Additive problems with almost prime squares (Q6093395)

A conjecture made by Hardy and Littlewood asserts that the number of representations of an integer \(n\) in the form \(n=p+x_1^2+x_2^2\) in primes \(p\) and non-zero integers \(x_1\), \(x_2\) as \(n\to\infty\) is asymptotic to \(\pi\mathfrak{S}(n)\mathrm{Li}(n)\), where \(\mathfrak{S}(n)\) is determined explicitly as the products running over primes \(p|n\). In the paper under review, the authors continue studying this conjecture by proving that there exists a constant \(C>0\) such that every sufficiently large integer \(n\equiv 1, 3\pmod{6}\) can be represented in the form \(n=p+x_1^2+x_2^2\), where \(p\) is a prime and \(x_1\), \(x_2\) are integers all of whose prime factors are greater than \(n^{1/C}\). Moreover, they show that the number of such representations is \(\gg\mathfrak{S}(n)n/\log^3 n\). In particular, every sufficiently large integer \(n\equiv 1, 3\pmod{6}\) can be written as the sum of a prime and two squares of integers with no more than \(C\) prime factors. Similarly, the authors treat representations of shifted primes \(p-1\) as sums of two almost prime squares, by showing that there exists a constant \(C>0\) such that the number of solutions to the equation \(p=x_1^2+x_2^2+1\) in primes \(p\leq x\) and \(x_1\), \(x_2\) all of whose prime factors are greater than \(p^{1/C}\) is \(\gg x/\log^3 x\). In particular, there are infinitely many primes shifted by one that can be written as two squares of almost primes. Finally, they consider the equation \(n=x_1^2+x_2^2+m\) concerning smooth numbers. The proofs of these arithmetic results combine a variety of different methods from different fields, including algebraic structure theory of orders in quadratic number fields, the full machinery of automorphic forms including some higher rank \(L\)-functions, and usual methods of analytic number theory.