Consider the following problem:
Show that there do no exist non-negative integers k and m such that ( k! + 48 = 48(k+1)^m )
If there are such integers then ( k! = 48 \times ((k+1)^m - 1 )).
Expanding we have ( k! = 48 \times (k^m + m \cdot k^{m-1} + {m \choose 2} k^{m-2} + \cdots + {m \choose {m-2}} k^2 + {m \choose {m-1} } k + 1 - 1 ) )
Hence ( k! = 48k\times ( k^{m-1} + m \cdot k^{m-2} + {m \choose 2} k^{m-3} + \cdots + {m \choose {m-2}} k + {m \choose {m-1} } ) )
Dividing by k on both sides we have ( (k-1)! = 48\times ( k^{m-1} + m \cdot k^{m-2} + {m \choose 2} k^{m-3} + \cdots + {m \choose {m-2}} k + {m \choose {m-1} } ) ) ... (equation (1) - we will use this again later)
This implies 48 divides (k-1)! or ( (k-1) \geq 6 \Rightarrow k \geq 7 )
In fact, we can quickly check that k = 7 and k = 8 does not work. After all, for k = 7, ( \frac{7!}{48} + 1 = 105 + 1 = 106 ) is not ( 8^m ) for any m. Similarly for k = 8, ( \frac{8!}{48} + 1 = 840 + 1 = 841 ) is not ( 9^m ) for any m.
( \frac{k!}{48} + 1 = (k+1)^m )
Since ( k > 8 ), then ( \frac{k!}{48} = \frac{1\times 2 \times 3 \times \cdots \times 8 \times \cdots k}{48} = 840 \times \cdots \times k ). Hence ( \frac{k!}{48} ) is even (as it is a multiple of 840). This implies ( \frac{k!}{48} + 1 ) is odd.
But that means ( (k+1)^m ) is odd or k+1 is odd. Hence k is even.
Let ( k = 2k_1 ).
We need to show that ( \frac {(k-1)!}{48} ) is divisible by 2 and (k_1 ). Since we showed ( k \geq 9 ) in claim 1, and then showed k is even; this implies k is at least 10 or more.
( \frac {(k-1)!}{48} \ = \frac {1 \times 2 \times 3 \times \cdots \times 9 \times \cdots \times (k-1)} {48} \ = 1 \times 5 \times 3 \times 7 \times 8 \times 9 \times \cdots \times (k-1) )
For k = 10, we see that (1 \times 5 \times 3 \times 7 \times 8 \times 9 ) is easily divisible by 10 (5 exists and 8 supplies one 2).
For k = 12, we see that (1 \times 5 \times 3 \times 7 \times 8 \times 9 \times 10 \times 11 ) is also divisible by 12 (3 exists and 8 supplies one 4).
From k=14 onwards, this will be trivially true because: ( k/2 \geq 7 ) will be present in the product (as it is not canceled off by division by 48), and another 2 will be supplied by the first even number that follows k/2. (There will be one such even number, because ( \frac{k}{2} + 2 \leq k-1) as ( 6 \leq k ) (in fact it 6 is much less than k as we have reached 14 or more).
From Equation 1, we have:
( (k-1)! = 48\times ( k^{m-1} + m \cdot k^{m-2} + {m \choose 2} k^{m-3} + \cdots + {m \choose {m-2}} k + {m \choose {m-1} } ) )
This implies ( \frac {(k-1)!}{48} = k \times (k^{m-2} + m \cdot k^{m-3} + \cdots + {m \choose {m-2}} ) + m )
Therefore ( \frac {(k-1)!}{48} - k \times (k^{m-2} + m \cdot k^{m-3} + \cdots + {m \choose {m-2}} ) = m )
From Claim 3 we know k divides ( \frac {(k-1)!}{48} ). Hence the LHS of the above equation is divisible by k. This implies RHS is also divisible by k. Hence k divides m.
Let ( m = Qk ) where Q could be 1.
Use A. M. - G.M. Inequality to establish:
$$ \frac { 1+ 2 + \cdots + k}{k} > (1 \times 2 \times 3 \times \cdots \times k ) ^{\frac{1}{k}} $$
Strictly inequality holds, as the numbers are not all equal. This implies:
$$ \frac { (k+1)^k}{2^k} > k! $$
Therefore ( \frac { (k+1)^k}{2^k} + 48 > k! + 48 ) ... (2)
Clearly ( 48 (k+1)^k > \frac { (k+1)^k}{2^k} + 48 ) ... (3).
Why? Divide both sides by ( (k+1)^k ). Then you have ( 48 > \frac { 1}{2^k} + \frac{48}{(k+1)^k} ). k is 9 or more, right hand sides is (much) smaller than 1. Hence 48 is certainly greater than RHS.
Combining (2) and (3) we have ( 48 (k+1)^k > k! + 48 ). Since k divides m we have ( 48 (k+1)^m > 48 (k+1)^k > k! + 48 ).
Even before we start working on this problem, we should check some basic cases (where k and m are small; that is 0, 1, 2 etc.).
Also see

Publishing original scientific research in a peer-reviewed journal is an achievement many scholars work toward for years. Achieving this feat early in one's academic journey requires exceptional intellectual rigor, creativity, and persistence. Aaditya Punatar and Muskan Sadana have reached this remarkable milestone, successfully publishing their collaborative research paper in the prestigious Journal of Emerging Investigators […]

Her project on quantum-accelerated Steiner trees for 15-minute cities earned a place among the finalists at the world's largest pre-college STEM competition. Prishaa Shrimali of Marymount School of New York, was selected as a Grand Finalist at the 2026 Regeneron International Science and Engineering Fair (ISEF) in Phoenix, Arizona. Her project was Optimizing Urban Accessibility: […]

In 2026, the following Cheenta students have been successful for Indian Statistical Institute's M.Stat Entrance. They ranked within the first 50 in the entire country in these entrances. I.S.I. M.Stat Entrance

In 2026, the following Cheenta students have been successful for Indian Statistical Institute's B.Stat Entrance and Chennai Mathematical Institute's B.Sc. Math Entrance. They ranked within the first 200 in the entire country in these entrances. Most of these students attended the problem solving workshops regularly, which happen 5 days every week. CMI B.Sc. Math Entrance […]