Find the sum of all positive integers n such that n ! 1 ! × 2 ! × 3 ! × ⋯ × 2 0 0 ! is a perfect square.
This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try
refreshing the page, (b) enabling javascript if it is disabled on your browser and,
finally, (c)
loading the
non-javascript version of this page
. We're sorry about the hassle.
Very nicely done! It is great that you know the Bertand's Postulate, hopefully you have also seen its proof. Here is a pretty good one, try to understand it, if you have not done so already: http://en.wikipedia.org/wiki/Proof of Bertrand's_postulate
As you mentioned in your solution, the Bertand's Postulate is not really necessary. Besides 211, you just need one more prime, for example 157. Also note that most proofs of Bertrand's Postulate actually involve direct checking for numbers in this range.
Exquisite!
We work first with the numerator. Rewrite the expression by factoring the highest even term from each even factorial term. ( 1 ! ) 2 ∗ 2 ∗ ( 3 ! ) 2 ∗ 4 ∗ . . . ∗ ( 1 9 9 ! ) 2 ∗ 2 0 0
Then regroup the terms,
= ( 1 ! ∗ 3 ! ∗ . . . ∗ 1 9 9 ! ) 2 ∗ ( 2 ∗ 4 ∗ . . . ∗ 2 0 0 )
Then factor out 2^100,
= ( 1 ! ∗ 3 ! ∗ . . . ∗ 1 9 9 ! ) 2 ) ∗ 2 1 0 0 ∗ 1 0 0 !
Removing the largest square part, the question reduces to:
When is 1 0 0 ! / n ! a perfect square?
Clearly for n = 100 and n = 99, we have perfect squares. We shall verify there are no others.
For n = 98, we have (99)(100) is not a perfect square.
For n = 97, we have (98)(99)(100) is not a perfect square.
For n < 97, we have 97 is a prime factor of 100!/n! with multiplicity one. Hence 100!/n! is not a perfect square.
The answer is then 99 + 100 = 199.
Nicely done!
What if n > 1 0 0 ?
Log in to reply
Then the number is not integral. We were asked to find only integral 'perfect squares' right?
Let A = n ! 1 ! × 2 ! × 3 ! × ⋯ × 2 0 0 ! .
One of the requirements for A to be a perfect square, is that A is integer. For n ≥ 2 1 1 , the denominator is divisible by the prime number 2 1 1 , whereas the numerator isn't (it can be trivially decomposed into factors that are all smaller). Therefore, A is then not integer, so we must have n < 2 1 1 . Note that this is a necessary condition, but not necessarily a sufficient one (it is, but I won't bother proving it). For the moment, let us not worry about this, and see for what values n < 2 1 1 , A is the square of a rational number.
Note that, in order to test whether a number is a rational square, we can freely divide and multiply by any square: the resulting number is a rational square if and only if the original number is. We will make full use of this.
First off, note that for instance 5 9 ! × 6 0 ! can be written as 6 0 × ( 5 9 ! ) 2 , so we can do away with the square and be left with 6 0 . Generalizing this, if we divide A by ( 1 ! ) 2 , ( 3 ! ) 2 , ( 5 ! ) 2 , and all the way up to ( 1 9 9 ! ) 2 , A reduces to
A 1 = n ! 2 × 4 × 6 × ⋯ × 2 0 0
We can divide this by ( 2 5 0 ) 2 , and be left with
A 2 = n ! 1 0 0 !
Finally, let us multiply by ( n ! ) 2 , to get A 3 = 1 0 0 ! × n ! .
An integer (like A 3 ) is a square if and only if every prime number divides it an even number of times (this is easy to prove using unique prime factorization).
Let us consider the prime factor 9 7 (not coincidentally the largest prime number below 100). It divides 1 0 0 ! once, and n ! is divisible ⌊ n / 9 7 ⌋ times. Clearly we must then have that ⌊ n / 9 7 ⌋ is odd, which together with the restriction n < 2 1 1 , means that
9 7 ≤ n < 2 × 9 7
Now consider the prime factor 1 0 1 . This does not divide 1 0 0 ! , so ⌊ n / 1 0 1 ⌋ must be even. This (together with n < 2 1 1 ) means that either n < 1 0 1 or n ≥ 2 0 2 .
Combining the two regimes, we are left with the possibilities n ∈ { 9 7 , 9 8 , 9 9 , 1 0 0 } . For n = 9 7 and n = 9 8 , we find A 3 = 9 8 × 9 9 × 1 0 0 and A 3 = 9 9 × 1 0 0 respectively, which are both obviously not squares. For n = 9 9 and n = 1 0 0 , we find A 3 = 1 0 0 and A 3 = 1 , which are squares. Hence, in these cases, and only in these cases, is A the square of a rational number, and potentially integer.
Finally, we should remember to check whether A is integer in these cases. This is however trivially true for all n ≤ 2 0 0 .
The only solutions are thus n = 9 9 and n = 1 0 0 , and the final answer is 9 9 + 1 0 0 = 1 9 9
This is a very nice solution! A little misprint near the end: when checking n = 9 7 , 9 8 , 9 9 , 1 0 0 what is currently A 3 was clearly meant to be A 2
Let P = 1 ! × 2 ! × 3 ! × . . . × 2 0 0 ! . Then, using the identity n ! = ( n − 1 ) ! × n , we obtain
P = 1 ! × 2 ! × 3 ! × . . . × 2 0 0 ! = 2 × 3 ! × ( 3 ! × 4 ) × 5 ! × ( 5 ! × 6 ) × . . . × 1 9 9 ! × ( 1 9 9 ! × 2 0 0 ) = ( 3 ! × 5 ! × . . . × 1 9 9 ! ) 2 × ( 2 × 4 × . . . × 2 0 0 ) = ( 3 ! × 5 ! × . . . × 1 9 9 ! ) 2 × 2 1 0 0 × 1 0 0 ! ( ∗ )
We want to divide P by n ! to obtain a square. Observe that the first two terms in ( ∗ ) are both squares, so we can divide ( ∗ ) by 1 0 0 ! to obtain a square. Since 1 0 0 ! = 9 9 ! × 1 0 2 , we can also divide ( ∗ ) by 9 9 ! to obtain another square.
Now, we shall show that there are no other possible n . Observe that 2 1 1 is a prime, so if n ≥ 2 1 1 , then 2 1 1 appears as a prime factor in the denominator but not in the numerator, and we would have n ! P being a non-integer. So n < 2 1 1 . Furthermore, since 1 9 9 is prime, so we can easily see from ( ∗ ) that 1 9 9 appears exactly twice as factors of P . If 1 9 9 ≤ n < 2 1 1 , then n ! would have exactly one factor of 1 9 9 ; thus if n ! P is an integer, then it will have exactly one factor of 1 9 9 and cannot be a square. So n < 1 9 9 .
Now, because 1 0 1 is prime, the only times it appears as factors of ( ∗ ) are in 1 0 1 ! , 1 0 3 ! , . . . , 1 9 9 ! , which are all squared. Therefore, 1 0 1 appears an even number of times. If we choose 1 0 1 ≤ n ≤ 1 9 8 , then n ! would have exactly one factor of 1 0 1 and thus n ! P will have an odd number of factors of 1 0 1 and cannot be a square. So n < 1 0 1 . But 9 7 is also prime, and besides appearing as a factor in the squared term of ( ∗ ) , it also appears exactly once in the factor 1 0 0 ! . So 9 7 appears as a factor of P an odd number of times, so we would need to cancel out at least one factor of 9 7 in the denominator. So 9 7 ∣ n ! ⇒ n ≥ 9 7 . Thus, we have 9 7 ≤ n ≤ 1 0 0 .
We may easily check that n = 9 7 or n = 9 8 do not make n ! P a square, by observing that 9 7 ! P = ( 3 ! × 5 ! × . . . × 1 9 9 ! ) 2 × 2 1 0 0 × 9 8 × 9 9 and that 9 8 ! P = ( 3 ! × 5 ! × . . . × 1 9 9 ! ) 2 × 2 1 0 0 × 9 9 . Clearly, 9 9 and 9 8 × 9 9 are not squares, so n ! P is not a square for n = 9 7 and n = 9 8 .
Therefore, the only possible n are n = 9 9 and n = 1 0 0 , so the sum of all n is 1 9 9 .
Nicely done!
By taking the number of times each number occurs across the factorials, the numerator can be re-written: 2 1 9 9 ∗ 3 1 9 8 ∗ 4 ∗ 1 9 7 ∗ . . . ∗ 1 9 9 2 ∗ 2 0 0 1 .
Now, since even bases have odd powers, the numerator is a perfect square times 2 ∗ 4 ∗ 6 ∗ 8 ∗ . . . ∗ 2 0 0 = 2 1 0 0 ∗ 1 0 0 !
Thus, the fraction is a perfect square only when n! has the same primes with odd powers as 100!. Since 100! is divisible by the primes 53, 59, 61, ... 89, 97, exactly once, n! must have an odd number of all of them. This can only occur when ( 2 k − 1 ) ∗ 9 7 ≤ n < ( 2 k ∗ 5 3 ) , for some integer k. Furthermore, if n ≥ 2 1 1 , n! will contain primes the denominator does not, and therefore the original fraction won't even be an integer. Combining these, we see 9 7 ≤ n < 1 0 6
For n < 1 0 0 , n ! 1 0 0 ! is a perfect square iff n! has the same prime with odd powers as 100!.
For n = 97, 9 7 ! 1 0 0 ! = 1 0 0 ∗ 9 9 ∗ 9 8 , which is not a perfect square.
For n = 98 9 8 ! 1 0 0 ! = 1 0 0 ∗ 9 9 , which is also not a perfect square.
For n = 99 9 7 ! 1 0 0 ! = 1 0 0 , a perfect square, so n = 99 is a solution
For 1 0 0 < n < 1 0 6 , 101 divides n! exactly once, so 1 0 0 ! n ! cannot be a perfect square.
The only solutions are 100 and 99, giving a sum of 199
You mean 4 1 9 7 or 4 ∗ 1 9 7 ?
Log in to reply
Sorry, yes I meant 4 1 9 7 , I messed up the formatting.
this part "Thus, the fraction is a perfect square only when n! has the same primes with odd powers as 100!" is true but not clear enough, you should explain it more as others may won't understand
Why can you conclude that ( 2 k − 1 ) ∗ 9 7 ≤ n < ( 2 k ∗ 5 3 ) for some integer k ?
I think the solution is 9 8 + 9 9 .
Log in to reply
OH
This is impossible for 9 8 ! contains 8 factors of prime 11 while 9 9 ! has 9 factors of 11.
By the identity: n ! ( n + 1 ) ! = ( n ! ) 2 ( n + 1 ) for all n ∈ N , the product of the factorials 1 ! ⋅ 2 ! ⋯ 2 0 0 ! can be expressed as ( 1 ! ) 2 2 ⋅ ( 3 ! ) 2 4 ⋅ ( 5 ! ) 2 6 ⋯ ( 1 9 7 ! ) 2 1 9 8 ⋅ ( 1 9 9 ! ) 2 2 0 0 . Since 2 ⋅ 4 ⋯ 2 0 0 = ( 2 ⋅ 1 ) ( 2 ⋅ 2 ) ⋯ ( 2 ⋅ 1 0 0 ) = 2 1 0 0 1 0 0 ! , the product above becomes 1 0 0 ! ⋅ 2 1 0 0 A 2 = 1 0 0 ! ( 2 5 0 A ) 2 , where A = 1 ! ⋅ 3 ! ⋅ 5 ! ⋯ 1 9 9 ! . Let n be a positive integer such that n ! 1 0 0 ! ⋅ ( 2 5 0 A ) 2 ( ∗ ) is a perfect square. Notice that n ≤ 2 1 0 . For otherwise the numerator of ( ∗ ) would contain the prime factor 2 1 1 , which is impossible. Since there are an even number of prime factors that are larger than 1 0 0 in the numerator of ( ∗ ) (as factors of A 2 ) while each prime factor between 1 0 0 and 2 1 0 appears only once in n ! , it follows that n ≤ 1 0 0 . Since there is an odd number of prime factor 9 7 in the numerator of ( ∗ ) , it follows that n ≥ 9 7 . Hence there are only 4 cases to consider for the value of n : 9 7 , 9 8 , 9 9 , and 1 0 0 . We note a basic fact from number theory that for all positive integers a and b if a b 2 is a perfect square, then a is a perfect square.
case I
If n = 9 7 then ( ∗ ) becomes 9 8 ⋅ 9 9 ⋅ 1 0 0 ( 2 5 0 A ) 2 = ( 9 8 ) ( 9 9 ) ( 1 0 ⋅ 2 5 0 A ) 2 . Since ( 9 8 ) ( 9 9 ) , ending in a 2 , is not a perfect square, this case is impossible.
case II
If n = 9 8 then ( ∗ ) becomes 9 9 ⋅ 1 0 0 ( 2 5 0 A ) 2 = 9 9 ( 1 0 ⋅ 2 5 0 A ) 2 . Since 9 9 is not a perfect square, this case is impossible.
case III
If n = 9 9 then ( ∗ ) becomes 1 0 0 ( 2 5 0 A ) 2 = ( 1 0 ⋅ 2 5 0 A ) 2 , which is a perfect square, as required.
case IV
If n = 1 0 0 then ( ∗ ) becomes ( 2 5 0 A ) 2 , which is a perfect square, as required.
Hence there are only two values of n namely n = 9 9 and n = 1 0 0 for which ( ∗ ) is a perfect square. The sum of these n 's equal 1 9 9 .
first consider the numerator n! * (n+1)!=(n!)^2 (n+1) using this we pair (1! * 2!) (3! * 4!) (5! * 6!) *(7! * 8!) ........... (197! *198!) *(199! * 200!) we pair up all of the squares as a single one say k^2 then numerator reduces to (k^2) 2 * 4 * 6 * 8 *......198 * 200 taking 2 common and clubbing it with the square part we get numerator as K^2 *100! it can be easily seen that can be either only 100 or 99. since if n>100 then denominator after cancellation with 100! from numerator will not be a perfect square. and also if n<99 then it leaves 99 in the numerator which cannot be reduced .
To begin with, let us modify the expression 1 ! × 2 ! × 3 ! × ⋯ × 2 0 0 ! a bit.
1 ! × 2 ! × 3 ! × ⋯ × 2 0 0 ! =
= 1 × 2 × 3 ! × ( 3 ! × 4 ) × 5 ! × ( 5 ! × 6 ) × ⋯ × 1 9 9 ! × ( 1 9 9 ! × 2 0 0 ) =
= ( 3 ! × 5 ! × 7 ! × ⋯ × 1 9 9 ! ) 2 × 2 × 4 × 6 × ⋯ × 2 0 0 =
= ( 3 ! × 5 ! × 7 ! × ⋯ × 1 9 9 ! ) 2 × 2 × 1 × 2 × 2 × 2 × 3 × ⋯ × 2 × 1 0 0 =
= ( 3 ! × 5 ! × 7 ! × ⋯ × 1 9 9 ! ) 2 × 2 1 0 0 × ( 1 × 2 × 3 × 4 × ⋯ × 1 0 0 ) =
= ( 3 ! × 5 ! × 7 ! × ⋯ × 1 9 9 ! ) 2 × 2 1 0 0 × 1 0 0 ! =
= ( 3 ! × 5 ! × 7 ! × ⋯ × 1 9 9 ! × 2 5 0 ) 2 × 1 0 0 !
Therefore, the original expression will now be
n ! ( 3 ! × 5 ! × 7 ! × ⋯ × 1 9 9 ! × 2 5 0 ) 2 × 1 0 0 !
We have now achieved a complete square with the exception of 1 0 0 ! .
Speaking of which, let us see if any factorial can be a perfect square. If the largest prime factor p of the factorial m ! (where m is a positive integer) does not occur in the factorial again, no matter how you expand it essentially means that the factorial can not be a perfect square. If you do not encounter integers of the type 2 p , 3 p or any t p (where t is a positive integer), in the expansion of the factorial, then the best case scenario we could possibly get would be to have a perfect square times a prime number, which is in no case a perfect square.
Consequently, we learn that 1 0 0 ! is in no way a perfect square because 9 7 is the largest prime in the expansion and it is not a divisor of 9 8 , 9 9 or 1 0 0 . This means that if we get rid of it the expression is going to be a perfect square. Thus, one possible solution for n is n = 1 0 0 .
However, note that 1 0 0 = 1 0 2 and 1 0 0 ! = 9 9 ! × 1 0 0 . This allows us to tweak the original expression even further, giving us
n ! ( 3 ! × 5 ! × 7 ! × ⋯ × 1 9 9 ! × 2 5 0 × 1 0 2 ) 2 × 9 9 !
which essentially provides us with the second solution for n , which is n = 9 9 .
The only thing left at this point is to see if there are any other solutions for n . Since 9 9 and 9 8 are not a perfect squares, we may not use the method we did for getting n = 9 9 . Furthermore, since 9 7 is a prime integer it is evident that n ! 9 9 ! can not be a square for any 1 ≤ n ≤ 9 9 , again due to the fact that it is not a divisor of 9 8 and 9 9 . Thus, any sum left of 9 9 ! for the fraction can not be a square because 9 7 does not repeat for any expansion.
We can now see that the only solutions are n 1 = 1 0 0 and n 2 = 9 9 . Therefore, the required sum is n 1 + n 2 = 1 9 9 .
Note that ( 2 k ) ! ( 2 k − 1 ) ! = ( 2 k − 1 ) ! 2 ⋅ 2 k
So we can rewrite the original expression X as: X = n ! 1 ( k = 1 ∏ k = 1 0 0 ( 2 k − 1 ) ! ) 2 ⋅ k = 1 ∏ k = 1 0 0 ( 2 k ) = n ! 1 ( k = 1 ∏ k = 1 0 0 ( 2 k − 1 ) ! ) 2 ⋅ 2 1 0 0 ⋅ 1 0 0 !
The terms in the numerator are a perfect square except for 1 0 0 ! . So in order for X to be a perfect square, it is necessary and sufficient that Y = n ! 1 0 0 ! be a perfect square.
This can't happen if n > 1 0 0 : n ! has 1 0 1 as a prime factor, so it doesn't divide evenly into 1 0 0 .
Now, if n < 9 7 then Y has exactly one factor of 9 7 in it (a prime number). So it can't be a square.
And if 9 7 ≤ n < 9 9 then Y has exactly one factor of 1 1 in it, so it can't be a square either.
All that remains is to check n = 1 0 0 , where we get Y = 1 , and n = 9 9 we get Y = 1 0 0 . These are both squares, so these are the only two solutions. They sum to 1 9 9 .
Hm, published this ages ago and it got phantom unpublished :) (Same thing happened with my solution to the "n-gon" combinatorics problem - it unpublished even after I'd put a comment on it). Assuming this is a bug rather than a CM action?
the answer is given in the title of the question -_-
We can rewrite the given number as: n ! ( 1 ! ) 2 ⋅ ( 3 ! ) 2 ⋅ ( 5 ! ) 2 ⋯ ( 1 9 9 ! ) 2 ⋅ 2 ⋅ 4 ⋅ 6 ⋯ 2 0 0 = n ! ( 1 ! ⋅ 3 ! ⋅ 5 ! ⋯ 1 9 9 ! ) 2 ⋅ 2 1 0 0 ⋅ 1 0 0 ! = ( 1 ! ⋅ 3 ! ⋅ 5 ! ⋯ 1 9 9 ! ⋅ 2 5 0 ) 2 ⋅ n ! 1 0 0 !
Now we have to find some n for which ( 1 ! ⋅ 3 ! ⋅ 5 ! ⋯ 1 9 9 ! ⋅ 2 5 0 ) 2 ⋅ n ! 1 0 0 ! is a square.
We can see that the only available numbers for n are 1 0 0 and 9 9 .
To understand why we have to notice that 1 0 1 and 9 7 are prime numbers, so n can not be smaller than 9 7 in order to have an even exponent of 9 7 in the prime factorization of the given number. If n is greater of equal to 1 0 1 it has to be bigger than 2 0 2 in order to have an even exponent of 1 0 1 in the prime factorization of the given number. But by induction n must be infinite. In fact we know that between x and 2 x with x > 1 there is always a prime number, therefore there will be a prime between 1 0 1 and 2 0 2 . In order to have an even exponent of that prime n must be bigger that 2 times the prime... and so on...
The only possible values of n are: 1 0 0 , 9 9 , 9 8 and 9 7 . And the only that actually works are 1 0 0 and 9 9
So the answer is 1 9 9
Problem Loading...
Note Loading...
Set Loading...
Let S = 1 ! × 2 ! × 3 ! × ⋯ × 2 0 0 ! . We have
S = 1 ! × 2 ! × 3 ! × 4 ! × ⋯ × 1 9 9 ! × 2 0 0 ! = 1 ! × ( 1 ! ⋅ 2 ) × 3 ! × ( 3 ! ⋅ 4 ) × ⋯ × 1 9 9 ! × ( 1 9 9 ! ⋅ 2 0 0 ) = ( 1 ! × 3 ! × ⋯ × 1 9 9 ! ) 2 × ( 2 × 4 × ⋯ × 2 0 0 ) = ( 1 ! × 3 ! × ⋯ × 1 9 9 ! ) 2 × 2 1 0 0 × 1 0 0 ! (*)
If n = 1 0 0 , then n ! S = ( 1 ! × 3 ! × ⋯ × 1 9 9 ! ) 2 × 2 1 0 0 , and if n = 9 9 , then n ! S = ( 1 ! × 3 ! × ⋯ × 1 9 9 ! ) 2 × 2 1 0 0 × 1 0 0 . Therefore, 9 9 , 1 0 0 are values of n . We prove that there are no other values.
Assume that there are other values of n . Let n ! S = m 2 for some integer m , and let S = k 2 ⋅ 1 0 0 ! where k 2 is the perfect square from ( ∗ ) . Then we have k 2 ⋅ 1 0 0 ! = m 2 ⋅ n ! If 1 ≤ n ≤ 9 6 , the right-hand side of this equation will contain an even number of factors of the prime 9 7 , while the left-hand side will contain an odd number of factors of 9 7 (since 1 0 0 ! is divisible by 9 7 only once). This is a contradiction, therefore no value of n < 9 7 will work. By inspection, n = 9 7 , 9 8 do not work, and we found already that n = 9 9 , 1 0 0 are solutions, so we now have to consider n ≥ 1 0 1 .
Take the equation k 2 ⋅ 1 0 0 ! = m 2 ⋅ n ! once again. If 1 0 1 ≤ n ≤ 2 0 1 , then n ! is only divisible once by the prime 1 0 1 , so the right-hand side of the equation contains an odd number of factors of 1 0 1 . We know that 1 0 1 ∣ 1 0 0 ! , so the left-hand side has an even number of factors of 1 0 1 , again a contradiction. Therefore, any other values of n must satisfy n ≥ 2 0 2 .
Take n ≥ 2 0 2 . By Bertrand's Postulate, there exists a prime p between ⌊ n / 2 ⌋ and n . Then n ! is only divisible by p once, and p ∣ 1 0 0 ! . So, p appears an odd number of times in m 2 ⋅ n ! but appears an even number of times in k 2 ⋅ 1 0 0 ! , another contradiction.
Therefore, no other values of n exist such that n ! S is a perfect square. The answer is 1 0 0 + 9 9 = 1 9 9 .
Note: An alternate way to check n ≥ 2 0 2 would be to use the prime 2 1 1 and then check n = 2 0 2 , 2 0 3 , … , 2 1 0 ; I used a different way to avoid excessive computation.