(I sometimes get depressed because so many of Asimov's books, especially his essay collections and early books on science, are not in-print and haven't been in my lifetime. Most of my books by Asimov have been purchased at used bookstores, and almost all of them are thirty years old or more.)
One of the books that I found myself reading was Only a Trillion (Jenkins' review here, in which he mentions in passing what I will spend some energy on). It's one of Asimov's very early (1957) science essay collections, and chronologically number 24 in Asimov's list of published books (the fact that this is one of his "early" works and is only his 24th is amazing in itself -- many authors could produce 24 books in a lifetime and be considered "prolific") -- the book deals primarily with matters of chemistry and a bit of atomic physics, concentrating on the distribution of the various types of atoms in the Earth's crust, and in the universe in general.
I won't make detailed commentary here about the nature of Asimov's science, as I'm not (yet) educated enough in the modern-day nature of chemistry to comment directly as to how much of what he says is no longer considered accurate. But there is one essay, "The Unblind Workings of Chance", that seems so relevant today that I almost pinched myself when I read it.
The question is, how did the atoms in those small molecules [here he refers to water, carbon dioxide, and ammonia in particular, but to "atoms that make up nucleoproteins" in general] manage to place themselves just so in order that the first nucleoprotein molecule might be formed? Once one nucleoprotein molecule exists, it can guide the formation of others. But how was the first one formed?
Could it have been the result of the blind workings of chance? Could the atoms have just happened to bump one another and stuck together in the right pattern -- just by chance, after a billion years of random trying?
...I will only say that the chances are more infinitesimal than you or I can imagine. So infinitesimal, that if the known universe were crammed with nothing but people and each person performed the test twenty times a second (a hundred times a second, a thousand times, what's the difference!) for a billion years (or a trillion or a trillion trillion), the chances of any one of those humans coming up with a perfect nucleoprotein pattern at any instant in all that time is still infinitesimal.
That kind of thing was pointed out, rather triumphantly, by Lecomte du Nouy, in a book named Human Destiny, published in 1947. His point of view was that this proved it to be completely unreasonable to suppose that life had originiated by the blind workings of change, and that therefore there must have been some directing intelligence behind its origin. (Asimov, p. 102-103)
The de Nouy book is on Amazon here, and I must admit I haven't read it. I'll see if I can't find it in one of my local libraries, though, because this argument sounds like something I've heard before, doesn't it? Why, it's that horrid old "Argument from Design" dressed up in the language of "irreducible complexity" and "tornado in a junkyard", being refuted by Asimov (a biochemist, not a biologist) decades before "Intelligent Design" was a glimmer in anyone's eye, hell, three years before even the foundational book of the original scientific creationism movement, The Genesis Flood was published. (Isn't it ironic that The Genesis Flood is still in print, while the Asimov book is not?)
Asimov triumphantly defeats the hoary old probability argument just a few paragraphs later. The whole essay is worth a read, and goes into much more detail, but:
We have no right to assume they [hydrogen and oxygen molecules] combine at random, and, as a matter of fact, they don't. The chemical properties of the hydrogen and oxygen atoms are such that the combination H-O-H is the only one that has any reasonable probability at all, so it is the only combination formed.
So one does not and must not ask: what are the chances that a nucleoprotein molecule is built up through the blind workings of chnace?
One must ask: what are the chances that a nucleoprotein molecule is built up through the known laws of physics and chemistry -- the very definitely unblind workings of chance? (Asimov, p. 104-105)
He then goes on to describe in some detail how individual atoms are made into monomers, which can then (with introduction of energy) be converted via peptide bonds into proteins. Basically laying to rest a whole sheaf of those ignorant probability calculations that we see so often from anti-evolutionists everywhere. Many large libraries should still have a copy of the book, for anyone interesting in reading it.
Only a Trillion, Asimov, Isaac. 1957. Abelard-Schuman Ltd.