Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.

com
Paper I (b) Differentiate the result in (a) w.r.t. x,
n
1 1 1  1 1 1 
1. (1 + 2 x) n = ∑ ak x k --- <*>  −  +  − 
6  ( x + 1) 2 ( x + 2) 2  6  ( x − 1) 2 ( x − 2) 2 
k =0

(a) Put x = 1 , we have

n
∑ ak = (1 + 2 ⋅1) n = 3 n =
[
( x 2 − 1)( x 2 − 4) − x 2 x( x 2 − 1) + 2 x( x 2 − 4) ]
2 2 2 2
k =0 ( x − 1) ( x − 4)
n
x 4 − 5 x 2 + 4 − 2 x 2 (2 x 2 − 5)
(b) Differentiate <*> w.r.t. x, 2n(1 + 2 x) n −1 = ∑ ka k x k −1 =
k =1 ( x 2 − 1) 2 ( x 2 − 4) 2
n
− 3x 4 + 5 x 2 + 4
Put x = 1 , we have ∑ ka k = 2n(1 + 2 ⋅1) n −1 = 2n3 n −1 =
k =1 ( x 2 − 1) 2 ( x 2 − 4) 2
n n n
3x 4 − 5x 2 − 4
(c) ∑ (3k + 1)a k = 3 ∑ ka k + ∑ a k i.e.
k =0 k =0 k =0 ( x 2 − 1) 2 ( x 2 − 4) 2
n n
1 1 1  1 1 1 
= 3∑ ka k + ∑ a k = 2n3 n + 3 n = (2n + 1)3 n =−  −  −  − 
k =1 k =0 6  ( x + 1) 2 ( x + 2) 2  6  ( x − 1) 2 ( x − 2) 2 
n
3k 4 − 5k 2 − 4
(c) ∑ 2
x x k =3 ( k − 1) 2 (k 2 − 4) 2
2. (a) Consider =
( x 2 − 1)( x 2 − 4) ( x + 1)( x − 1)( x + 2)( x − 2) 1 n  1 1  1 n  1 1 
=− ∑  −  − ∑  − 
6 k =3  (k + 1) 2 (k + 2) 2  6 k =3  (k − 1) 2 (k − 2) 2 
A B C D
= + + +
x +1 x −1 x + 2 x − 2 1 1 1  1 1 1 
=−  2 −  − − + 
Then, x = A( x − 1)( x + 2)( x − 2) + B( x + 1)( x + 2)( x − 2) 6  4 ( n + 2) 2  6  12 (n − 1) 2
 


+C ( x − 1)( x + 1)( x − 2) + D( x − 1)( x + 1)( x + 2) 1 1 1
→ − × 2 + ×1 as n → ∞
Put x = −1 , 1, −2 and 2, 6 4 6

−1 −1 1 1 5
we have A = , B= , C= and D = =
6 6 6 6 32

x 1 1 1 1 1 1
Hence, =− ( − )− ( − )
( x 2 − 1)( x 2 − 4) 6 x +1 x + 2 6 x −1 x − 2

Page 1
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
3. (a) f ( x) = q1 ( x)( x − 1)( x − 4) + (− x + k ) and f ( x) = q 2 ( x)( x − 4) 2 + (kx − 10)  π π
 cos − sin 
Put x = 4 , −4 + k = f (4) = 4k − 10 , 4. (a) Required matrix =  2 2  =  0 − 1
 sin π π  
 cos   1 0 
i.e. k=2  2 2 
2
(b) Let g ( x) = ax + bx + c . We have  x   0 − 11   − 3 
(b) (i) Let Q = ( x, y ) .   =    =  
3 2
x + ax + bx + c = f ( x) = q1 ( x )( x − 1)( x − 4) + (− x + 2) --- <*>  y   1 0  3  1 
3 2 2 i.e. Q = (−3, 1)
x + ax + bx + c = f ( x) = q 2 ( x)( x − 4) + (2 x − 10) --- <**>
* Put x = 1 into <*>: a+b+c = 0 --- <1>  0 − 1
(ii) (1) M =  
* Put x = 4 into <**>: 16a + 4b + c = −66 --- <2>  −1 0 
* Differentiate <**>,  3π 3π 
 0 − 1  cos 2( 4 ) sin 2( 4 ) 
2 2
3x + 2ax + b = 2( x − 4)q 2 ( x) + ( x − 4) q ' 2 ( x) + 2 (2) Note M =   =  
 − 1 0   sin 2( 3π ) − cos 2( 3π ) 
Put x = 4 , 8a + b = −46 --- <3>  4 4 
3π
∴ T represents the reflection in the line y = x(tan ) = −x .
Now, < 2 > − < 1 >: 5a + b = −22 4
And <3>: 8a + b = −46 (3) Since reflection is area invariant, so
So, a = −8 , b = 18 . Sub, into <1>, c = −10 area of ∆O ' P' Q ' = area of ∆OPQ
2
Thus, g ( x) = −8 x + 18 x − 10 1
= (OP)(OQ )
Alternatively 2
Q deg f ( x ) = 3 and leading coefficient of f (x) is 1 1
2
= ( 3 2 + 12 )( 3 2 + 12 ) = 5
∴ f ( x) = ( x + A)( x − 4) + (2 x − 10) , and recall that 2
f ( x) = q1 ( x)( x − 1)( x − 4) + (− x + 2)
∴ (1 + A)(1 − 4) 2 + (2 × 1 − 10) = f (1) = q1 (1)(0)(1 − 4) + (−1 + 2)
∴ A=0, then f ( x) = x( x − 4) 2 + (2 x − 10)
= x 3 − 8 x 2 + 18 x − 10
i.e. g ( x) = f ( x) − x 3 = − 8 x 2 + 18 x − 10

(c) Remainder = [ g (−1)]3 = [−8 − 18 − 10] 3 = = −46656

Page 2
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
5. (a) Consider | z − a |= r where a ∈ C and r > 0 . 1 1 1 1
6. (a) Apply AM ≥ GM on 2 sets of numbers { α, β, γ, δ } and { , , , },
Then, 2 2
r =| z − a | = ( z − a)( z − a ) = zz − az − a z + | a | 2 α β γ δ

zz = az + a z − (| a | 2 −r 2 ) 1 1 1 1 1 1/ 4
we have α + β + γ + δ ≥ 4(αβγδ)1 / 4 and + + + ≥ 4( ) .
Compare the given equation , we have α β γ δ αβγδ

The product of the above 2 inequalities gives the required result.

a = 12 − 16i and | a | 2 −r 2 = 375 , r = (12 2 + 16 2 ) − 375 = 5

So, S represents a circle with centre G at 12 − 16i , radius equals 5. Alternatively

(b) z1 should represent a the complex number of a point (say Q) lies on the Apply Cauchy-Schwartz’ s inequality on 2 sets of numbers
circle such that it is collinear with the origin and the centre of the circle. 1 1 1 1
{ α , β , γ , δ } and { , , , }.
α β γ δ
Note OQ = 12 2 + 16 2 − 5 = 15 and OG = 12 2 + 16 2 = 20

OQ (b) Using the result in (a) by renaming α , β , γ and δ by

∴ z1 = (12 − 16i ) = 9 − 12i
OG β + γ + δ , α + γ + δ , α + β + δ and α + β + γ respectively.
Then,

[3(α + β + γ + δ)] 1
+
1
+
1
+
1 
 ≥ 16
β+ γ +δ α + γ +δ α+β+δ α+β+ γ 
i.e. result follows.
(c) By (b), we have
16 α + β + γ + δ α + β + γ + δ α + β + γ + δ α + β + γ + δ
≤ + + +
3 β+ γ +δ α+γ+δ α+β+δ α+β+ γ

α β γ δ
= (1 + ) + (1 + ) + (1 + ) + (1 + )
β+γ+δ α+γ+δ α+β+δ α +β+ γ

α β γ δ 16 4
∴ + + + ≥ −4 =
β+ γ +δ α+ γ +δ α+β+δ α+β+ γ 3 3

Page 3
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
7. (a) (i) Coefficient determinant D x + y + z = 2 − −− <1>

1 1 1 (ii) (E): x − 2z = 1 − −− < 2 >
3 x + 4 y + 6 z = b − −− < 3 >
= a 0 − 4 = −12 + 4a − a(a + 4) + 16 = −(a + 2)(a − 2) 
3 4 a+4 On solving <1> and <2>, we have x = 1+ 2 z and y = 1− 3 z
(E) has a unique solution For consistent, the above solution must satisfy <3>,
iff D≠0 ∴ b = 3(1 + 2 z ) + 4(1 − 3z ) + 6 z = 7
iff a ≠ 2 and a ≠ −2 Solution set = {( 1+ 2t , 1− 3t , t): t ∈ ℜ }
iff a < −2 or −2 < a < 2 or a > 2 (b) Solve the first 2 equations of (F): x = −1− 2 z and y = 3 + z
Now For consistent, the above solution must satisfy the third and fourth equation:

2 1 1 ∴ λ = 3(−1 − 2 z ) + 4(3 + z ) + 2 z = 9
D x = 2 0 − 4 = 32 − 2a − 4b and µ = 7(−1 − 2 z ) + 17(3 + z ) − 3z = 44
b 4 a+4 (c) Note the first equations of (G) equals the system (E)

1 2 1 2
by setting a = and b = 5 .
D y = a 2 − 4 = ab + 4b − 2a 2 − 6a − 22 3
3 b a+4 By (a), x = 3 , y = −1 , z = 0 which is obviously not satisfying the last

1 1 2 equation of (G).
D z = a 0 2 = 8a − ab − 2 i.e. (G) is inconsistent.
3 4 b

Dx 2a + 4b − 32 Dy 2a 2 + 6a + 22 − ab − 4b
x= = , y= =
D (a + 2)(a − 2) D (a + 2)(a − 2)

Dz ab + 2 − 8a
z= ==
D (a + 2)(a − 2)

Page 4
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
8. (a) ‘ ⇒ ’ If λ is a repeated root of p( x) = 0 , then 1 1
Consider f ( ) = 6 f (µ ) = 0 (by <1>)
2
p( x) = ( x − λ ) q1 ( x) for some polynomial q1 ( x) . µ µ
Therefore, p' ( x) = ( x − λ)[2q1 ( x) + ( x − α)q1 ' ( x)] . 1 1 1 1 1 1
f ' ( ) = 6 5 + 5a 4 + 4b 3 + 3c 2 + 2b + a
So, p (λ ) = p ' (λ ) = 0 . µ µ µ µ µ µ

‘ ⇐ ’ If p(λ) = p' (λ) = 0 , 1

= (aµ 5 + 2bµ 4 + 3cµ 3 + 4bµ 2 + 5aµ + 6)
then p( x) = ( x − λ)q 2 ( x ) for some polynomial q 2 ( x) . µ5
So, p' ( x) = q 2 ( x) + ( x − λ)q 2 ' ( x) = 0 (by <*>)
Q p ' (λ ) = 0 1
i.e. is a repeated root of f ( x) = 0 . (by (a))
∴ q 2 (λ ) = 0 µ
Therefore, q 2 ( x) = ( x − λ)q 3 ( x) for some polynomial q 3 ( x) . (c) (i) Let α be the repeated root of g ( x ) = 0 .
Thus, p( x) = ( x − λ) 2 q 3 ( x) .
0 = g (α) = 4α 6 − 16α 5 + 17α 4 − 7α 3 + 17α 2 − 16α + 4
Hence, λ is a repeated root of p( x) = 0 . By (a), 
0 = g ' (α) = 24α 5 − 80α 4 + 68α 3 − 21α 2 + 34α − 16
(b) (i) Suppose the converse that the root µ = 0 , then
f (µ) = f (0) = 1 ≠ 0 , which leads to a contradiction. By trial and error, α = 2 .
i.e. µ≠0 ∴ one of the repeated roots is 2.
6 5 4 3 2
(ii) f ( x) = x + ax + bx + cx + bx + ax + 1 1
5 4 3 2
(ii) By (b), is also a repeated root of g ( x ) = 0 .
f ' ( x) = 6 x + 5ax + 4bx + 3cx + 2bx + a 2
It is given that µ is a repeated root of f ( x) = 0 , so we have ∴ ( x − 2) 2 (2 x − 1) 2 is a factor of g (x)
µ 6 + aµ 5 + bµ 4 + cµ 3 + bµ 2 + aµ + 1 = f (µ) = 0 --- <1> By noting that deg g ( x) = 6 and leading coefficient of g (x) is 4
5 4 3 2
6µ + 5aµ + 4bµ + 3cµ + 2bµ + a = f ' (µ) = 0 --- <2> we have, g (x) = ( x − 2) 2 (2 x − 1) 2 ( x 2 + bx + c)
Now 6× < 1 > −µ× < 2 > : Put x = 0 , 4 = g (0) = (−2) 2 (−1) 2 c , c =1
5 4 3 2 2 2
aµ + 2bµ + 3cµ + 4bµ + 5aµ + 6 = 0 --- <*> Put x = 1 , 3 = g (1) = (−1) (1) (1 + b + 1) , b =1
Thus, g (x) = ( x − 2) 2 (2 x − 1) 2 ( x 2 + x + 1)
for which ( x 2 + x + 1) cannot be factorized into linear
polynomials with real coefficients.
i.e. g (x) cannot be factorized in such a way too.

Page 5
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
9. (a) (i) n =1, x1 > y1 is trivial. 5 1 2 7
(v) 4 x n +1 + 3 y n +1 = 4( x n + y n ) + 3( x n + y n ) = 4 x n + 3 y n
Assume x k > y k for some positive integer k. 6 6 9 9

5 1 2 7 (vi) By (a)(v), 4xn + 3 yn = 4 x n −1 + 3 y n −1

Consider x k +1 − y k +1 = ( x k + y k ) − ( x k + y k )
6 6 9 9 = 4 x n −2 + 3 y n −2

11 =…
= (xk − y k ) > 0 by assumption.
18 = 4 x1 + 3 y1
The statement is also true for n = k + 1 . Taking limit on both sides,
By induction, x n > y n for all positive integers n.
lim (4 x n + 3 y n ) = lim (4 x1 + 3 y1 )
n →∞ n →∞
5 1 1
(ii) x n +1 − x n = ( x n + y n ) − x n = ( y n − x n ) < 0 (by (a)(i))
6 6 6
4 lim x n + 3 lim y n = 4 x1 + 3 y1
n →∞ n →∞
i.e. {x n } is strictly decreasing.

2 7 2 4 x1 + 3 y1
y n +1 − y n = ( x n + y n ) − y n = ( x n − y n ) > 0 (by (a)(i)) i.e. lim x n = (as lim x n = lim y n )
9 9 9 n →∞ 7 n →∞ n →∞

i.e. { y n } is strictly increasing. (b) Define X n = − x n and Yn = − y n for all positive integers n.
(iii) By (a)(i) and (ii), we have Then, (1) X 1 = − x1 > − y1 = Y1
x1 > x 2 > ... > x n −1 > x n > y n > y n −1 > ... > y 2 > y1 5 1 5 1
(2) X n +1 = − x n +1 = ( − x n ) + ( − y n ) = X n + Yn ,
∴ {x n } is strictly decreasing and bounded below by y1 , 6 6 6 6
{ y n } is strictly increasing and bounded above by x1 . 2 7 2 7
Yn +1 = − y n +1 = ( − x n ) + ( − y n ) = X n + Yn
i.e. Both {x n } and { y n } are convergent sequences. 9 9 9 9

5 1 By (a)(iii), { X n } and {Yn } are convergent sequences.

(iv) Consider lim x n +1 = lim ( x n + y n )
n →∞ n →∞ 6 6 i.e. {x n } and { y n } are convergent sequences.

5 1
∴ lim x n = lim x n + lim y n (Q lim x n , lim y n exist)
n →∞ 6 n → ∞ 6 n→ ∞ n →∞ n →∞

i.e. lim x n = lim y n

n →∞ n →∞

Page 6
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
10. (a) ‘ ⇒ ’ M is orthogonal.  cos θ sin θ 
‘ ⇐ ’ When M =   ,
MM T = I ,  − sin θ cos θ 
∴ 1 =| I |=| MM T |=| M | ⋅ | M T |=| M | 2 , | M |= ±1  cos θ sin θ  cos θ − sin θ 
MM T =    = I
a b   − sin θ cos θ  sin θ cos θ 
Let M =   .
c d   cos θ sin θ 
When M =   ,
Then, MM T = I gives M T = M −1  sin θ − cos θ 
a c  1  d − b  cos θ sin θ  cos θ sin θ 
∴   =   MM T =    = I
b d  | M |  − c a   sin θ − cos θ  sin θ − cos θ 
d −c
∴ a= and b =
|M | |M | (b) (i) The statement is trivial for n = 1 .
Case (1) | M |= 1 :  cos kθ sin kθ 
Assume M k =   for some positive integer k.
a = d , b = −c , ad − bc = 1  − sin kθ cos kθ 
∴ a2 + b2 = 1  cos θ sin θ  cos kθ sin kθ 
M k +1 = MM k =   
∴ −1 ≤ a, b ≤ 1 as M is a real matrix.  − sin θ cos θ  − sin kθ cos kθ 
∴ there exits θ ∈ ℜ such that  cos θ cos kθ − sin θ sin kθ sin kθ cos θ + cos kθ sin θ 
=  
a = cos θ and b = sin θ  − (sin kθ cos θ + cos kθ sin θ) cos θ cos kθ − sin θ sin kθ 
 a b   a b   cos θ sin θ   cos(k + 1)θ sin( k + 1)θ 
i.e. M =   =   =   =  
 c d   − b a   − sin θ cos θ   − sin( k + 1)θ cos(k + 1)θ 
Case (1) | M |= −1 : By induction, the statement is true for all positive integers n.
a = −d , b = c , ad − bc = −1  cos θ sin θ 
(ii) For M =   , note that M 2 = I
∴ 2
a +b =12
 sin θ − cos θ 
∴ 0 ≤ a, b ≤ 1 as M is a real matrix. M for n = 2k − 1
∴ Mn = , where k is a positive integer.
∴ there exits θ ∈ ℜ such that I for n = 2k

a = cos θ and b = sin θ

 a b   a b   cos θ sin θ 
i.e. M =   =   =  
 c d   b − a   sin θ − cos θ 

Page 7
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
 cos θ sin θ  11. (a) (i) Let f ( x) = x − 1 − ln x for x > 0 .
(c) It is not possible for X to be in the form of   , or else,
 sin θ − cos θ  < 0 for 0 < x < 1
1 x −1 
by (b)(ii), X 400
=X 2 ( 200 )
= I ≠ −I . f ' ( x) = 1 − = = 0 for x =1
x x 
> 0 for x >1
 cos θ sin θ 
So the orthogonal matrix X must be in the form of   .
 − sin θ cos θ  ∴ f ( x) attains its absolute minimum at x = 1 .

 cos 400θ sin 400θ  ∴ f ( x) ≥ f (1) = 1 − 1 − ln 1 = 0 for x > 0 .

Now − I = X 400 =  
 − sin 400θ cos 400θ  i.e. ln x ≤ x − 1 for all x > 0
cos 400θ = −1 (ii) ln bi ≤ bi − 1 (by (a) with bi > 0 )
 , 400θ = (2m + 1)π for m ∈ Z.
sin 400θ = 0 a i ln bi ≤ a i bi − a i (Q a i > 0 )

 cos θ sin θ  n n
Thus, X =   ∑ ai ln bi ≤ ∑ (a i bi − a i ) = 1 − 1 = 0
 − sin θ cos θ  i =1 i =1

 (2m + 1)π (2m + 1)π  n  n  n

 cos sin  ∑ ln bi a i
≤0, ln ∏ bi ai  ≤ 0 , i.e. ∏ bi a i
≤1.
= 400 400  for m ∈ Z. 
i =1  i =1  i =1
 − sin (2m + 1)π (2m + 1)π 
 cos 
 400 400  (b) (i) We are going to prove x x −1 ≥ 1 for x > 0 :
(d) We are going to show that MM T = I gives ( M n )( M n ) T = I for all (1) For x ≥ 1 , x −1 ≥ 0
positive integers n: so x x −1 ≥ 10 = 1
When n = 1 , the statement is trivial. (2) For 0 < x < 1 , 1− x > 0
k k T
Assume ( M )(M ) = I for some positive integer k. so x 1− x < 11− x = 1 , ∴ x x −1 > 1
Consider ( M k +1 )( M k +1 ) T = ( M )(M k )( M k ) T ( M k ) T Combine two cases, x x −1 ≥ 1 for x > 0 :
= ( M )( I )( M T ) (by assumption) i.e. xx ≥ x for x > 0
=I (by definition) n n
(ii) c i ci ≥ c i (by (b)(i) with c i > 0 ), then ∏ c i c ≥ ∏ ci
i
=1.
∴ by induction, the statement is correct for all positive integers n. i =1 i =1

i.e. M 401 is orthogonal matrix.

Page 8
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
n

xi
∑ xi n
(c) (i) Put a i = n
and bi = i =1
n
xi
xi n ∏ xi
∑ xi ∑ xi 2
i =1
(ii) Put c i = 1
such that c i > 0 and ∏ ci = 1
=1.
i =1 i =1 i =1 ×n
 n n  n n
 ∏ xi   ∏ xi 
a i , bi > 0 ,    
such that  i =1   i =1 
n Then, by (b)(ii),
n n xi
∑ xi 1
∑ ai = ∑ n
= i =1
n
= 1,
  n
xi
1   
xi   n  1n
i =1 i =1
∑ xi ∑ xi    ∏ xi  n  
 
   ∏ xi 
  i =1 
i =1 i =1    i =1  
n
 xi   n  xi  
1≤ ∏ 1  = ∏  1  
  n i =1    i =1   n  
  ∑ xi 2  n n n
n
 x n2
   ∏ x i      ∏ xi   
∑ a i bi = ∑  n i =
i =1
=1. 
  i =1 

    i =1  
 
i =1 
n  

 ∑ xi ∑ xi
i =1 2 2

 i =1  i =1 xi
 
  n
Then, by (a)(ii),  
n
 xi 
∏ xi x i

xi i =1
∴ 1≤ ∏  =
 n  ∑n xi 1 x1 + x2 +...+ xn
 ∑ xi  i =1 i =1 
 n n   n  n
n
 i =1    ∏ x i    ∏ xi 
 
∏ n xi  ≤1  
i =1 
  i =1    i =1 

 ∑ xi
2

 i =1  x1 + x2 +...+ xn
n 
 n 
1 i.e. x1 x2
x1 x 2 ...x n xn
≥  x1 x 2 ...x n 
 
 ∑ xi  n
∑ xi  
 i =1  n xi  i =1
∴  n  ∏ x i 
 ≤1
 ∑ xi 2  i =1 
 
 i =1  --- End of Solutions of Paper I ---
n
∑ xi
 n 2  i =1
 ∑ xi 
n
 
i.e. ∏ xi x i
≤  i =n1 
i =1  ∑ xi 
 
 i =1 

Page 9
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
Paper II 3. (a) (i)
y
0 0
( ) ( )
x − sin x 0 1 − cos x 0
sin x 2
1. (a) lim = lim = lim =0
x→0 x sin x x →0 x cos x + sin x x →0 2 cos x − x sin x
y = f (x )
(b) f (x) is differentiate at x = 0 and hence continuous at x = 0
x
∴ lim− f ( x) = f (0)
x →0
O 1 2
x
lim (sin x + a ⋅ ) = 3 + b (0) + 0 2
x→0 − sin x
0 + a = 3 , i.e. a=3
For f (x) is differentiate at x = 0 , (ii) Q f (−3) = 2 = f (0) but −3 ≠ 0

f ( x ) − f (0) f ( x ) − f ( 0) ∴ f (x) is not injective.

lim− = lim+
x→0 x−0 x → 0 x−0 (b) (i) g (− x) = f ((− x) − 1) + f (−(− x) − 1)
= f ( x − 1) + f (− x − 1) = g ( x) for all x ∈ ℜ
3x
(sin x + )−3
sin x (3 + bx + x 2 ) − 3 ∴ g (x) is even.
lim = lim+
x→0 − x x →0 x

 sin x x − sin x 
lim−  + 3×  = lim (b + x)
x→0  x x sin x  x →0 +

1 + 3× 0 = b , b =1

2. (a) f (0) = 3 , g (0) = −1

(b) f ' ( x) = g ( x) , g ' ( x) = f ( x)
(c) h' ( x ) = 2 f ( x ) f ' ( x ) − 2 g ( x ) g ' ( x ) = 2 f ( x ) g ( x ) − 2 f ( x ) g ( x ) = 0 ∀x
h( x ) = K ∀x , where K is a constant.
In particular, K = h(0) = [ f (0)] 2 − [ g (0)] 2 = 9 − 1 = 8
i.e. [ f ( x)] 2 − [ g ( x)] 2 = h( x) = 8

Page 10
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
(b) (ii)
5. (a) ∫ (5 − x) x − 1dx = − ∫ [( x − 1) − 4]( x − 1)1 / 2 dx
y
= − ∫ ( x − 1) 3 / 2 dx + 4∫ ( x − 1)1 / 2 dx
4
y = g (x) 2 8
=− ( x − 1) 5 / 2 + ( x − 1) 3 / 2 + C
5 3

5
(b) volume = π ∫ (5 − x) x − 1dx
1

5
x  2 8  128π
= π− ( x − 1) 5 / 2 + ( x − 1) 3 / 2  =
–4 –2 O 2 4  5 3 1 15

6. (a) x 3 − 13x + 12 = ( x − 1)( x 2 + x − 12) = ( x − 1)( x − 3)( x + 4)

4
dy
∫−4 g ( x)dx = area under graph of y = g (x) over [ −4 , 4]
(b) (i)
dy dt 4t
= = =t,
dy
=p (parameter of P is p)
dx dx 4 dx P
(4 + 8) × 4 dt
= = 24
2 −1
∴ slope of normal at P =
p
1
x⋅ − ln x
d  ln x  x 1 − ln x Equation of normal at P: x + py = (4 p ) + p(2 p 2 + 1)
4. (a)  = 2
=
dx  x  x x2 x + py = 2 p 3 + 5 p
k 3 −1 (ii) Put x = −24 and y = 31 , then
1 − ln(1 + ) 2 n 1 − ln(1 + k × )
1 2n n = lim 3 − 1 2n
(b) lim ∑ ∑ 0 = p 3 − 13 p + 12 = ( p − 1)( p − 3)( p + 4)
n →∞ n k =1 k n →∞ 2n k =1 3 −1 2
(1 + ) 2 (1 + k × )
n 2n I,e, p = 1, 3 or −4
3
3 1 − ln x  ln x  1
=∫ dx =   = ln 3
1
x2  x 1 3

Page 11
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
x −1 ( x + 3) − 4 1 4 (e)
7. (a) f ( x) = 3
= 3
= 2
− 3
( x + 3) ( x + 3) ( x + 3) ( x + 3)
y
2 12 −2( x − 3)
f ' ( x) = − + =
( x + 3) 3 ( x + 3) 4 ( x + 3) 4

6 48 6( x − 5)
f " ( x) = 4
− 5
=
( x + 3) ( x + 3) ( x + 3) 5
(b) (i) x > 1 or x < −3
(ii) x < −3 or −3 < x < 3 y = f ( x)

(iii) x > 5 or x < −3 1

(3, )
108 1
1 1 (5, )
(c) max point = (3, ), point of inflection = (5, ) 128
108 128 x
O 1
(d) vertical asymptote: x = −3
Let y = mx + c be the non-vertical asymptote. −1
27
f ( x) x −1
m = lim = lim =0
x →∞ x x →∞ x ( x + 3) 3

x −1
c = lim [ f ( x) − mx] = lim =0
x →∞ x →∞ ( x + 3) 3
∴ y = 0 is the horizontal asymptote.

 1
1 for k ≤ 0 or k > 108

 1
(f) n(k ) = 2 for k =
 108
 1
3 for 0 < k < 108


Page 12
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
8. (a) Differentiate y (1 + x 2 ) = 1 w.r.t. x by ( n + 2 ) times, 1 −2 x
(c) y= 2
, y' =
1+ x (1 + x 2 ) 2
[
0 = (1 + x ) y 2
]( n + 2)

6x 2 − 2
n+ 2
y" =
−2
(1 + x 2 ) 3
[(1 + x 2
]
) − x ⋅ 4x =
(1 + x 3 ) 3
= ∑ 2 ( r ) ( n + 2− r )
n + 2 C r (1 + x ) y
r =0

= n + 2 C 0 (1 + x 2 ) y ( n + 2) + n + 2 C1 (2 x) y ( n +1) + n + 2 C 2 (2) y ( n ) −2 x
When n = 1 , f 1 ( x) = (1 + x 2 ) 2 y ' = (1 + x 2 ) 2 × = −2 x
= (1 + x ) y 2 ( n + 2)
+ 2(n + 2) xy ( n +1)
+ (n + 2)(n + 1) y (n) (1 + x 2 ) 2
(b) (i) Multiple (1 + x 2 ) n + 2 on both sides of (a), then result follows. ∴ f 1 ( x) is a polynomial of degree 1 and
(ii) f n ( x) = (1 + x 2 ) n +1 y ( n ) the coefficient of x 1 = − 2 = (−1)1 (1 + 1)!
f n ' ( x) = (1 + x 2 ) n +1 y ( n +1) + 2(n + 1)(1 + x 2 ) n xy ( n ) 6x 2 − 2
When n = 2 , f 2 ( x) = (1 + x 2 ) 3 y" = (1 + x 2 ) 3 × 2 3
= 6x 2 − 2
Therefore, (1 + x )
2 2 n+2 ( n +1) 2 n +1 (n) ∴ f 2 ( x) is a polynomial of degree 2 and
(1 + x ) f n ' ( x) = (1 + x ) y + 2(n + 1)(1 + x ) xy
∴ (1 + x 2 ) f n ' ( x ) = (1 + x 2 ) n + 2 y ( n +1) + 2(n + 1) xf n ( x) the coefficient of x 2 = 6 = (−1) 2 (2 + 1)!
Differentiate once, Assume f k ( x) = [(−1) k (k + 1)! ]x k + q1 ( x) and
(1 + x 2 ) f n " ( x) + 2 xf n ' ( x ) f k +1 ( x ) = [(−1) k +1 (k + 2)! ]x k +1 + q 2 ( x) for some positive
= (1 + x 2 ) n + 2 y ( n + 2) + 2(n + 2)(1 + x 2 ) n +1 xy ( n +1) integer k, where deg q1 ( x) ≤ k − 1 and deg q 2 ( x) ≤ k .
+2(n + 1) f n ( x) + 2(n + 1) xf n ' ( x) Consider n = k + 2 , by (b)(i),
2 n +1 2 ( n + 2) ( n +1)
= (1 + x ) [(1 + x ) y + 2(n + 2) xy ] f k + 2 ( x) = −2(k + 2) xf k +1 ( x) − (k + 1)(k + 2)(1 + x 2 ) f k ( x )
+2(n + 1) f n ( x) + 2(n + 1) xf n ' ( x) {
= −2(k + 2) x [(−1) k +1 (k + 2)! ]x k +1 + q 2 ( x) }
= − (1 + x ) 2 n +1
(n + 2)(n + 1) y (n)
(by (a)) 2
{
− (k + 1)(k + 2)(1 + x ) [(−1) (k + 1)! ]x k + q1 ( x)
k
}
+2(n + 1) f n ( x) + 2(n + 1) xf n ' ( x) which is obviously a polynomial of degree k + 2 , with
= −(n + 2)(n + 1) f n ( x) +2(n + 1) f n ( x) + 2(n + 1) xf n ' ( x) the coefficient of x k + 2
i.e. (1 + x 2 ) f n " ( x) − 2nxf n ' ( x) + n(n + 1) f n ( x) = 0 = − 2(k + 2)[(−1) k +1 (k + 2)! ] − (k + 1)(k + 2)[(−1) k (k + 1)! ]
= (−1) k + 2 (k + 2)(k + 1)!{2(k + 2) − (k + 1)}
= (−1) k + 2 (k + 2)(k + 1)! (k + 3)
= (−1) k + 2 (k + 3)! . The proof is completed.

Page 13
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
(d) By (c), we can let 9. (a) * The inequality is obviously true for n = 1 .
n n n −1 n−2
f n ( x) = (−1) (n + 1)! x + a n −1 x + a n−2 x + ... + a1 x + a 0 2n 4
* We are going to show by induction that ≤ for integer n ≥ 2 .
where a i ∈ ℜ n! n
n −1 n−2
n
f n ' ( x) = (−1) n(n + 1)! x + (n − 1)a n −1 x + (n − 2)a n − 2 x n −3 + ... + a1 22 4
When n = 2 , =2= ∴ the statement is true for n = 2 .
f n " ( x) = (−1) n n(n − 1)(n + 1)! x n − 2 + (n − 1)(n − 2)a n −1 x n −3 2! 2
+ (n − 2)(n − 3)a n − 2 x n − 4 + ... + 2a 2 2k 4
Assume ≤ for some integer k ≥ 2 .
k! k
By (b)(ii), Consider n = k + 1 ,
(1 + x 2 ) f n " ( x) − 2nxf n ' ( x) + n(n + 1) f n ( x) = 0 2 k +1 2k 2 4 2
= ≤ (by assumption)
Compare coefficient of x n −1 : (k + 1)! k! k + 1 k k + 1
(n − 1)(n − 2)a n −1 − 2n(n − 1)a n −1 + n(n + 1)a n −1 = 0 4
≤ (Q k ≥ 2 )
2a n −1 = 0 , ∴ a n −1 = 0 k +1

2n 4 4
Now 0 < ≤ and lim 0 = 0 = lim
n n! n n →∞ n →∞ n
Now ∑ ak = sum of roots of f n ( x) = 0
k =1 2n
∴ lim =0 (by sandwich theorem)
n →∞ n!
(−1) coefficient of x n −1
=
coefficient of x n e −1 e
I 1 = ∫ x −3 ln xdx =
2 ∫1
(b) (i) ln xdx − 2
1
−a n −1
=
(−1) n (n + 1)!
=0
=−
2
[
1 −2 e
]
1 e
x ln x 1 + ∫ x − 2 d ln x
2 1

e −2 1 e −3
=− + ∫ x dx
2 2 1

e −2 1 − 1 − 2
=−
2
+ ( )x
2 2
[ ] e
1

1
= (1 − 3e − 2 )
4

Page 14
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com

e −1 e e −2 r + 1
I n +1 = ∫ x −3 (ln x) n +1 dx =
2 ∫1
(ii) (ln x) n +1 dx − 2 I r +1 = − + Ir
1
2 2
e −2 r + 1  1 r 
=−
2
[
1 −2
x (ln x) n +1 ]
e
1 +
1 e −2
2 ∫1
x d (ln x) n +1 =−
2
+
2 
1
(r! ) r +1 − 2 ∑
− k )!2
1
k +1


2 e k = 0 (r 
e −2 n + 1 e −3  1 1 1 r
1 
= (r + 1)! r + 2 − 2 ∑ 
2 ∫1
=− + x (ln x) n dx − 2
2 e ⋅ 2 ⋅ (r + 1)! e − k )!2 k +2 
2  k =0 (r 
e −2 n + 1  1 1 1 r +1
1 
=− + In = (r + 1)! r + 2 − 2 − 2 ∑ 
2 e ⋅ 2 ⋅ (r + 1)! e (r − t + 1)!2 t +1 
2 2  t =1 
 1 1 n
1  (put t = k + 1 )
(iii) When n = 1 , n! n +1 − 2 ∑ 
2 k +1   1 
 e k = 0 ( n − k )!2  1 r +1
1
= (r + 1)! r + 2 − 2 ∑ 
2 k +1 
1 1 1
1  e k = 0 ( r + 1 − k )!2 
= −
4 e2
∑ k +1
k = 0 (1 − k )!2 The proof is completed.
1 1 1 1 (iv) For 1 ≤ x ≤ e , e −2 ≤ x −2 ≤ 1
= − ( + )
4 e2 2 4 ∴ e −2 x −1 (ln x) n ≤ x −3 (ln x) n ≤ x −1 (ln x) n

1 e e e
= (1 − 3e − 2 ) = I 1 ∴ e −2 ∫ x −1 (ln x) n dx ≤ ∫ x −3 (ln x) n dx ≤ ∫ x −1 (ln x ) n dx
4 1 1 1

 1 1 r
1 
Assume I r = r!  r +1 − 2 ∑  for some e e
2 k +1  ∴ e −2 ∫ x −1 (ln x) n dx ≤ I n ≤ ∫ x −1 (ln x) n dx
− k )!2 1 1
 e k =0 (r 
positive integer r. e
e −1 e  (ln x) n +1  1
where ∫ 1
x (ln x) n dx = ∫ (ln x) n d ln x = 
1 +
 =
+1
 n 1  1 n

1 1
i.e. ≤ In ≤
e 2 (n + 1) n +1

Page 15
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
n
1 1 n
2 n −k 1 n
2k
H ' ( x) = 2 ∫ f (t ) g (t )dt  f ( x) g ( x)
x
(c) Note that ∑ k +1
= n +1 ∑ (n − k )! = n +1 ∑ 10. (a) (i) (1)
 0 
k = 0 ( n − k )!2 2 k =0 2 k = 0 k!

 1 1 1 n
2k   n
2k 
 = n! 1 − 1 −  ∫ ( f (t )) 2 dt ( g ( x)) 2 −  ∫ ( g (t )) 2 dt ( f ( x )) 2
x x
I n = n!  n +1 − 2 n +1
2 ∑  2 n +1  e 2 ∑ 

(by (iii))
 0   0 
 e 2 k = 0 k!   k = 0 k! 
1 n!  1 n
2k 
≤ 1
∴ 2
≤ n +1 
1− 2 ∑  n +1
(by (iv))
x
{
= − ∫ [ f (t ) g ( x)] 2 − 2[ f (t ) g ( x)][ f ( x) g (t )] + [ f ( x) g (t )] 2 dt }
k = 0 k!
e (n + 1) 0
2  e 
1 2 n +1 1 n
2k 2 n +1 x
∴ 2 ( n + 1)!
≤ 1− 2 ∑ ≤
(n + 1)!
= − ∫ [ f (t ) g ( x ) − f ( x) g (t )] 2 dt
k = 0 k!
e e 0

2 n +1 n
2k 2 n +1 ≤0
∴ e 2 [1 − ]≤ ∑ ≤ e2 −
(n + 1)! k =0 k! (n + 1)! i.e. H ( x) is decreasing on I.

2 n +1 2 n +1 (2) Q H ( x) is decreasing on I.
As lim e 2 [1 − ] = e 2 = lim [e 2 − ]
n →∞ (n + 1)! n→∞ (n + 1)! ∴ H ( x ) ≤ H (0 ) = 0 for all x ∈ I .
∞
2k n
2k 2
 x f (t ) g (t )dt  ≤  x( f (t )) 2 dt  x( g (t )) 2 dt 
∴ ∑ k! n→∞ ∑ k! = e 2
= lim (by sandwich theorem) i.e.
 ∫0   ∫0  ∫0 
k =0 k =0

for all x ∈ I .

Page 16
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
(ii) (1) Put f (t ) = 1 and g (t ) = h' (t ) 11. (a) Suppose the converse that sin θ = 0 or sin φ = 0 .
such that f and g are continuous on I. * Consider sin θ = 0 , P = (0, 20 cos θ )
By (a)(i)(2), result follows by noting that h(0) = 0 . As O, P and Q are collinear,

x 1 x we have sin φ = 0
∫ 0 (h' (t )) dt = ∫ (h' (t )) 2 dt − ∫ (h' (t )) 2 dt
2
(2) * Note that
0 1 Thus, cos θ ≠ 0 and cos φ ≠ 0

1 2x 2 y dy
≤ ∫ (h' (t )) 2 dt * Now differentiate the equation of (E): + =0
0 144 400 dx
(Q (h' (t )) 2 ≥ 0 and x ≤ 1 ) dy 5 sin θ dy 5 sin φ
∴ =− , =−
dx P 3 cos θ dx Q 3 cos φ
x
* (h( x)) 2 ≤ x ∫ (h' (t )) 2 dt (by (a)(ii)(2))
0 (as cos θ ≠ 0 and cos φ ≠ 0 )

1 dy dy
≤ x ∫ (h' (t )) 2 dt (proved before) * ∴ =0, =0
0 dx P dx Q

Taking integration for the above inequality over [0, 1] ∴ the slope of tangents at P
=the slope of tangents at Q
dt ≤ ∫  x ∫ (h' (t )) 2 dt dx
1 1 1
∫ 0 (h( x))
2
0 0  As P and Q are distinct points with equal
slopes of tangents, the two tangents cannot
=    dt 
1 1
∫ 0 xdx  ⋅  ∫ 0 (h' (t ))
2
  have an intersection.

1 1 1 1 Contradiction!
 ∫ 0 (h' (t )) dt  = ∫ 0 (h' ( x )) dx
2 2
=
2   2 Similar argument can be applied on the case sin φ = 0 .
To sum up, sin θ ≠ 0 and sin φ ≠ 0
(b) Put h( x) = ln(sec x + tan x) for x ∈ [0, 1]

sec x tan x + sec 2 x

such that h' ( x ) = = sec x which is continuous on [0, 1]
sec x + tan x
and h ( 0) = 0 .

1 1 1 2 1 1
∫0 (ln(sec x + tan x)) sec xdx = [tan x ]10 = tan 1
2 ∫0
2
dx ≤
2 2

Page 17
Suggested Solution for Pure Mathematics 2010 By Y.K. Ng (last update: 14/4/2010) ykng2007@hotmail.com
dy 5 sin θ  x(5 sin θ) + y (3 cos θ) = 60
(b) (i) =− (ii) 
dx P 3 cos θ  x(5 sin φ) + y (3 cos φ) = 60
Equation of tangent at P: 5 sin θ 60
x(5 sin θ) + y (3 cos θ) = (12 sin θ)(5 sin θ) + (20 cos θ)(3 cos θ) 5 sin φ 60 20(sin θ − sin φ) 20(sin θ − sin φ)
y= = =
∴ x(5 sin θ) + y (3 cos θ) = 60 5 sin θ 3 cos θ sin θ cos φ − cos θ sin φ sin(θ − φ)

Equation of tangent at Q: 5 sin φ 3 cos φ

x(5 sin φ) + y (3 cos φ) = 60 5 sin(θ − φ)

20 ×
(ii) Suppose the converse that sin(θ − φ) = 0 . 4
= = 25 (note: sin(θ − φ) ≠ 0 )
Then, θ − φ = nπ for n ∈ Z. sin( θ − φ)

θ = nπ + φ Substitute into the first equation,

∴ sin θ = sin( nπ + φ) = sin nπ cos φ + cos nπ sin φ 60 − (25)(3 cos θ)
x= (note: sin θ ≠ 0 )
n 5 sin θ
= (−1) sin φ
cos θ = cos(nπ + φ) = cos nπ cos φ − sin nπ sin φ 3(4 − 5 cos θ)
=
n sin θ
= (−1) cos φ
∴ P = ( 12(−1) n sin φ, 20(−1) n cos φ ), Q =( 12 sin φ, 20 cos φ ) 3(4 − 5 cos θ)
R=( , 25)
Obvious that the line PQ passes through the origin and hence it sin θ
cannot pass through A (0, 16). 20 cos θ − 16 5 cos θ − 4
(c) slope of PA = =
Contradiction! 12 sin θ − 0 3 sin θ
(c) (i) slope of AP = slope of PQ (as A, P and Q are collinear) 25 − 26 3 sin θ
slope of AR = =
20 cos θ − 16 20 cos θ − 20 cos φ 3(4 − 5 cos θ) 4 − 5 cos θ
= −0
12 sin θ − 0 12 sin θ − 12 sin φ sin θ

(5 cos θ − 4)(sin θ − sin φ) = 5(cos θ − cos φ) sin θ ∴ (slope of PA)( slope of AR) = −1
5 cos θ(sin θ − sin φ) − 4(sin θ − sin φ) = 5 sin θ(cos θ − cos φ) i.e. ∆PAR is a right angled triangle.
4(sin θ − sin φ) = 5 cos θ(sin θ − sin φ) − 5 sin θ(cos θ − cos φ)
= 5(sin θ cos φ − cos θ sin φ) --- End of Solutions of Paper II ---
= 5 sin(θ − φ)

Page 18