 5
5   2 5  lim  lim 2
n n n  n n n

(b) lim 
 n  n  2 n3 
 2  = lim 
 
 n  n  2  n 2  1  n3  n  1 
 = lim 
 n 3  n 2  2n 

n 
 n 1 n  1  n 
  n  1 n 2

 1  
 n   n  1 n 2  1  
1 2 1 2
1  2 1  lim  lim 2
= lim n n =
n  n n  n =1
n   1  1   1  1 
1  1  2  1  lim  1  lim 2 
 n  n   n n  n n 
4 4
 3  3
4
 2    2  lim  4
 2n  3  n n  n  2  16
(c) lim   = lim   =  =  =
n  3n  7
  n  3  7   3  lim 7   3  81
 n  n  n 

2 4 2 4
 5 lim 2  lim 5
2n  4 n
5 2 2 0
= lim n n =
n  n n n
(d) lim 7 = =0
n  3n  n  10
3
1 10 1 10
3  4  7 3  lim 4  lim 7 3
n 

n n n  n n  n
1 1
 2 lim n  2
1  2 10n 10 n n  10 2
(e) lim = lim = =
n 5  3  10 n n  5 5
n
 3 lim n  3 3
10 n  10

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 8
Limit of sequence notes By Mr. Francis Hung
4. Theorems on limits
3n 2  4n
Example 4.2 Find lim , using theorems on limits.
n  2n  1

4
3
3n 2  4n n
lim = lim n 
n  2n  1 n  1
2
n
4
3
 lim n =  and lim n =3
n  n  1 2
2
n
3n  4n
2
 lim =
n  2n  1

a n m  a n m 1    am
Example 4.3 Find lim 0 m 1 m1 , where ai and bj are constants and m is a positive
n  b n  b n
0 1    bm
integer, Also b0  0.
a a
a0  1    mm
a0 n m  a1n m 1    am n n = a0
lim = lim
n  b n  b n
0
m
1
m 1
   bm n b  b1    bm b0
0
n nm
bn  c n
Example 4.4 If a > b > c > 0, prove that lim n n = 0.
n  a  c
n n
b c
b c
n n     00
a a
lim n n = lim  n  n = = 0.
n  a  c n 
 a   c  1 0
   
a a
Example 4.5 Find the limit: lim n
n 
 n2  1  n . 
lim n
n 
 
n 2  1  n = lim n
n 
 n2  1  n   n2  1  n
n 1  n
2
= lim n 
n 
n2  1  n2
n 1  n
2
= lim
n 
n
n 1  n
2

1 1
= lim =
n  1 2
1 2 1
n
Example 4.6 Eavluate each of the following limits.
1
(a) lim m (where m is a positive integer.)
n  n

3n3  n2  n
(b) lim .
n  5n 3  1
k n   k  1
n

(c) lim n1 (where k is a positive number.)

k   k  1
n  n 1

(d) lim
n 
 n 1  n . 
(e) lim  3
n 1  3 n.
n 

3 1
Ans. (a) 0; (b) ; (c) ; (d) 0; (e) 0.
5 k 1

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 9
Limit of sequence notes Edited by Mr. Francis Hung
4. Theorems on limits
5. Squeezing principle (p.42)
n
5.1 Find lim n .
n 2

nn  1 nn  1
2n = (1 + 1)n = 1 + n + +  > for n  2
2 2
1 2
0< < for n  2
2 n
nn  1
n 2
0< < for n  2
2 n
n  1
n 2
lim 0  lim  lim =0
n  n 2 n n n  1

n
By squeezing principle, lim =0
n 2 n

n100
4.2 Find lim .
n 1.01n

nn  1 n  100 

1.01n = (1 + 0.01)n = 1 + 0.01n +  + 0.01101+ 
101!
nn  1 n  100 
> 0.01101 for n  101.
101!
1 101!
0< < for n  101.
1.01 nn  1 n  100   0.01
n 101

n100 n100
0< < 101! for n  101.
1.01n nn  1 n  100  0.01101
n100 1
lim 0  lim  101! lim =0
n  n 1.01 n n  1  2   100 
n1  1  1    0.01
101

 n  n   n 
n100
By squeezing principle, lim =0
n 1.01n

n!
4.3 Find lim n .
n n

n! 1 2 3 n 1
0< =    <
nn n n n n n
n! 1
lim 0  lim n
 lim = 0
n  n n n  n

n!
By squeezing principle, lim =0
n nn

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 10
Limit of sequence notes By Mr. Francis Hung
4. Theorems on limits
an
4.4 Prove that lim = 0 for any real number a .
n n!

0n
(a) For a = 0, lim = 0 is obviously true.
n n!

(b) For a > 0, there exists k N such that a  k, it follows that

a a a
1 > > >
k k 1 k  2
When n > k, we have
an  a a a  a a a
0< =         
n!  1 2 k  k  1 k  2 n 
n k k n
ak  a  ak  k 1  a 
<   =    
k!  k 1  k!  a   k  1 

k  1k   a 
n
=  
k!  k 1

Because k is a constant, so as
k  1k . As 0 <
a
< 1, so that,
k! k 1

k  1 
k

a  k  1
n k
 lim 
 a 
n
lim   =  =0
n  k!  k 1 k! n   k  1 

an
and hence, by squeezing principle, lim = 0.
n n!

bn
(c) For a < 0, let b = –a (where b > 0), by the above result, lim = 0.
n n!

bn an bn
–  
n! n! n!
bn an bn
0 = – lim  lim  lim =0
n n! n n! n n!

an
By squeezing principle, lim = 0.
n n!

an
We conclude that for all real number a, lim = 0.
n n!

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 11
Limit of sequence notes By Mr. Francis Hung
4. Theorems on limits
4.5 (a) Evaluate the limits:
1
(i) lim 2 n ;
n
1
(ii) lim n n .
n 

(b) Let 0  r  n, prove that C rn11t n r 1  C rn t n 1 r for t  1, where C rn are binomial
coefficients.
Hence, or otherwise, prove by induction that t n  1n  t n 1  1n 1 for t  1.
1 1

Let x, y > 0, an = x n  y n n , prove that

1

(c)
(i) {an} is strictly decreasing;
(ii) lim a n  maxx, y, by using squeezing principle.
n 

(d) Suppose xn  0 is a monotonic increasing sequence tends to a.

 
1

Prove that lim x1n  x2n    xnn n = a.

n 

(a) (i)
1
1 1
Clearly 2 n > 1, otherwise 2 n  1  2 = 2 n  1n = 1  2  1 !!!   1 n

Let 2 n = 1+ hn, where hn > 0

2 = ( 2 n )n = (1 + hn)n = 1 + nhn +  > nhn
1

2
0 < hn <
n
2
lim 0  lim hn  lim =0
n  n n n
By squeezing principle, lim hn = 0  lim 2 n = lim1  hn  = 1
1

n n n

(ii) Clearly n
1
n

1
1
> 1, otherwise n  1  n = n
n
   1 = 1  n  1 !!!
1
n
n
n

Let lim n n = 1+ kn, where kn > 0

n 
1 nn  1 2 nn  1 2
n = ( n n )n = (1 + kn)n = 1 + nkn + kn +  > k n for n  2
2 2
2
0 < kn < for n  2
n 1
2
lim 0  lim k n  lim =0
n  n n  n 1
By squeezing principle, lim k n = 0  lim n n = lim1  k n  = 1
1

n n n

(b)
n  1nn  1n  r  1t ; C nt n 1 r = nn  1 n  r  1t n1r
Crn11t n r 1 =
n  r 1

r  1! r
r!
n1 n  r 1
Cr 1 t
=
n  1t n  r 1
n  1 t nr  1 for t  1 and 0  r  n.
 n1r =
n  n1 r
Cr t r  1t r  1
 C rn11t n r 1  C rn t n 1 r for t  1 and 0  r  n.
(t + 1)2 = t2 + 2t + 1 > t2 + 1  t + 1 > t 2  12 . It is true for n = 1.
1

n 1 n 1 n n
(tn + 1)n+1 =  Crn1t nr = 1   Crn 1t nr = 1   Crn11t n r 1 >  Crnt n 1r = (tn+1 + 1)n
r 0 r 1 r 0 r 0

 t n  1  t n 1  1
1 1
n n 1
for t  1

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 12
Limit of sequence notes By Mr. Francis Hung
4. Theorems on limits
1

 x  n  n
x
  = y    1 = y t n  1n
1 1

(c) (i) Let t = ; an = x n  y n n

y  y  

   
1 1

an 1 y t n1  1 n1 t n1  1 n1

= = < 1 by (b)
   
1 1
an y tn 1 n tn 1 n
an+1 < an {an} is strictly decreasing
W.L.O.G. let x  y, y = 0  y n n  x n  y n n   y n  y n n = 2 n y
1 1 1 1
(ii)
 
1
lim 2 n y = y by (a)(i)
n

By squeezing principle, lim x n  y n n = y

1

n

Similarly if y  x, lim x n  y n n = x
1

n

 lim a n  maxx, y
n 
1
   x   
1 1 1

(d) xn = x n n
n
n
1  x2n    xnn n  xnn  xnn    xnn n
 n n xn
1
a = lim xn  lim x  x    x   lim n
1
n n n
1 2 n
n n
xn = 1(a) = a by (a)(ii)
n n n

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 13
Limit of sequence notes By Mr. Francis Hung
5. Monotonic convergent theorem
Example 5.1 (Example of monotonic convergent theorem)
Find the limit of the sequence

2, 2 2 , 2 2 2 ,  .

Can you prove that the limit exists?

Solution:

Let x1 = 2 , x2 = 2  2 ,  , xn = 2  2  2  ... 2 (there are n ’s and n 2’s inside.)

Then x2 = 2  x1 , x3 = 2  x2 ,  , xn = 2  x n 1 .

First, we try to prove that {xn} is a monotonic increasing sequence.

2 2  2
x2 – x1 = 2  2  2 =  2  2  2  
  2 2  2

 
2
 2  2   2 2

2 2 2 2
=    0
2 2  2 2 2  2 2 2  2
 x2 > x1
Suppose xk  xk–1.

xk 1  xk  2  xk  2  xk 1 =  
2  xk  2  xk 1 
2  xk  2  xk 1
2  xk  2  xk 1
2  xk  2  xk 1 xk  xk 1
= = > 0 by induction assumption
2  xk  2  xk 1 2  xk  2  xk 1
xk+1 > xk for k  1
By the principal of mathematical induction, {xn} is a monotonic increasing sequence.
 {xn} is a monotonic increasing sequence.

Claim: xn < 2 for all integer n.

Proof: By mathematical induction.
n = 1, x1 = 2 < 2
It is true for n = 1.
Suppose xk < 2 for k > 1

Then xk+1 – 2 =  2  xk  2   2  xk  2
2  xk  2
xk  2
= <0
2  xk  2
 xk+1 < 2
So the statement is also true for n = k + 1
By the principle of mathematical induction, xn < 2 for all integer n.
Since {xn} is an monotonic increasing sequence and is bounded above by 2, by the monotonic
convergent theorem, lim xn exists.
n

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 14
Limit of sequence notes By Mr. Francis Hung
5. Monotonic convergent theorem
Let lim xn = x.
n

Then x2 = 2  2  2  
x2 = 2 + x
x2 – x – 2 = 0
(x – 2)(x + 1) = 0
x = 2 or x = –1

As x = 2  2  2   > 0
So x = –1 is rejected
Therefore, x = 2 only.

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 15
Limit of sequence notes By Mr. Francis Hung
6. An important limit: the number e
6 An important limit: the number e
n
 1
Theorem 6.1 The sequence {an} where an = 1   is monotonic increasing and bounded above by
 n
3, and hence is convergent.
Proof: By the binomial theorem, for any positive integer n,
n n r n r
 1 n 1  n 1 
1   =  C r   = 1   C r  
 n  r 0  n  r 1 n
nn  1n  r  1  1 
n r
=1    
r 1 r! n
n
1  0  1  2   r  1 
=1    1  1  1  1  
r 1 r!  n  n  n   n 
n 
1 r 1  k 
= 1      1   
r 1  r! k  0  n 
n 
1 r 1  k 
< 1      1  
r 1  r! k  0  n  1 
n 1 
1 r 1  k 
< 1      1  
r 1  r! k  0  n  1 
n 1
 1 
= 1  
 n 1
 The sequence is monotonic increasing.
n n 
 1 1 r 1  k 
Next, 1   = 1      1  
 n r 1  r! k  0  n 
n
1 1 1 1 1
<1    1   
r 1 r! 1! 2! 3! n!
1 1 1
<1  1   
1 2 2  3 n  1  n
 1 1 1  1 1
= 1  1  1            
 2  2 3  n 1 n 
1
=1+1+1–
n
<3
This shows that {an} is bounded above by 3. By monotonic convergent theorem, {an} converges.

n
 1
The limit of this sequence is denoted by e, i.e. e  lim 1   .
n  n
n n 
 1 1 r 1  k  
1
Let n  , then 1    1      1    1  
 n r 1  r! k  0  n  r 1 r!
n
 1 1 1 1
this suggests e  lim 1   = 1         2.71828 
n  n 1! 2! k!

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 16
Limit of sequence notes By Mr. Francis Hung
6. An important limit: the number e
Corollary 6.2: For any rational number p,
pn
 1
(a) lim 1    e p
n   n
n
 p
(b) lim 1    e p
n  n
Proof:
(a) p = 0, the result is obvious.
p  ℕ. Induction on p.
p = 1, the result is obvious.
kn
 1
Suppose lim 1    e k for some k ℕ.
n   n
k 1n kn n kn n
 1  1  1  1  1
lim 1    lim 1   1    lim 1   lim 1    e k  e  e k 1
n  n n  n  n n    n n  n
pn
 1
By MI, lim 1    e p  p  ℕ.
n   n
If p < 0 and –p  ℕ, let q = –p > 0
pn  qn
 1  1 1 1
lim 1    lim 1    qn
 q
 ep
n   n n   n  1 e
lim 1  
n   n
1
If p = , where q ℕ.
q
1 1
Claim: If 1 < x < y, then 1  x q  y q .
1 1
Proof: By contradiction, if y q  x q
q q
 1q   1q 
y
   x 
   
y  x, which is a contradiction.
1 1
 1  x q  y q  (*)
n
 1
 1   is monotonic increasing.
 n
n n 1
 1  1 
 1    1  
 n   n 1
n n 1
 1 q  1  q
1    1   (by the result of (*))
 n   n 1
n
 1 q
 1   is monotonic increasing.
 n
n
 1
Moreover, 1   < 3
 n
n
 1 q 1

1   < 3 q (by the result of (*))

 n

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 17
Limit of sequence notes By Mr. Francis Hung
6. An important limit: the number e
n
 1 q 1

1   is bounded above by 3 q .
 n
n n
 1 q  1 q
By monotonic convergent theorem, lim 1   exists. Let lim 1   = l.
n   n n   n
q
 1  qn 
lim 1     e   q
n    n 
 
n
 1 q
 l = lim 1  
n   n
m
When p = , where q  0, m, q ℕ. and (m, q) = 1
q
m m
 1
pn
 1 q
mn
 1  qn    1 q 
n
m
 1 m
lim 1    lim 1    lim 1      lim 1      e q   e q  e p
n  n n   n n    n   n n    
 
m
When p = – , where q  0, m, q ℕ. and (m, q) = 1
q
pn  mn
 1  1 q
1 1  mq
lim 1    lim 1     e  ep
 1  
m
n  n n  n mn
q
e q
lim 1   
n   n 

The theorem is proved.
n
 p
(b) To prove lim 1    e p
n  n
p = 0, the result is obvious.
p ℕ, Induction on p.
n
 1
p = 1, lim1    e1 ; it is true for p = 1.
n 
 n
n
 k
Suppose lim1    e k , k ℕ
n 
 n
n n n
 k 1  1  k 
lim1    lim1   1  
n 
 n  n n   n  1 
n 1
 k 
n 1  
 1  n 1
 lim1   lim
n 
 n  n1 1  k 
 
 n 1
ep
=e = ep+1
1
 It is also true for p = k + 1
n
 p
By the principal of mathematical induction, lim1    e p for p ℕ
n 
 n

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 18
Limit of sequence notes By Mr. Francis Hung
6. An important limit: the number e
n n
 p  q
p < 0, let q = –p > 0, lim1    lim1  
n
 n n  
 n
1
= n
, valid for n > q
 n 
lim 
n  n  q 
 
1 1
= n q
 q
 q   q 
lim1   lim1  
n 
 nq n 
 nq
1
= q = ep
e
1
p = , where q ℕ. Let m = nq.
q
n m
n
 p  1   1 q m
lim 1    lim 1    lim 1    e q  e p (by the result of (a))
n   n n 
 nq  m   m
m
When p = , where q  0, mℤ, q ℕ. and (m, q) = 1
q
n k
n
 p  m  m q
lim 1    lim 1    lim 1   , where k = nqℕ.
n   n n 
 nq  k   k
1
  m 
k q
 
1
=  lim 1     e m q
 e p (by the above result and the result of (a))
k  k  
The theorem is proved.

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 19
Limit of sequence notes By Mr. Francis Hung
6. An important limit: the number e
Example 6.1 Advanced Level Pure Mathematics Calculus and Analytical Geometry I
by K.S. Ng, Y. K.kwok, p.90 Exercise 2D Q2(b)
n
 1  1 1 1
(i) lim 1   = n
= 3 n 1 1
= 3 n 1 1
n 3n   3n  
lim    1  3 3  1  3  1 3
n 3n  1  lim 1   lim 1   lim 1  
n   3n  1  n   3n  1  n   3n  1 
1
= m
; m = 3n – 1
 1 3
lim 1    1
m   m
1  1
= 1 e 3
e3
m m 1
 1  1 
Remark 1    1  
 m  m 1
m m1
 1 3  1  3
1    1  
 m   m 1
 It is a monotonic increasing sequence.
m
 1
Moreover, 1   < 3
 m
m
1 1
 1 3
 1    3 3  the sequence is bounded above by 3 3 .
 m
m
 1 3
By monotonic convergent theorem, lim 1   = l
m m
3
 m
m
m
1
 lim 1  1     3  e = lim 1  1    3  lim 1  1  3 = e 3
3
m m   m m m m
 
2 n 7 3
2 n 2n n2 
 4   2n  7   2n  3   4  2 2
(ii) lim 1   = lim   = lim   = lim 1  
n 2n  3  n 2 n  3  n    2n  7  n 2n  7 
2 n 7 3 m
 4  2  4 2  4 2
= lim 1    lim 1   = lim 1    1 ; m = 2n – 7
n 2n  7  n   2n  7  m m 
m
 1  1  1  1  2
= lim 1  1  1  1  
m  m  m  1  m  2  m  3 
m m m m
 1 2  1 2  1 2  1 2
= lim 1   lim 1   lim 1   lim 1  
m m  m m  1  m m  2  m m  3 
1 1 1 1
= e 2  e 2  e 2  e 2 = e2
n n n n
 2 8   4  2   4  2
(iii) lim 1   2  = lim 1  1   = lim 1   lim 1   = e4e–2 = e2
n n n  n  n  n  n n  n n
n n n n
 3 5   5  1   5   1  5 1 3
(iv) lim 1   2  = lim 1  1   = lim 1   lim 1   = e 4  e 2  e 4
n   4n 8n  n  4n  2n  n 4n  n 2n 

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 20
Limit of sequence notes By Mr. Francis Hung
6. An important limit: the number e
Example 6.2
n 2 1 n 2 1 n2 1
 n2 1  1   1   1 
lim  2   lim 1  2   lim 1  2  lim 1  2   e
n   n  n   n  n   n 
n
 n 
Example 6.3
nn  1 n  k  1      
k nk
Let  be fixed real number and k is a positive integer, find lim   1   .
n k! n  n
n
 
1  
nn  1 n  k  1      
k nk
1 nn  1n  k  1 k  n
lim   1   = lim  k
  k
n k! n  n n k! n  
1  
 n
1 n n 1 n  2 n  k  1 k e 
= lim      
n   k! n n n n 1
k  1  k 
= lim 1  1    1    1 
1 1 2
 e
n   k!  n  n  n 
1
= k e  
k!
n

Remark: lim 1    e  ℝ.
n   n
Proof: ℝ, we can find a sequence of rational numbers {qm} so that lim qm   .
m 
n
 q  lim q m
lim lim 1  m   lim e q m  e n  e  (We have assumed that f (x) = ex is continuous.)
m   n   n  m

Example 6.4
(a) Prove the following inequalities:
n
 n 1
(i) n! <   for n > 1
 2 
n n
n n
(ii)    n! e 
e 2
 1  n 
where e is the limit of the sequence 1    .
 n  
n!
(b) Using (a)(ii) to prove that lim n  0 .
n  n

(a) (i) By the inequality of the means, we have

1
1 n  n  1
2
1 1
2  n  1  2  n  1  n  1
2 2
1 1
3  n  2  3  n  2  n  1
2 2

1
n 1  n  1
2
n
n  1
Multiplying, n! <   .
 2 

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 21
Limit of sequence notes By Mr. Francis Hung
6. An important limit: the number e
n n n n
n 1  n   1 
(ii) By (i), n! <  n
    1    e  .
 2  2  n 2
 1 
n
As 1    is an increasing sequence with e as its limit, so that
n 
 
n
 1
1    e for n = 1, 2, 
 n
2 3 n
 1  1
 1
 , 1    e
1
 1    e , 1    e ,  1    e ,
 1   2   3  n
2 3 n
Multiplying 1  1   1   1    e n
1 1 1 1
 1  2  3  n
2 3 n
 2  3   4   n 1
         en
 1  2   3   n 
n  1n  en
n!
n n  1n  n! e n 
n n
    
e en 2
1 n e
(b) By (a)(ii),n
 n  n! n  n n
e 2
1 n! e
 
en nn 2n
1 e
As n  ,  0, 0
en 2n
n!
By squeezing principle, lim 0.
n  nn
Example 6.5
n
 1
Prove that xn = 1   is monotonic increasing and bounded above. While the sequence
 n
n1
 1
yn = 1   is monotonic decreasing and bounded below.
 n
Hence show that they have the same limit: lim xn  lim yn  e .
n n

yn
[Hint: Consider and use Bernoulli inequality: (1 + x)n  1 + nx for x > –1.]
yn1
n1
 1 
1  
 nn  2  
n n
xn1  n  1  1   1   1 
  2  
1   1  2 
1 
xn  1
n
 n  1   n  1   n  1   n  1 
1  
 n
 n  1   n 2  n  1 n  2  n3  3n 2  3n  2
 1  2  1  = 2   3 1
 n  1  n  1   n  1  n  1  n  3n  3n  1
2

 xn+1 > xn
The fact that xn is bounded above by 3 has already been proved.
 lim xn exists.
n

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 22
Limit of sequence notes By Mr. Francis Hung
6. An important limit: the number e
n 1
 1
1   n1
 n  12   n  1  
n 1
yn  n 1   n 1 
=      1  
yn1  1 
n 2
 nn  2  n  2   nn  2   n  2 
1  
 n 1
 n  1  n  1   n 2  3n  1 n  1 n3  4n 2  4n  1
 1      3 1
 nn  2 n  2   nn  2   n  2 n  4n 2  4n
 yn > yn+1, so it is monotonic decreasing.
Clearly yn is bounded below by 0
 lim yn exists.
n

Let lim xn = p, lim yn = q

n n
n 1 n
 1  1  1
q = lim1    lim1    lim1    lim xn 1  lim xn  p
n
 n n
 n n 
 n  n n

Example 6.6
n n1
 1  1
(a) Let an = 1   , bn = 1   , n = 1, 2, 
 n  n
Without reference to binomial theorem, show that {an} is monotonic increasing and {bn} is
monotonic decreasing.
Hence, determine which is larger number (1000000)1000000 or (1000001)999999.
n n
n n
(b) From the results of (a) show that    n! en  1  .
e e
n
n
For n > 6, derive the sharper inequality n! n  .
e
x
 1
(a) Let y = 1   .
 x
 1
ln y = x ln 1  
 x
x
 1   1 x  1 
y '  1   ln1       
 x    x  1  1x  x 2 
x
 1   1 1 
= 1   ln1    
 x    x  1 x 
 1 1
Let z = ln1   
 x  1 x
1  1  1
z’ =  2  
1  x  x  1  x 2
1

1 1 1  x  x 1
=  =  <0
x1  x  1  x 2
x1  x 2
x1  x 2
 z is strictly decreasing
  1 1 
 x > 0, z(x) > lim z  x  = lim ln1    = ln 1 – 0 = 0
x  x    x  1  x 
z>0
y’ > 0

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 23
Limit of sequence notes By Mr. Francis Hung
6. An important limit: the number e
y is increasing
n
 1
an = 1   is monotonic increasing.
 n
x 1
 1
Let y = 1   .
 x
 1
ln y = (x + 1) ln 1  
 x
x 1
 1    1  x  1  1 
y '  1   ln1       2 
 x    x  1  1x  x 
x 1
 1    1  1
= 1   ln1    
 x    x  x
 1 1
Let z = ln1   
 x x
1  1 1
z’ =   
1  1x  x 2  x 2
1 1  x 1 x 1
=  2 = 2  >0
x1  x  x x 1  x  x1  x 2
 z is strictly increasing
  1  1
 x > 0, z(x) < lim z  x  = limln1     = ln 1 – 0 = 0
x  x
  x  x
z<0
y’ < 0
y is decreasing
n 1
 1
bn = 1   is monotonic decreasing.
 n
n
 1
(n + 1) > e > 1   it is true for n > 1
 n
n
 1  1
n 1     1  
 n  n
n 1
 1
n  1  
 n
n 1
 n 1
n 
 n 
n n  n  1
n 1

Put n = 1000000, then (1000000)1000000 > (1000001)999999.

n
n
(b) The fact that    n! has been proved in Example 14 (a)(ii).
e
 bn  to e
2 3 4 n 1
 1  1  1  1
 e < 1   , e  1   , e  1   ,  , e  1  
 1  2  3  n

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 24
Limit of sequence notes By Mr. Francis Hung
6. An important limit: the number e
2 3 4 n 1
 1  1  1  1
Multiplying e n  1   1   1    1  
 1  2  3  n 
2 3 4 n 1
 2   3   4   n 1
e        
n

1 2 3  n 

e 
n n  1
n 1

n!
1 n!
 n 
e n  1n1
n n n n
n n!n n  n   1  n 
en  1   en  1  e  n! 1     n! n!
e n  1n 1
 n 1  n   n 1
n n
n n
    n! en  1 
e e
n
n
To prove that for n > 6, n! n  .
e
Induction on n.
7
7
When n = 7, L.H.S. = 7! = 5040, R.H.S. = 7  = 5257
e
 L.H.S. < R.H.S., it is true for n = 7
k
k
Suppose k! k   for some positive integer k > 6.
e
 bn  to e
k 1
 1
 e < 1  
 k
e k < (k + 1)k+1
k+1

k k 1 k  1
k 2
(k + 1)! = (k + 1)k! < k  1  k  ; by MI, the statement is true for n > 6.
e e k 1

C:\Users\85290\Dropbox\Data\MathsData\Pure_Maths\Sequence&Series\notes\limit_of_sequence_note.docx Page 25