Precious metals have been lent out at interest for at least 4500 years. It's called capitalism, the old name was usury.
Imagine that 4500 years ago someone very poor put aside as an investment the tiniest grain of silver she could scrape off, and lent it out at 2% annual interest. And imagine that up to the present day, her descendants didn't draw on that store, but kept reinvesting it with the profits in the same way, at the same rate of interest.
Of course, over the great majority of that time, no-one would have lent at anything like so low a rate. But let's stick with my scenario.
By now, that family's hoard of gold would have grown so large, that to create it we would have to turn every single atom in the universe, (we can't observe it any farther even in principle, no matter how big our telescopes) into gold, and then add as many other universes as there are atoms in this one, and turn them all to gold, and still we wouldn't be done.
You have got the maths a bit wrong . 1.02 raised to the power  4500 is  about  5 x 10^38.  That will raise 1g of silver to 5 x 10^35 kg of silver.  The mass of the sum is about 2 x 10^30 kg. Converting the 70% hydrogen atoms (atomic mass 1) and 28% helium (atomic mass 4) to silver (atomic mass 107) gives about 1.26 x 10^32 kg.  You would need about 4000 suns turned into silver.  There are about  10^11 stars in our galaxy and about 10^11 galaxies in the observable universe.  Assuming our star is about average mass you are out by a factor of about 2.5 x 10^18.
Good catch! I was hoping someone would correct that.
I should have let her lend at 7%. Still many times less than the lowest rate you could borrow at for most of the time elapsed since then.
Imagine that 4500 years ago someone very poor put aside as an investment the tiniest grain of silver she could scrape off, and lent it out at 2% annual interest. And imagine that up to the present day, her descendants didn't draw on that store, but kept reinvesting it with the profits in the same way, at the same rate of interest.
Let me suggest another way to think about this.

Ultimately, lending money and getting back more can be thought of as supported by the growth of the economy. But only in recent years has the economy grown as fast as 2% per year. 4500 years ago the economy was dominated by farming. As argued in economist Robin Hanson's analysis, the economy at that time took about 900 years to double in size, compared to about 20 years today. An appropriate interest rate at that time would have been something like 0.08%. Today the equivalent would be about 3-4%.

Over the time period since 2500 BC, world product has increased by roughly a factor of 50,000. That is a realistic estimate of how much an investment should be expected to grow in that interval of 4500 years. It's a substantial amount but not the extravagant figure you calculated. And it corresponds to using the interest rates above, 0.08% for most of the period, ramping up to 3-4% today.

I think it's more useful to consider what profit you could make from real investments, not what rate of interest we might declare acceptable in our minds. The first is a very important historical fact, the second an abstraction.
Even today no investor will settle for 3-4%, though now you can get that with virtually no risk. In ancient times all commercial enterprises, including moneylending, expected huge rates of return, and were correspondingly risky by our standards. From such considerations, Keynes foresaw a future time when real interest rates would reach zero.
If we want to know whether any activity, including interest bearing loan, is sustainable over some time period, we need to consider its real conditions, not those we would prefer in retrospect.