Let (x,y,z) be a solution of the problem. Then, notice that x,y are odd, hence z is also odd, and co-prime with xy. The equation can be written successively x8+2x4y4+y8=4z4, or (x4−y4)2=4z4−4x4y4, or z4−(xy)4=(2x4−y4)2.
We prove that the equation a4−b4=c2 (∗)), has no solutions (a,b,c), where a,b,c are positive integers and a,b are co-prime. Assume the contrary to be true. Consider (a0,b0,c0)∈N3 the solution of the above equation with a0 minimum.
If b0 is odd, from b04+c02=a04 we deduce that there exist positive integers m,n,m>n, of different parities, such that a0=m2+n2, b02=m2−n2, and c0=2mn. It follows that m4−n4=(a0b0)2, i.e. the triple (m,n,ab) is a solution of equation (∗), with m,n,ab>0 and (m,n)=1. This contradicts the minimality of a0.
If b is even, there exist m,n∈N, m>n, such that a2=m2+n2, b2=2mn, and c=m2−n2. We may assume that m is even, n is odd. From b2=2mn it follows that 2m=p2, n=q2, i.e. m=2p12, n=q2, with q odd, (p1,q)=1. Thus, a2=(2p12)2+(q2)2, which means that there exist r,s∈N, r>s, such that a=r2+s2, 2p12=2rs, q2=r2−s2. From (r,s)=1 and rs=p12 it follows that r=u2, s=v2, and (u,v)=1. This means that u4−v4=q2, i.e. (u,v,q) is a solution of equation (∗) with u,v,q∈N, (u,v)=1, and u<a0, which contradicts the choice of a0.
Getting back to the equation
z4−(xy)4=(2x4−y4)2,
from the above we can see that the only solutions it may have are with x4−y4=0, i.e. with x=y. But, as (x,y)=1, it follows that one must have x=y=1, and finally that x=y=z=1 is the only solution.