1. **Prove that a,b,c contain the same set of prime factors.**
Assume, without loss of generality, that p is a prime factor of a but not of b. Then p∣a and p∤b.
Since p∣a, we have p∣ap−1−1 by Fermat's Little Theorem. Therefore, p∣(ap−1−1)(bp−1−1)(cp−1−1).
However, since p∤b, p∤bp−1−1. Thus, p∤(ap−1−1)(bp−1−1)(cp−1−1)+1.
This implies p∤((ap−1−1)(bp−1−1)(cp−1−1)+1)3, but p∣(abc)p−1, which is a contradiction. Hence, a,b,c must contain the same set of prime factors.
2. **Prove that the exponent of each prime in a,b,c is the same.**
Let x,y,z be the maximum exponents of a prime p in a,b,c respectively. Assume without loss of generality that x≤y≤z.
Consider the expression (ab+bc+ca−a−b−c)3. Since abc∣((a−1)(b−1)(c−1)+1)3, we have abc∣(ab+bc+ca−a−b−c)3.
- Case 1: y and z are strictly greater than x.
In this case, px∣(ab+bc+ca−a−b−c) and p3x∣(ab+bc+ca−a−b−c)3. However, px+y+z∣abc and since x+y+z>3x, this is a contradiction.
- Case 2: One of y and z is equal to x.
Without loss of generality, assume y=x. We need to prove that z=x=y.
If p∤pxb+c, then proceeding as in Case 1 gives the desired result.
If p∣pxb+c, then (abc)2∣((a2−1)(b2−1)(c2−1)+1)3. This implies (abc)2∣(a2b2+b2c2+c2a2−a2−b2−c2)3.
If p∤p2xb2+c2, we have the desired result. Otherwise, p∣p2xb2+c2.
This implies p∣p2x(b+c)2−p2xb2+c2⇒p∣p2x2bc, which is impossible for a prime p not equal to 2. Hence, we conclude that all primes except 2 occur in the same exponents.
3. **Handle the case when p=2.**
Case 1 yields that 2 also occurs in the same exponents in two of a,b,c, say a and b. Hence, we should have a=b and c=2r⋅a.
Plugging this into the original equation gives:
2rna3n∣(2⋅2rna2n+a2n−2rnan−2an)3
2rna3n∣a3n(2⋅2rnan+an−2rn−2)3
2rn∣(2⋅2rnan+an−2rn−2)3
Since 2∣a, we can choose a sufficiently large n such that (2⋅2rnan+an−2rn−2)3 is fully divisible by 23, which is a contradiction since we have 2rn∣23, implying r=0.
Therefore, a=b=c.
■