it was eventually found that the cause of these striking colors was the precipitation of the gold metal in tiny particles that were all roughly the same size, that is to say, the particles were "monodisperse"

suspensions of tiny particles that are all about the same size display unusual optical properties in bulk, like optical absorption in sharply defined bands. strong fluorescence is another property shown by such particles

basically you're seeing quantum mechanical behavior that would be obscured in a suspension of particles of different sizes. I don't know all the science I admit

why does this curious behavior occur? why doesn't the precipitated gold all just stick together in large particles? the answer is that sometimes it does, and you just get dark-colored crud settling out of solution. it requires carefully controlled circumstances to get a suspension of particles that are all the same size

but these conditions are fairly easy to achieve as it turns out, and it's fairly easy to tell when the reaction works right. you get beautifully colored pink and purple suspensions of gold nanoparticles. I've carried out this reaction myself!

the key is the presence of substances in solution that adsorb strongly to the surface of the gold metal, coating the tiny particle of metal as it forms with a molecular monolayer that protects it from further reaction. these are called "capping agents"

many many substances adsorb to gold in this fashion. sulfur compounds are especially keen to adsorb to gold, but nitrogenous and oxygenated organic compounds also work. citric acid is a common capping agent

and it turns out that amino acids also can work for this purpose

there's been a loooooot of work done in finding and fine-tuning sets of conditions of chemical precipitation of gold that will yield suspensions of particles of different sizes and shapes

not just gold, either

silver metal can be precipitated in like fashion, in monodisperse nanoparticles. so, with stronger chemical reducing agents, can copper, nickel, and other metals

indeed, a huge range of insoluble substances that can be prepared by chemical precipitation, not just metals, have been prepared in the form of monodisperse nanoparticles. metal oxides, sulfides, selenides, all sorts of things

now

the optical properties of these monodisperse nanoparticle dispersions is greatly dependent on their size, shape, _and_ their surface chemistry as it turns out

like, for example, one can use the gold nanoparticle's high affinity for sulfur compounds as a way to attach a chemical "receptor" to its surface. like, you can put a thiol "tail" on something like a crown ether, so you can tether the crown ether to the gold particle

and then you'll have gold particles that are capable of recognizing a specific chemical, and it'll show up as a change in the particles' optical properties!

so I found myself thinking, what would happen if you capped gold nanoparticles with a _mixture_ of amino acids?

one imagines the various amino acids jostling for positions around the gold particles as they form, coordinating with the gold surface and with each other

each mixture of amino acids would likely produce a suspension of gold nanoparticles that had unique surface geometry and chemistry

the particles would have amino acids adsorbed to their surfaces in different arrangements, mimicking the differing geometries you get from amino acids covalently bonded with one another in proteins

well then, why not prepare a number of suspensions of nanoparticles of gold (or other substances, but gold is so convenient and it's been the most heavily used for this sort of thing) capped with different mixtures of amino acids

then see which one of these nanoparticle preparations shows the greatest affinity for some target molecule, something you want to detect and recognize, like, estradiol, say

then prepare a new set of nanoparticle suspensions using conditions that were slight modifications of the conditions that made the most promising nanoparticles in the first round of tests

see what I'm getting at? I think it might be possible to make nanoparticles that are tuned to recognize a particular substance through an evolutionary process

and the beauty of this idea is that it's very low tech. it's chemistry I can do in my kitchen, practically

a huge range of amino acids are available off the shelf as B.S. "dietary supplements". I'd have to purify the shit out of them, of course, before using them in an actual experiment, but at least they're purchasable

gold chloride solution can be bought on eBay, and I don't have to work with gold anyway. metal sulfides are an option

so when I have time, I think I'm gonna start working on this idea. I love it because I can start with really crude, see-Spot-run chemistry. looking for really basic property changes. does it change color? does it fluoresce? stuff like that

but I can refine things too, use spectrophotochemical detection to look for subtle shifts in absorption bands, work towards proper fluorescent spectrometry too

maybe even start to think about methods for determine particle sizes and aggregation more directly

@kara we don't have a full handle around all this yet but it sounds super fucking rad

@Fidgetcetera thanks ^^

of relevance to my idea, amino-acid-modified ZnS nanoparticles: https://www.sciencedirect.com/science/article/pii/S0167577X10007974