Multiplicative combination algorithm

Problem:

Given a number n, is there an efficient algorithm to obtain a list of 2-combinations from the set {1...n}, sorted by the value of the product of the combination?

I need this in order to determine the largest product of two *-digit numbers that satisfies a certain condition. If the list is unsorted, I must first determine all combinations that satisfy the condition, then iterate through those to find the combination with the largest product, which is inefficient.

As an example, given n = 3, the combinations possible are:

Combination:      Product:
   3, 3              9
   3, 2              6
   3, 1              3
   2, 2              4
   2, 1              2
   1, 1              1

Sorted by the value of the product in descending order, this is:

Combination:      Product:
   3, 3              9
   2, 3              6
   2, 2              4
   1, 3              3
   1, 2              2
   1, 1              1

Extra background:

I just solved a Project Euler question regarding finding the largest palindromic number that is a product of two 3 digit numbers. My approach was to iterate downward from 999 (the largest 3 digit number) with two factors and find the product of each combination, additionally checking whether the number was palindromic:

def maxpal():
    for i in reversed(range(100,1000)):

        # Since we only want unique combinations, we only
        # need to iterate up to i

        for j in reversed(range(100,i)):   
            if str(i*j) == str(i*j)[::-1]:
                yield i*j

print max(maxpal())

Note that the first list in the example iterates over factors in exactly the same order as this code. My initial assumption was that since I was iterating downwards, the first palindrome I found would be the largest one. This is clearly not the case, since j iterates all the way to 100 before i