I haven't posted here in a while. I'm going to try posting something blank and building it up over time, like a wiki article. It surely won't ever get posted if I write it in a text editor. So here we go: Space Food. Part 1/ 3. It doesn't have any pictures yet. You might want to wait to read it till I have pictures.
Mars is a dark, frozen, irradiated vacuum covered in poisonous soil, and I want to go there. If other people go there, that would be cool too. Let's talk about how to make food on Mars, because people need food, for now at least, until we figure out whole brain emulation.
This post looks at the manufacture of carbohydrates, amino acids, fatty acids, vitamins, and bio-available sources of essential chemical elements by three methods (physicochemical, microbial, and botanical), with some consideration for physicochemical, microbial, and botanical recycling of human waste material into those nutrients. The post is mainly a study of possibility; of what nutrients we can currently manufacture in which ways if we want to.
Physicochemical manufacturing:
It would be great if we could recycle all our human waste products into food by doing a little chemistry, but we can't just yet. Food molecules are built in nature by enzymatic molecular assembly, but human chemistry isn't that advanced yet, and mostly consists of the less precise methodology of putting things into jars along-side poisonous corrosive catalysts and then heating them. We can make a few foods non-enzymatically. Let's look at carbohydrates first. For variety, I will sometimes call a non-enzymatic synthesis a total synthesis or a synthesis from scratch or a purely chemical synthesis or an abiotic synthesis.
In practice, for space missions, it makes sense to chemically synthesize carbohydrates and maybe essential fatty acids from recycled water and CO2, and to then get the rest of your amino acids, vitamins, and trace chemical elements from plants and rapidly growing edible microorganisms like microalgae and yeast. But I'm going to see if we can make all of it chemically. That means making like 50 complicated chemicals, instead of growing four plants, some mushrooms, and a tank of green goop, but I'm curious. It's very hard to do, but I think we can make most of the essential nutrients non-enzymatically. So that's the layout of the essay; a huge section on physicochemical manufacture, and then two small sections on microbial manufacture and plant manufacture, which are probably the more important sections for the practicality of space diets in closed ecological life-support systems, but they're so much easier, that there's less to write about. (Part II) (Part III)
:: Chemical Carbs
There is no good standard definition for carbohydrates. Organic chemists use the term a little differently from nutritionists, and neither of them has a very good definition either. In this essay, "carbohydrates" will refer to edible inessential oxyhydrocarbons that also aren't fatty acids. This includes sugars and starches, as well as some other things like ethanol, but excludes indigestible dietary fibers like cellulose. Carbohydrates, by any definition, aren't essential for human metabolism, but it's probably a good idea to get some of our calories from them, just in case high protein diets (and high RNA diets, which we'll talk about more in the microbial food section) turn out to stress the kidneys or result in accumulation of uric acid salts in the body. The carbohydrates that we're good at making non-enzymatically are alcohols of various kinds: glycerol, the semi-edible ethanol, propylene glycol, and 1,3-propanediol. Propylene glycol is also known as 1,2-propanediol, but it'll use the first name to distinguish it from the 1,3 isomer.
How are they made? Ethanol is made by hydrating ethylene (i.e. water is a reactant) or hydrogenating CO2 (i.e. H2 is a reactant). 1,3-Propanediol is made by hydroformylating ethylene oxide. Hydroformylation basically means reacting something with CO and H2 gasses, the mixture of which, syngas, is easily made by the steam reformation of methane. Propylene glycol is made in a few ways, including hydrolysis of propylene oxide. Glycerol is also made a few ways. One good one is hydrolyzing epichlorohydrin, made by the base-catalyzed cyclization of dichlorohydrin, made by the oxidation of allyl chloride, made by chlorinating propylene. Another route to glycerol is hydrolyzing allyl alcohol, made by high-temperature isomerization of propylene oxide, made by oxidizing propylene.
Just how digestible are these? Toxicology measures oral toxicity in terms of median lethal doses, i.e. the dose that will kill half of a studied population of creatures, often rats, mice, or rabbits. These are usually significantly correlated with humans, but it's also not unusual for two species to have median lethal doses for a compound that differ by a few grams of substance / kilogram body weight. Two compounds known to be more toxic to humans than to laboratory animals are methanol and ethylene glycol (not propylene glycol, which was one of the food alcohols listed above). But the LD50 numbers are usually a really good start for classifying substances by oral toxicity in humans. Lots of material safety data sheets list glycerol's LD50 at 12,600 and that's wrong. I think the figure comes from a 1945 report by the Federation of American Societies for Experimental Biology. Most studies since then put it above 25,000 mg/kg. I'm calling it 27,200 mg/kg, as reported in "Comparative toxicity of synthetic and natural glycerin" (Hine, 1953). You can just think of it as 27 g/kg. I'll try to use oral LD50s based on rats everywhere in this essay. The LD50 for ethanol is 9,000 mg/kg, and for propylene glycol it's 20,000 mg/kg. Finally, 1,3-propanediol clocks in at 15,670 mg/kg. Compare glucose (oral LD50: 25,800 mg/kg). I've read that some other diols have low toxicity: 1,5-pentanediol, 1,3-octanediol, and 1,10-decanediol all supposedly have LD50s above 10,000 mg/kg. I'm a little suspicious though. I think we should just focus on the smaller diols, which are better studied, have higher LD50s, and are probably easier to produce.
I should note that these are all acute doses, as in a short term investigation. There's no guarantee in these numbers that the substance won't have chronic effects, such as through mutagenicity. This is just a preliminary identification of potential food compounds. Anyway, I'd call a chemical with an LD50 of 10 g/kg edible as a carbohydrate source, and anything below 10 g/kg I would call semi-edible as a carbohydrate at best. But that's my own rule of thumb for including them here, not an industry standard. Edible flavorings will be treated later separately; they make up a much smaller portion of the diet, so their LD50s can be lower.
Glycerol (in fats and oils) and ethanol are normal components of human diets. We know the enzymes involved in their digestion and the metabolic products generated. Glycerol in particular can be digested by more than one pathway, which makes it particularly attractive if you're worried about people with alcohol dehydrogenase enzyme deficiency. But what about propylene glycol and 1,3-propanediol? The LD50s tell us that they're non-toxic in acute doses, but are they digestible - can we get out calories from them?
Propylene glycol is used as antifreeze and e-cigarette liquid, but it's also a food additive. It is significantly less toxic than ethanol, and that's in humans too, not just in rats. The other famous glycol antifreeze, ethylene glycol, is a terrible poison that will kill you three different ways at once, but propylene glycol is a normal non-poisonous carbohydrate. Propylene glycol is metabolized in the liver by the same machinery as alcohol. In particular, alcohol dehydrogenase turns it into lactaldehyde, and then aldehyde dehydrogenase turns that into lactic acid. So far as I can tell, toxic doses aren't enjoyable, the way they sometimes are with ethanol. The last carbohydrate on the list, 1,3-propanediol, is an isomer of propylene glycol. It's occasionally used as a food additive, but less so. I imagine it's digested by the same enzymes as propylene glycol, but maybe the end result is something a little different from lactic acid. The LD50 of 1,3-propanediol is also a little lower than propylene glycol, so unless the synthesis is way easier or its metabolic byproduct is interesting in some way, I'd just stick with propylene glycol.
After trying to figure out the metabolic products by hand, my best guess is 1,3-propanediol is dehydrogenated at one end to form 4-hydroxybutyraldehyde, which is then oxidized to form 4-hydroxybutanoic acid, better known to drug users as GHB. That can't be right, can it? I think I would have heard of that. Maybe instead alcohol dehydrogenase takes 1,3-propanediol and dehydrogenates it at both ends, forming succinaldehyde, which is then oxidized to form succinic acid. Succinic acid doesn't have a particularly high LD50 (2,260 mg/kg), but it's an intermediate of the citric acid cycle, so we know where it's going. I wonder if my banana friend would try consuming 1,3-propanediol and report to me whether she experienced any symptoms like those of GHB use.
Lactic acid is another carb that we can make from scratch. It's a chiral compound, having left-handed (levo-) and right-handed (dextro-) versions, and the total synthesis that I've seen produces a mixture of both enantiomers, a racemate. This is a slight problem because we can only digest L-lactic acid, while D-lactic acid is toxic (see "D-Lactic Acid as a Metabolite: Toxicology, Diagnosis, and Detection"). A material safety data sheet lists the racemic mixture at (LD50 = 3,543 mg/kg), which is not great, and another lists the purified L-Lactic acid at 3,730 mg/kg, and why isn't that figure appreciably higher than the other one? Sad. The main reason I could see to make lactic acid non-enzymatically for extraterrestrial diets is that it can be used to make lactylates. Lactylates are very non-toxic food additives and I'll probably talk about them in the section on fatty acids and their esters, rather than here with the carbohydrates. The LD50 of sodium stearoyl lactylate was once reported > 25,000 mg/kg, which puts it up there with glycerol and glucose. Pretty impressive. Lactic acid itself is made from acetaldehyde and hydrogen cyanide, HCN. In particular, acetaldehyde and HCN form a cyanohydrin called lactonitrile, which is then hydrolyzed in sulfuric acid to make racemic lactic acid. I don't know if we'd have to isolate the levo enantiomer to make lactylates. If we do, there are some methods for separating and purifying L-lactic acid from the racemic mixture (see "A Technical Overview on Alternative Methods for the Separation of Racemic Carboxylic Acids: Lactic Acid as a Compound"), but if you're throwing away half of your synthesized product at best, that's not very efficient. Maybe we don't need lactylates.
Are there any other carbs that we can make non-enzymatically that provide calories and have low toxicity and might be metabolized by different enzymes than the above, in case we want to diversify our diet for safety? I'm not sure, but I'd really like to find something that isn't an alcohol or polyol. Saccharin is a synthetic sweetener that we can make from scratch, but it doesn't have digestible calories (also it has nitrogen and sulfur, so I wouldn't call it a carb anyway). Let's start by finding something non-toxic enough to be worth making and then I'll see if we can actually make it. How about organic acids found in foods? L-Ascorbic acid is vitamin C, and it has an oral LD50 of 11,900 mg/kg, which is good enough to also be a carbohydrate! But I'm not sure we can make it in any decent quantities. We can make it from glucose, whose difficult synthesis is discussed a little below. I discuss other possible synthesis routes in the vitamin section. It's not really promising. What other food acids are there? Citric acid's oral LD50 is only 3,000 mg/kg in rats, which may be fine for a flavoring, but it's too low for a carb. Malic acid is similar (I see an MSDS listing 1,600 to 3,200 mg/kg, but I'm not sure if that's for the dextro enantiomer of for the racemic mixture of dextro- and levo-). I don't see good numbers for tartaric acid, but it's probably just as low. Sodium tartrate is only 4,360 mg/kg in rabbits. For fumaric acid, I see one source listing LD50 at 6,800 mg/kg, and another one listing 9,300 mg/kg for female rats and 10,700 mg/kg for male rats. To me, that sounds semi-edible. Kind of weird though, because it's closely related to malic and maleic acid, which have low LD50s. The median lethal acute oral dose in rats for a certain salt of oxalic acid, sodium oxalate, is often reported as 11,160 mg/kg, which sounds good, but I think the number is either wrong or irrelevant. The LD50s are much lower in other animals and we know that oxalate accumulation causes kidney failure in humans. How about other acids? Pyruvic acid is important in cellular metabolism, which sounds like a sign of digestibility, but a salt of it, sodium pyruvate, has a reported LD50 of just 5,600 mg/kg. Pyruvic acid is a keto acid (it has a carboxyl acid group and a ketone group). Let's look up some other famous keto acids. Alpha-ketoglutaric acid is a keto acid made by removing an amine group from glutamic acid. It's part of the body's citric acid cycle, and it's a partial antidote for cyanide poisoning. Its LD50 is 7,100 mg/kg. Cool, but not cool enough. Levulinic acid has LD50 = 1,850 mg/kg. No thank you. Another keto acid, acetoacetic acid, maybe though? For "acetoacetic acid ethyl ester", better known as ethyl acetoacetate, I see a single paper claiming an oral LD50 range of 3,980 to 12,000 mg/kg, which seems quite a wide margin. Maybe it's a small sample size confidence interval thing? And then I see an MSDS listing just the low number, 3,980. But then another site has acetoacetic acid n-Butyl ester (aka butyl 3-oxobutanoate) at 11,260 mg/kg. Two of those three numbers are big! So maybe acetoacetic acid esters are edible? Potentially cool. Esters are usually made by combining an organic acid with an alcohol (n-Butanol in this case). You might be wondering if alcohol esters of the previous organic acids have higher median lethal doses. I'll spare you all the numbers; the LD50s of organic acid esters are generally a little higher than the corresponding acids, but not higher by enough to turn somewhat non-toxic organic acids into highly non-toxic carbohydrates.
What other chemicals might be edible? Acetoacetic in that last ester is not just structurally a keto acid, it's also functionally a ketone body in animal metabolism, which we'll talk about more in the fatty acid section. Acetone is another ketone body, but not a keto acid, and its oral LD50 in rats is 5,800 mg/kg, which is kind of impressive for something I think of as an industrial solvent. It's a metabolic intermediate, so maybe I shouldn't be surprised, but I am. It's still not food, but it's less toxic than citric acid, at least in rats. Back on alcohols for a second, there's an artificial carbohydrate called 1,3-butanediol which is not a ketone body or a keto acid, but it's metabolized (in the liver, by alcohol dehydrogenase) to the ketone body beta-hydroxybutyric acid, which the body can use for energy as a glucose alternative. In addition to being digestible, 1,3-butanediol is non-toxic (LD50 18,610 mg/kg). That's a good high number, but you eat carbohydrates to balance out ketogenesis, not to cause it. And if it's being metabolized by the same enzymes as ethanol and propylene glycol, then it's not adding the kind of diversity to the chemical menu in the way that I'm searching for. Fun fact about 1,3-butanediol, one of its isomers is a moderately famous drug; 1,4-butanediol is sold on the street, and the usual alcohol-digesting enzymes in the liver convert it into gamma-hydroxybutyric acid, rather than beta-hydroxybutyric acid, and the gamma- version (GHB) is a potent anesthetic and euphoriant.
| Chemical | LD50 |
|---|---|
| Glycerol | 27,200 mg/kg |
| Glucose | 25,800 mg/kg |
| Propylene Glycol | 20,000 mg/kg |
| 1,3-Butanediol | 18,610 mg/kg |
| 1,3-Propanediol | 15,670 mg/kg |
| L-Ascorbic Acid | 11,900 mg/kg |
| Butyl Acetoacetate | 11,260 mg/kg |
| Ethanol | 9,000 mg/kg |
| Alpha-Ketoglutaric Acid | 7,100 mg/kg |
| Fumaric Acid | 6,800 mg/kg, 9,300 mg/kg (female), 10,700 mg/kg (male) |
| Acetone | 5,800 mg/kg |
| Sodium Pyruvate | 5,600 mg/kg |
| Sodium Tartrate | 4,360 mg/kg (rabbit) |
| Ethyl Acetoacetate | 3,980 to 12,000 mg/kg (?) |
| L-Lactic Acid | 3,730 mg/kg |
| Citric Acid | 3,000 mg/kg |
| Levulinic Acid | 1,850 mg/kg |
In short, I'd say that glycerol is really great for a space carbohydrate, even though it might not be the easiest of the food alcohols to make. Propylene glycol also has significant merit, although I worry about the possibility that some people (an unlucky 50% of east Asians) might be mostly unable to digest it, because, while they have a little alcohol dehydrogenase enzyme in their intracellular fluid (their cytosol), they don't have the second alcohol dehydrogenase enzyme that most people have in their mitochondria. If a significant portion of people can't drink propylene glycol, then that's not a great carbohydrate for space. Also, can we just fix our mitochondria? That would be cool. One day when I'm a gene therapy research czar, I'll get right on that. Anyway, if you want to diversify your carbohydrates for metabolic reasons, I'd look into acetoacetic acid (since two of its esters have LD50s over 10 mg/kg (depending on the study)) or fumaric acid (about which we have similar hope and similar uncertainty).
I should find a more efficient way to look these up. Like a list of all the metabolic intermediates and food additives that are oxyhydrocarbons. I saw a website that listed an LD50 of 26,000 mg/kg in mice for a vitamin A derivative called all-trans-retinoic acid. It's more commonly known as isotretinoin, and it's sold under the name Accutane for treating acne. I don't believe the number, but I got very briefly excited about the prospect of eating large amounts of acne cream in space. Ooh, the Hazardous Substances Data Bank is available for download. I hope it's in an easily read format.
"Sugars! What about sugars?!" Monosaccharides and disaccharides (better known as sugars) and polysaccharides (including starch, pectin, and natural gums) are the usual things that we think of as carbohydrates, but we're not good at making them from scratch. Let's talk about them anyway. Between the two classes, there are also oligosaccharides (molecules made of a few monosaccharide molecules). They're not as famous, and mostly not digestible, but we'll touch on them too.
What are sugars, really? We just don't know. Sugars are sweet tasting and water-soluble. That's actually a common definition, though it doesn't tell you much about the molecule. We could have done better: we could have said that carbohydrates have the form C_n (H2O)_m, as the name suggests, and sugars have n = m. Most of the sugars do - glucose, fructose, ribose, arabinose, mannose, galactose, et cetera. That could have been part of their definition. But someone decided that the deoxy sugars, like deoxyribose and fucose, should be included too. They don't even fit the carbohydrate pattern with n != m. To be fair, it's not just them; other things I'm considering carbohydrates break the pattern too. Glycerol and ethanol for example. But we could have had nice formulaic definitions.
The two famous methods of making sugars are Butlerov's formose reaction and Kiliani-Fischer synthesis. They both make normal sugars with n = m. People who research extraterrestrial chemical food synthesis *love* the formose reaction, but there's not much to love, in my opinion. You start with formaldehyde: CH2O. The guy who invented the formose reaction, Butlerov, also discovered formaldehyde. With its one carbon, you'd think it would the simplest sugar, a "monose". But it's not sweet and it's not edible. It's not called a sugar, but we can use it to make sugars. Here's the setup: first you heat up a solution of formaldehyde and water, with some calcium hydroxide as a catalyst. Next, nothing. That's the whole synthesis. Adding the water isn't even a step: formaldehyde is already normally stored in water, because if you leave pure formaldehyde sitting out, it spontaneously polymerizes into paraformaldehyde, with 8 to 100 moner subunits. The calcium hydroxide just helps it to join up into sugars instead.
Down at the molecular level, there are lots of things that happen in the formose flask. The first one that's of interest here is the condensation of two molecules of formaldehyde to form glycolaldehyde. Glycolaldehyde a two-carbon molecule with n = m, and people call it a diose, but not a sugar, though it is reportedly sweet-tasting. I don't know if it's edible or not, but I doubt it. Keeping heating the formaldehyde flask and soon three-carbon sugars form, trioses! These ones are traditionally called sugars, even though they aren't used as food. One of them, dihydroxyacetone, is used in sunless tanning lotion, and might be a little toxic just on your skin. The other triose, glyceraldehyde, is a chiral molecule with L- and D- forms, and it's an intermediate in a human metabolic pathway, so you'd think we could digest some of it. But I didn't see any online sources mentioning its edibility or use as food after some quick searches. Back in the flask, tetroses form, and pentoses, and hexoses. And the chemists say, "STOP!", because hexose sugars are the bestose sugars, but higher carbohydrates form anyway. While all of those things are forming, so-called "Cannizzaro reactions" are breaking the products down into alcohols and carboxylic acids, such as methanol and formate. And other reactions are happening too. For example, some molecules are rearranging - turning from ketone sugars into aldehyde sugars and back, for example. Among the reaction products, you get some sugar alcohols that are used as diet sweeteners, like sorbitol, xylitol, erythritol. Sugar alcohols are defined to have at least one OH group attached to every carbon. And more complex versions of those form too - branched polyols. And other stuff. In every batch, there are going to be some big weird hydrocarbons that no one's ever seen before. You should imagine "Entry Of The Gladiators" playing. Lots of sugars are produced, a large portion even, but there's too much junk. The end result of the reaction is a sweet-tasting syrup called formose, and it's not food. If you feed it to rats, they die. If you feed it to yeast, they don't ferment. And of course it's not food. We use formaldehyde to preserve organs. Dihydroxyacetone is tanning cream. Just 10 ml of methanol will make you blind. Formate fucks with your mitochondria. Even among the produced sugars, most of them are chiral and we can only digest the dextro enantiomer, but the formose reaction will produce both forms equally.
So lots of people tried fixing the reaction. They use a different inorganic mineral as a catalyst, or organic bases. Or some weird organo-metal compound. Or they stop the reaction before the formaldehyde is all used up, so that there are fewer big crazy molecules made. Or they try running the reaction in a microporous zeolite that will keep big molecules from forming, simply by limiting the space in which reactions happen. Or instead of doing the reaction in a heated flask with water, they do it in a solid-state form with a ball mill providing the activation energy. Or they use UV light instead of heat. Or they do it at high pressure. Most of this work was done in the early days of space flight, in the 60s, 70s, and 80s. We tried, we failed, and some people are still trying. I wish them the best of luck, but I'm not particularly hopeful about formose reaction variants. Some of them are selective for specific reaction products, familiar things like 2,4-bis(hydroxymethyl)-3-pentulose or 3,3-di-c-(hydroxymethyl)-3-deoxy-furanorono-1,4-lactone or 3-c-(hydroxymethyl)pentofuranose. They just don't selectively produce sugar.
I read a recent bachelor's thesis (Headley 2019) that designed a processing plant for filtering sugars from formose syrup using mostly size-exclusion chromatography and ion-exchange chromatography. I know there's some large-scale continuous chromatography in industry, but I'm used to seeing chromatography applied in flask and vial-scale batch process for analysis, so I'm not sure how much merit his system has for food production. But in principle, maybe? Maybe with one of the formose modifications and some filtering, you can get a product with a high enough portion of sugar and a low enough portion of (formaldehyde, dihydroxyacetone, methanol, formate, ...) that you could grow yeast on it, if not feed it to animals or people. Although if we're allowing microbes like yeast into the diet, then there are also microbes that make pure sugar, so again, not there's much point in doing a formose reaction.
Or maybe I spoke too soon! There's a polyol compound called pentaerythritol that's solid at standard conditions with an oral LD50 of 18,500 mg/kg in rats. Its synthesis is almost the same as the formose reaction: mix formaldehyde and acetaldehyde in water with calcium hydroxide catalyst. Here are some more details. I'm not sure which details make the difference compared to formose - less heat and the stoichiometric addition of acetaldehyde, presumably. I don't know that humans have ever tried eating it. But if it's as edible to us as it is to rats, and if it's also chronically non-toxic, and if perhaps it is even palatable, then we have an easy synthesis of a solid carbohydrate. Nice. Normal food is often solid. But still no glucose. Acetaldehyde is made in the Wacker process (the oxidation of ethylene) or by the hydration of acetylene.
What about the other famous sugar-synthesis reaction, Kiliani–Fischer synthesis? I'll write about that one soon. It might work. You have control over what you make, at the expense of the reaction being really slow. Every carbon addition takes multiple steps, and you need to do chemical separations in between. Also the yields aren't great, and lots of cyanide compounds are involved all throughout, which is not ideal for food. The reaction conditions are quite harsh.
Oh whoa, nevermind. Organo-catalyzed aldol reactions for sugar synthesis have taken off in the last twenty years with very little fanfare. I'm going to start by reading "Total Asymmetric Synthesis of Monosaccharides and Analogues" (Robina et al, 2011). Maybe we can do sugars now. That would be amazing. You can do a lot more things once you have sugars. You can do vitamin C for example, which is essential to the human diet.
To make glucose their way (one of the simpler of their ways at least) you need silyl compounds to use as protecting groups and some proline as a chiral catalyst. We can make proline, That's not a problem. I talk about it in the amino acid section. I know the theory of silyl protecting groups (how to put them on, where to put them, how to take them off) but I don't know how easy it is to make them and use them at scale for making large amounts of sugar. Gonna have to look into that.
Assuming we have monosaccharides, can we then make them into polysaccharides? Can we link up glucose into amylose, for example? I don't think so, no. But we can make some smaller versions: some disaccharides and oligosaccharides. There's been much less work done on this topic than on sugar synthesis, but some. Caramelization does it a little bit, but there's a better way. Selective glycosylation reactions. We'll get to it. I think progress on this subject would be a big step toward improving the normalcy and palatability of the non-enzymatic diet. If you can make sugar into amylose, you can make flour. And from there you can make bread and pasta. We'll see later that we can already make oil and margarine. And I have some more reading to do, but I think we can make a complete protein amino-acid shake too. Honestly, that's already a better diet than many people on earth are accustomed to. Provided we can figure out how to introduce lots of alpha-(1,4) glycosidic bonds to glucose.
So, we're not good at making sugars and starches non-enzymatically yet, so far as I've read the literature. But since carbohydrates aren't essential, I don't think it's particularly important which ones we're good at making. What we're good at making is glycerol, and I think it might be worth the effort to make from scratch in space or on Mars: it's a pretty easy synthesis from simple gasses (easy enough for earth factories or NASA to do it, that is. It's perhaps not quite a backyard shed synthesis), and it's very digestible, and it gives us a source of carbohydrate calories. Also, glycerol is clear and slightly sweet and otherwise basically flavorless, and you can eat quite a lot of it compared to glucose without your blood sugar spiking (glucose has a glycemic index defined to be 100. A G.I. of 55 and below is considered low. I've seen glycerol's glycemic index reported as 3 and as 5. Pretty damn good.) I don't know about you, but flavorless slightly-sweet transparent nectar is exactly the kind of thing that my temperate and ascetic sides want to subsist on, while simultaneously my nerdy side is excited by the prospect of chemically-derived space slime. If you're also interesting space slime, do stick around for the section on microbial manufacturing.
But man can not live on glycerol alone. In addition to needing calories, there are some essential nutrients that we need, like nine to fifteen particular amino acids, and two to seven fatty acids with double bonds in just the right places that human metabolism can't introduce itself, a bunch of vitamins, and a few trace elements that might need to be processed into specific compounds to be bioavailable. Which essential nutrients can we make non-enzymatically?
: Fatty Acids and Triglycerides
Carboxylic acids are hydrocarbons that have a COOH group sticking off one end, more or less. If the rest of the molecule besides the COOH group is an aliphatic/non-aromatic hydrocarbon chain, then the carboxylic acid is called a fatty acid. In practice, they're not just aliphatic, they're even unbranched. A straight line or a slightly curved line. Formic acid (sprayed by wood ants for defense) and acetic acid (found in vinegar) are the first two saturated fatty acids - the shortest and simplest ones. Organisms in nature synthesize mostly fatty acids with an even number of carbons in the chain, up to lengths of about 28 carbons. They're conventionally divided into four groups: short-chain (1 to 5 carbons), medium-chain (6 to 12), long-chain (13 to 21), and very long-chain (22 or more). Fatty acids with odd-numbers of carbon atoms in their chains do exist and they're metabolized by the human body a little differently than the even ones. There are many fatty acids, but only two are normally considered essential to human metabolism: linoleic acid (ending in LEIC, just called LA for short) and alpha-linolenic acid (ending in LENIC, called ALA for short). Horrible names. There are two more fatty acids that infants can't make fast enough from other sources: arachidonic acid (AA) and docosahexaenoic acid (DHA), so those are called conditionally essential. But I think infants are essential, so maybe we should just say that there are four essential fatty acids. No, don't worry, I'll use the standard terminology. A few more might become essential if you have some other metabolic disorder besides (or in addition to) being an infant: docosapentaenoic acid (DPA), eicosapentaenoic acid (EPA), and gamma-linolenic acid (GLA). I'll call those conditionally essential too. Essential: (LA, ALA). Conditionally essential: (AA, DHA, DPA, EPA, GLA). All five are long-chain, and they all have carbon double bonds (they're partly unsaturated).
A triglyceride is a combination of glycerol with three fatty acids, not necessarily the same three. Triglycerides are more famously known as fats and oils, depending on whether they're solid or liquid at room temperature. Triglycerides taken from plants and animals contain mostly long-chain fatty acid subunits, but that's not a chemical requirement. To make triglycerides, you mix together the fatty acids and the glycerol with some sulfuric acid as a catalyst, and then they combine (in a process called esterification). We're good at digesting fatty acids in the form of triglycerides, and normal earth diets include lots of different triglycerides, and they're really high in digestible calories per gram. So making synthetic triglycerides from scratch definitely has merit for the space food menu.
"Do we really need to make triglycerides of the fatty acids? Is the esterification essential?" Well, no. But the triglycerides generally have higher LD50s, which is good for not killing people. This is possible at the expense of being less readily absorbed. Not sure. Also people know how to cook with triglycerides. Even if it's not metabolically essential to esterify the fatty acids with glycerol, it would be nice if space food was a little bit like earth food, don't you think? And it's barely an expense: the sulfuric acid is a catalyst; theoretically, it doesn't get used up in the reaction.
So if we have fatty acids, then we can combine them with synthesized glycerol. But can we synthesize fatty acids? All of them, any of them?
Short-chains: We can make formic acid (1 carbon chain), and with an oral LD50 of 1.8 g/kg in mice, it's not particularly edible, but it's still sometimes used in dilute form as a food additive. It's also highly corrosive, like 10 times stronger than acetic acid (that might not sound bad if you think vinegar = acetic acid, but vinegar is only like a 6% solution, with the rest being water. Pure acetic acid will put lesion on your esophagus). In addition to being corrosive, formic acid is pretty flammable. And it decomposes readily, producing carbon dioxide that builds pressure in its container. We can do better. Acetic acid (2 carbons) is edible, and so is the triglyceride made with three molecules of acetic acid, triacetin. Acetic acid is fine at low concentrations, of course, and it seems like a good idea to have it in space diets, for normalcy if nothing else. The pure form with no water isn't much better than formic acid in terms of its LD50, but that's not how people make salad dressings. Acetic acid can be made in the Monsanto process by the carbonylation of methanol, i.e. by reacting with carbon monoxide (CO). Before Monsanto, acetic acid was made by the oxidation of acetaldehyde. When you combine an alcohol with an acid, you get an ester, and acetic acid has some decent esters in terms of toxicity: Isobutyl acetate (LD50: 13,400 mg/kg) and its isomer n-butyl acetate (LD50: 10,768 mg/kg) are both liquids used as flavorings and are found naturally in many fruits. They have a scent similar to bananas or apples. The other acetic acid esters I've seen aren't as edible, although some of them do smell nice - reminiscent of blueberries, for example.
The triglyceride made from glycerol and three molecules of acetic acid is called triacetin. It's absorbed readily by the body (particularly, through the hepatic portal vein) and the liver is able to turn it into the glycogen polysaccharide for energy storage. Some people in the late 60s found that baby rats could get half of their calories from triacetin, which is a lot ("Current research on regenerative systems", (Shapira et al. 1969)). So you'd think it would be highly edible, yeah? The oral LD50 for rats is only 3000 mg/kg according to PubChem. In clown units, that means that if two 150 lb rat-people each ate a cup of it in one sitting, one of them would probably die. I'm a little confused. I could try tracking down lots of different experiments giving different LD50s, but let's just find a better chemical to eat. The smoke point of triglyceride oils and fats increases with their chain length, so while triacetin is an oily liquid, I doubt it would even be useful for cooking.
Adding another carbon to the fatty acid chain, we get the three-carbon propionic acid, a liquid that smells like body odor. The triglyceride, tripropionin, has a nutty taste and an LD50 of 6,400 mg/kg. Getting better. The last short-chain saturated fatty acid is butryic acid, at four carbons. Butyric acid is formed naturally by your gut bacteria (colonocytes) when they ferment ingested carbohydrates and dietary fibers. Butyric acid is metabolized in the liver to form a ketone body, D-beta-hydroxybutyrate, which is a rabbit hole to some complicated stuff.
Complicated ketone body stuff: Butyric acid (4-carbons) and the medium-chain fatty acids are readily converted to ketone bodies. Ketone bodies can be synthesized by the human body from fatty acids and a few amino acids, particularly in times when it can't make much glucose. Ketone bodies are used as an alternative source of energy to glucose, particularly in the brain, and one reason you might not be able to synthesize much glucose is that you're ingesting few carbohydrates. Even-chain fatty acids are used to make ketone bodies with 4 carbon atoms, in particular acetoacetate, beta-hydroxybutyrate, and acetone. Are they a problem? Not usually: low-carb and high-fat diets (ketogenic diets) are sometimes used to treat epilepsy, so people can definitely live for extended periods of time off of ketogenic diets, and even do so with improved health. Admittedly, there are situations, such as disorders of fatty acid metabolism, where ketone bodies build up too much in the blood (hyperketonemia), making your blood too acidic (ketoacidosis). I once read (in "Short- And Medium-chain Fatty Acids In The Energy Metabolism" (Schönfeld and Wojtczak, 2016)) that butyric acid also inhibits glycolysis, the aerobic degradation of glucose into pyruvate, but I'm skeptical. It definitely has lots of complicated effects on metabolism through its activation of "G protein-coupled receptors" such as HCA2, but inhibiting glycolysis doesn't seem to be one of those. The impression I get is that you should eat some butyric acid to feed your gut bacteria (they produce it from fiber, but they also eat it. It's not a waste product for colonocytes), and maybe you should eat a different amount if you have epilepsy or diabetes. Interestingly, here is a paper claiming that you can treat cows suffering from overly-high tributeryn intake by administering either salts of the 3-carbon propionic acid or the chemical 1,2-propanediol (i.e. our good friend Propylene Glycol from the carbohydrate section). Some things about that paper are confusing to me too though.
Triglycerides of butyric acid, especially tributyrin, make up about 3% of butter, probably as a consequence of the metabolism of colonocytes in cow guts. Tributyrin is also used as an additive in margarine. It has an odor of vomit and rancid cheese when concentrated, but is perfectly palatable in small amounts as a flavoring. Butyric acid can be made by oxidation of butyraldehyde, which is made by hydroformylation of propylene. PubChem lists the oral LD50 of tributyrin at 3,200 mg/kg for rats and 12,800 mg/kg for mice. I have no earthly idea why there's a factor of four difference. Even if it turns out to be highly edible for humans, I wouldn't eat large amounts of it simply because of the butyric acid-stimulated ketosis thing. But if it smells like rancid cheese vomit, then eating too much probably isn't going to be a problem for anyone.
(There's a beautiful quote in the abstract of a government report, "Nonterrestrial Chemical Synthesis Of Food" (Siegel, 1966): "Questions might be raised as to the psychological acceptability of synthetic foods, but we can assume that space travelers are highly motivated in this regard." He didn't talk about tributyrin, but I can't help but laugh at the idea of some airforce chemist designing a diet with lots of rancid cheese vomit acid, shrugging his shoulders, and saying, "we can assume the astronauts are highly motivated".)
Medium and long-chains: Medium-chain fatty acids (6 to 12 carbons) are more readily digested than long-chains. They make up about 10% of animal milks. There are only four with straight chains (i.e.saturated, unbranched) and an even number of carbons: caproic (6), caprylic (8), capric (10), and lauric (12). I think they were found in goat milk first, hence the capra names. I don't know much about their synthesis. They are recommended for people with some genetic disorders of fatty-acid metabolism - certain enzyme deficiencies. So maybe they should be called conditionally essential? No, but let's put them in our space diets anyway. The saturated fatty acid with seven carbons (heptanoic acid) is also a little famous. It's esterified with glycerol to form triheptanoin and is then sold as a brand name medication to treat fatty-acid disorders. Having an odd-chain, heptanoic acid is metabolized into ketone bodies with five carbons, rather than the usual four carbon ketone bodies, namely it becomes 3-oxopentanoic acid and 3-hydroxypentanoic acid, both of which are able to enter the body's citric acid cycle.
As for medium-chain triglyceride toxicity, they're finally highly non-toxic. For example, tricaprylin, the glycerol triester of caprylic acid (8 carbons) has an LD50 of 34.2 g/kg in male and rats and 33.3 g/kg in female rats (see "Toxicity, teratogenicity, and pharmacology of tricaprylin" (Ohta et al. 1970)). I just saw that ethyl caprylate, the ester of caprylic acid and ethanol, has an oral LD50 of 25,960 mg/kg in rats. It's also used as a flavoring, and it has a scent like fruit and flowers. That sounds great. Maybe I should look up ethanol esters for other fatty acids.
I don't know the synthesis routes for individual medium-chain fatty acids, but I know a non-selective route. There was a really early non-enzymatic synthesis of triglycerides by a german chemist named Arthur Imhausen, using medium- to long-chain fatty acids as subunits. He made purely synthetic margarine in the 1940s, and the product was successfully used to feed soldiers. Nazi butter. That's where we're headed in this section. I've mostly read short scattered references to Imhausen's work, and I really need to research this better, but I think here's how Imahusen did it: first, he made long-chain hydrocarbons from CO2 and H2 by the Fischer-Tropsch process. It's one of my favorite chemical reactions. Want to see an animation with a pleasant male British voice-over? Fischer-Tropsch makes mostly saturated alkanes (i.e. the carbons have single bonds between them, and hydrogens all around), but side reactions also produce some alkenes with double bonds, and some alcohols, and some other junk (like maybe some slightly branched compounds or even cyclic ones). Fischer-Tropsch is a lot like making crude oil, with portions like natural gas, like kerosene, and like paraffin wax. Imhausen took the solid portion, the paraffin wax, as his starting material. German sources call it "gatsch" or sludge, but it's wax. I think Imhausen then melted and distilled the wax, and separated its components out by their boiling points. Larger hydrocarbons have higher boiling points, you see, and FT can make things up to like 70 carbons, while natural fatty acids are under 30 carbons in length, so there's a sweet spot of boiling points you want to capture. Then he oxidized the wax, turning it into fatty acids. I need to look up more about that step. Most references just say "catalytic oxidation", but I don't know what catalyst will turn wax and O2 into carboxylic acids. Chemically, I think the alkanes have to go through multiple intermediate stages of oxidation before the COOH group is formed, but macroscopically, it might just be one step for the manufacturers. I think Imhausen then reacted the fatty acids with an alkali, to make fatty acids salts (i.e. soap), and this was done as a filtering step to remove the nice fatty acids from the non-saponifiable side products of Fischer-Tropsch and the catalytic oxidization stage. Then you treat the soap with a mineral acid to get back your fatty acids and some mineral salts on the side. I've never tried turning soap back into oil using, say, hydrochloric acid or sulfuric acid, but that's on my to-do list once the pandemic ends and I've got some free time again. Finally, Imhausen combined the purified fatty acids with glycerol to make triglycerides.
I've got some big questions about Imhausen margarine.
* Does it contain lots of odd-numbered carbon-chain fatty acids, and if so, is that a problem for human metabolism? Based on extensive animal studies around WWII by E. Flossner, it doesn't seem to be a problem, but I'd like some more modern assurance.
* If the Fischer Tropsch process mostly makes saturated alkanes (with no double bonds between carbon atoms), is Imhausen margarine likely to contain any of the two essential fatty acids (which both have multiple double bonds in their carbon chains)? Between having to get the double bonds in the right places, and the fact that the essential fatty acids also have chiral centers (another dice roll), I'd say Imhausen margarine had a negligible amount of essential fatty acids, if any. So Imhausen gives us cheap calories and a normal foodstuff for human diets, but not an essential nutrient.
* What are the likely products of oxidizing the alcohols and other side reaction products from Fischer Tropsch? It seems that the main impurities in Imhausen margarine were slightly branched things (so-called iso-acids, being mostly like long straight fatty acids, but with small methyl, ethyl, and propyl side chains). The ethyl and propyl side chains of iso-acids are metabolized a little differently by the body than the straight portion, and this can cause an increase of dicarboxylic acids in the urine (aciduria), and maybe nausea or diarrhea. So if Imhausen margarine is used in a space diet, we should filter those out better. I hear it didn't hurt the German soldiers, but let's nevertheless, you know, more fully characterize and purify the stuff that we'd ideally be feeding to people as a major part of their diets for an extended period of time.
"Ideal for what, your technology fetish?" Harsh! Moreso I meant ideal for the sake of prudence. It would be prudent to have multiple independently functioning systems for sustaining astronauts separated from Earth, living in the dark, frozen, irradiated, evacuated, poisonous wasteland that is Mars. Like we could shoot for a chemical food factory that's nutritionally adequate by itself, and an adequate microbial food factory, and an adequate plant food factory. Let's get back to the chemical synthesis of fatty acids. We're almost at the essential ones:
: Long-chain Fatty Acids (13 to 21 carbons):
Skipping past some of the shorter ones for now, stearic acid is the unsaturated fatty acid with 18 carbons. It's not essential. I see an MSDS listing its oral LD100 (not LD50) in humans as 14,286 mg/kg. Apparently, the source is Gosselin RE et al. (1976). Sad. I don't know that I want to learn the story behind that figure. The LD50 in rats is just 4,600 mg/kg, but the triglyceride with three stearic acid molecules (tristearin) is much better at 20,000 g/kg. Another 18 carbon fatty acid, oleic acid, has one carbon double bond - i.e. one point of unsaturation - and it's the main component in olive oil triglycerides.
Those two are great for soaps and for food. Even more important are the essential and conditionally essential fatty acids. Let's talk about them all. Essential: linoleic acid (LA, C:18) and alpha-linolenic acid (ALA, C:18). Conditionally essential: arachidonic acid (AA, C:20), docosahexaenoic acid (DHA C:22), docosapentaenoic acid (DPA, C:22), eicosapentaenoic acid (EPA, C20), gamma-linolenic acid (GLA C:18).
...
At 20 carbons, eicosapentaenoic acid (EPA) is conditionally essential. It has multiple carbon double bonds (it's polyunsaturated), and the body can make it from essential ALA, but inefficiently. EPA is then made into DHA, which is hugely important. DHA is the main fatty acid in phospholipids in the brain and the retina, and maybe the skin. It's a good idea to get all of them in your diet.
...
Selective synthesis of long-long-chain fatty acids: For making long-chain hydrocarbons, in addition to the Fischer-Tropsch process, modern chemists also sometimes link up a small number of ethylene molecules (i.e. oligomerization, the small scale version of polymerization). Ethylene is not as simple a gas as the reactants in Fischer Tropsch, and that's a step back for feasibility in space, but if there's a possibility that oligomerization of ethylene gives us better control over the location of double bonds, then maybe it's worth looking into.
Oh wait, wait, wait, forget that! We can make specific long-chain fatty acids, including the seven famous polyunsaturated ones, the essential and conditionally essential ones! I just had to search google scholar better. Carbonyl olefination is the most common method of synthesis, but there are a bunch. I shouldn't have talked so long about Imhausen. We can make fatty acids. We'll be fine. I'll write it here when I understand it. //
Linoleic acid (LA): Linoleic acid is a fatty acid, which is a carboxylic acid with a long tail. It has a COOH carboxyl group on one end and 18 carbons in total. If we start counting down the tail starting with the carboxyl group carbon as "1", then all of the bonds between carbon atoms are single bonds except between 9 and 10 and between 12 and 13, which are both double bonds. All the double bonds of the essential and conditionally-essential fatty acids are in the cis- rather than trans- configuration, so that can go unsaid, and we can summarize linoleic acid with "lipid numbers" as 18:(9, 12). 18 carbons and two cis-double bonds, starting at positions 9 and 12 respectively. Much easier than writing (9Z, 12Z)-octadeca-9,12-dienoic acid.
Alpha-linolenic acid (ALA): ALA is the other essential fatty acid besides LA. While LA is 18:(9,12), ALA is 18:(9, 12, 15), i.e. it also has 18 carbon atoms, but it has an extra double bond between carbons 15 and 16, counting down the tail with the carbon of the COOH group as 1. While LA was made in 1950, ALA wasn't made till 1995. So that extra double bond was kind of tricky, I guess. But I think the new method also lets us make the conditionally-essential fatty acids, so it was a really good development. If you weren't feeling proud of your species, take heart remembering the stereospecific synthesis of skipped polyunsaturated fatty acids. Details:
...
I dunno, man, how about some references:
"Convergent and Stereospecific Synthesis of Complex Skipped Polyenes and Polyunsaturated Fatty Acids"
"Construction of (Z,Z ) skipped 1,4-dienes. Application to the synthesis of polyunsaturated fatty acids and derivatives"
"Synthesis of Naturally Occurring Unsaturated Fatty Acids by Sterically Controlled Carbonyl Olefination"
"Direct Preparation of (Z,Z)-1,4-Dienic Units with a New C6 Homologating Agent: Synthesis of α-Linolenic Acid"
...
Arachidonic acid (AA): The body can make this one from LA.
Gamma-linolenic acid (GLA): The body can also make this one from LA.
Eicosapentaenoic acid (EPA): The body makes this one from ALA. Kind of inefficiently though, so here's how you make it:
Docosapentaenoic acid (DPA): The body can make this one from EPA, which it can make from ALA. Another step of inefficiency. It's made by:
Docosahexaenoic acid (DHA): The body can make this one from DPA which it can make from EPA which it can make from ALA. Three steps of inefficiency now. Ouch. Synthesis:
:: Amino acids
What about other essential nutrients? Can we make any amino acids by non-enzymatic means? Yes, I think so! The oldest and most common way is Strecker synthesis, which I will focus on here, but there are some other options like Gabriel malonic ester synthesis or Bucherer–Bergs reactions followed by hydrolysis. Some history and terminology:
A carbon atom with a double bond to an oxygen atom is a carbonyl group. The carbon can then be covalently bonded to two more things. If the carbon is bonded to a hydrogen atom and to something else, then the molecule is an aldehyde. The defined parts are written CHO and called the aldehyde group, and the rest is called an R group, because undefined things sticking off any molecule in organic chemistry are R groups. If instead you have a carbonyl group and neither of the bonded things is a hydrogen atom, then it's more complex, and it's called a ketone. Those will come up later. They're twice as tricky as aldehydes. If the aldehyde has a straight chain, then it's named after a corresponding straight-chain carboxylic acid. 1 carbon: (formic acid, formaldehyde). 2 carbons: (acetic acid, acetaldehyde). 3 carbons: (propionic acid, propionaldehyde). And so on.
In 1850, Adolph Strecker tried to make lactic acid from acetaldehyde (CH3CHO), ammonia gas (NH3), and hydrogen cyanide (HCN) (also called prussic acid, which is liquid at standard conditions). But his reaction made alanine instead, the simplest alpha-amino acid after glycine. Easy! Amazing! Miraculous! One of the building blocks of life from such simple origins. What an achievement.
What exactly are amino acids such as alanine and glycine? Let's start with alpha-amino acids, the ones in proteins. Except for proline. Tricky proline. They all have an NH2 amine group sticking off a central carbon (the alpha carbon), and a COOH carboxylic acid group sticking off it too, and an undefined R group sticking off it, and the fourth valence electron of the alpha carbon is just bonded to a hydrogen atom. In alpha-amino acids, the COOH is also adjacent to the NH2, rather than opposite from it. In any amino acids, the R group is called a side chain and the rest of the molecule is called the backbone. All the alpha-amino acids have the same backbone, and they only differ in the side chain R group. And I think amino acids generally form solid white crystals with high melting points.
Nine of the proteinogenic amino acids are essential (histidine, isoleucine, leucine, lysine, methionine, phenylalanine, threonine, tryptophan, valine) and six more are conditionally essential (arginine, cysteine, glutamine, glycine, proline, tyrosine), and six of them our bodies can make on their own with no problems (alanine, asparagine, aspartic acid, glutamic acid, selenocysteine, serine), and one more doesn't occur in human proteins at all (pyrrolysine).
Strecker's synthesis preserved the R group of the starting aldehyde, namely the CH3 on the acetaldehyde, and transformed the rest of the aldehyde into the alpha backbone. And that's exactly what alanine is: a CH3 sidechain on the alpha backbone. The synthesis works with other aldehydes too. If Strecker had started with formaldehyde (HCHO) instead of acetaldehyde, then the R group connected to the CHO aldehyde group would have just been a hydrogen atom, and the resulting alpha-amino acid would have glycine. If you can get an aldehyde with the desired R group, then Strecker synthesis can turn it into an alpha-amino acid with the desired R group.
You can also perform Strecker synthesis with some slightly different starting chemicals. Ammonia is often mixed with hydrochloric acid to make ammonium chloride, (NH4Cl), and the hydrogen cyanide could be NaCN or KCN, for example. The CN is what does the work. The amino acid produced will still be the same.
The reaction has two main steps: First, the aldehyde becomes an aminonitrile (the carbonyl oxygen double bond is replaced with single bonds to an NH2 amine group and a CN cyanide group). This step is acid-catalyzed, which is why you need the prussic acid (HCN) or hydrochloric acid (HCl). Next, the CN group reacts with water and becomes a COOH. There are like eight little steps if you follow around all the electrons, but that's the gist.
Which alpha amino acids can we make by Strecker synthesis? I don't know offhand, but I'm going to go through every one to figure it out. I do know that Strecker can't make proline, because proline's side chain is connected to its backbone twice, making a little ring. Proline is a tricky bastard. We'll talk about that one separately.
I should mention that terrestrial life uses the levo- enantiomers of amino acids (like L-histidine, L-proline, L-glutamine, ...), except for glycine which is achiral (it's so simple with its hydrogen side chain that it doesn't have a chiral center). I'm not sure how to make amino acids with the preferred chirality. I've heard of "asymmetric Strecker synthesis", but I don't know the details. Chiral catalyst? Chiral replacement for NH3? Something. I think you can build the side-chain aldehyde first and worry about chirality later. So let's figure out if we can make the Strecker aldehydes.
I found a list that had all but three of them, all the technical names worked out, in wonderful paper, "Nature's Starships. II. Simulating The Synthesis Of Amino Acids In Meteorite Parent Bodies" (Cobb et al, 2015) . They omit cysteine and methionine and selenocysteine, because those contain an atom of sulfur , sulfur, and selenium, respectively, and those elements weren't part of the meteorite chemistry they were investigating. So I only have to figure out the names for three of them, and ACD/ChemSketch will look those up for me if I draw them.
Here are the essential amino acids and their corresponding Strecker aldehydes:
* lysine: 5-aminopentanal (C5H11NO): I haven't found a synthesis for this aldehyde, but Roger Gaudry made lysine a different way in 1948 from dihydropyran, and another lab did a similar synthesis from dihydropyran in 1949. Most of the routes I see for dihydropyran synthesis ultimately come from the dehydration of sugars and polysaccharides in plant biomass, so we're still not quite there. There's also an ugly synthesis of dihydropyran from ethyl lactate that involves silyl protecting groups and chelation-controlled alkylation. I'm going to look for something better. Maybe there's a way to do it starting with cyclopentanone or maleic anhydride.
* phenylalanine: phenyl-acetaldehyde (C8H8O). Wikipedia mentions seven synthetic routes to phenylacetaldehyde! Surely one of those can be worked into a total synthesis.
* threonine: 2-hydroxy-propanal (C3H6O2): More commonly known as lactaldehyde. You make it by ...
* tryptophan: indol-3-ylacetaldehyde (C10H9NO):
* valine: isobutyraldehyde (C4H8O): Isobutyraldehyde is made by the hydroformylation of propene.
Here are the conditionally essential amino acids and their aldehydes:
* glutamine: 4-oxobutanimide (C4H7NO2)
* glycine: formaldehyde (CH2O): Formaldehyde is made by combining methanol with oxygen (2 CH3OH + O2 = 2 CH2O + 2 H2O).
* tyrosine: p-hydroxyphenyl-acetaldehyde (C8H8O2)
: Pepetides and proteins
When you link amino acids up, you get pepetides. A molecule with two to fifteen amino acid blocks is an oligopeptide, and larger peptides are called a polypeptides. Fifty or more amino acid pieces and your polypeptide is called a protein. The protein albumin has almost 600 amino acid residues.
Can we make them non-ezymatically? We can definitely make small pepetides. I'm not sure about big ones. I will first look into whether any of the small ones are worth making .
...
:: Enzyme cofactor vitamins and choline
For even moderately long missions, it's probably more economical to bring vitamin supplements with you than to bring vitamin-making chemical equipment. (For short missions, it's probably also economical to bring fatty acid oils). But this essay isn't about bringing earth food on short missions, it's about our ability to make food for space missions of unspecified length. And if we want a self-sufficient Martian colony, then for very unspecified lengths. So what vitamins can we make for the long haul?
I know a little bit about vitamin synthesis, but not a ton. The short story is that, in practice, we can't make all of the essential vitamins non-enzymatically. But I think it's important to know which vitamins we're good at making, and which ones we can't make at all, and which ones we can theoretically make, but not fast enough to feed hungry astronauts and their Martian children, so I will include all of that here eventually. There are thirteen essential vitamins: (A, C, D, E, K, B1, B2, B3, B5, B6, B7, B9, and B12). We'll hit them all. Also choline. It's an essential nutrient. I don't know why it's not called a vitamin. Maybe because it's not an enzyme cofactor. But then neither is vitamin D. I'll write about its synthesis here with the vitamins. Because it probably is a vitamin.
: Vitamin A.
Structure: A vitamin name specifies a biological role, not a chemical. Vitamin A has a few forms, or "vitamers". The big two are retinol and retinal. Similar to how fatty acids can combine with the alcohol glycerol to make triglyceride esters, the retinol alcohol that we get from plants is often paired up as an ester with a fatty acid, palmitic acid, forming retinyl palmitate, another vitamin A vitamer. All the vitamers A are fat-soluble and they all have a very similar form - there's a special ring structure called a beta-ionone ring, which is in the chemical class of ketones, and there's a chain of four isoprenes molecules with some junk sticking off. Small isoprene chains are called terpenes (monoterpenes have two isoprenes, diterpenes have four, ...). Terpenes, and related terpenoids, are the main compounds in fragrant essential oils. If the word "isoprene" sounds familiar, but you just can't place it, long isoprene chains are called rubber. You've might have heard the phrase "isoprene rubber" at some point. Anyway, the chain structure on vitamin A, any vitamer, is a diterpenoid, having four isoprene units and some junk. Organisms with the enzyme beta-carotene dioxygenase, such as we humans, can also take photosynthetic pigments called carotenes, which have that same beta-ionone ring structure, and turn them into retinal. Those carotenes are called pro-vitamins or pre-vitamins, and I think synthesizing them would be just as good as synthesizing real vitamin A.
Function: Vitamin A is used in an enzymatic pathway in the retina of the human eye for light transduction and as a growth factor for epithelial cells. Epithelial cells are in your skin and mucous membranes, so vitamin A is important for immune function, and also in your urinary tract, making it important to male reproductive function.
Synthesis: Now, can we make vitamin A in a lab, any of the vitamers or pro-vitamins? Yes! Retinol was synthesized in 1946-1947 by two Dutch chemists, Arens and van Dorp. Kuhn and Morris reported an early synthesis, but it was garbage, apparently? Also in 1947, Otto Isler and his colleagues found a modified version of the Arens-van Dorp retinol synthesis which could be scaled up for industrial production. Isler is also known for achieving the industrial-scale synthesis of two other fat-soluble vitamins, vitamin E and vitamin K (in the form of phylloquinone).
"But how is it done, James? Is it a total synthesis, or do they start with some essential oil and just fiddle with it a bit, like cheaters, like dirty old fiddle 'em Bens?" Whoa. I might be inclined to take offense to that, if I knew what the hell you were saying. Both syntheses began with the beta-ionone ketone and built up its carbon sidechain in stages. The other chemicals used in the reactions, like propargyl bromide, methyl bromoactate, ethyl chloroacetate, and methyl vinyl ketone, are pretty simple, so if we can figure out how to make beta-ionone, then we can make vitamin A in space. We can make ionone from citral and acetone. Acetone is made in the cumene process, which sticks an oxygen atom onto propylene. What about citral? Can we make that? Citral is a monoterpene in essential oils, and you usually get it in nature as a mix with more of the trans-isomer (geranial) and less of the cis-isomer (neral). These words will be familiar to anyone who cares about perfumery. Citral can be synthesized from myrcene. Usually myrcene is made by pyrolysis of pinene. I've seen a patent with few details for making mycrene, "Direct synthesis of myrcene from isoprene" and I've seen a paper claiming pinene synthesis: "The total synthesis of alpha- and beta-pinene" (Thomas and Fallis, 1975). Good enough for me. Isoprene for the myrcene synthesis can be made a few ways, such as by reacting formaldehyde with isobutene. Isobutene is made by dehydrating isobutanol. Isobutanol can be made from syngas over a weird metal catalyst (Zr-Zn-Mn-Li-Pd). One way or another, we can make retinol is space. That's one vitamin down, twelve vitamins and choline still to go.
: Vitamin C
Function: The main vitamer of vitamin C is L-ascorbic acid, where L means the left-handed enantiomer. Salts of ascorbic acid, ascorbates, also work, and so does an oxidized version of ascorbic acid, dehydroascorbic acid. Vitamin C is a cofactor to the enzymes that introduce hydroxyl groups onto the amino acids proline and lysine, without which your body can't make the structural proteins collagen and elastin. Fun fact: collagen with the hydroxyl groups removed is called gelatin. Collagen is found in skin, cartilage, ligaments, tendons, blood vessels, corneas, bones, and the dentin layer of teeth. Vitamin C is a cofactor with some other enzymes too, without which we couldn't synthesize the neurotransmitter norepinephrine from dopamine. And also epinephrine is made from norepinephrine, so that would be doubly bad. Vitamin C is also necessary for carnitine synthesis.
Synthesis: Most animals and plants can make their own vitamin C. Humans, guinea pigs, and some bats have lost the enzyme L-gulonolactone oxidase, which turns sugar into vitamin C. Such a dumb situation. When I become the gene therapy research czar, I'm going to fix it so that humans don't need vitamin C in their diets anymore. The usual route for industrial synthesis of vitamin C is the Reichstein process which starts with glucose (which we're bad at making non-enzymatically) and it also uses acetic acid bacteria. We might not be out of luck though. I see a paper called "First synthesis of L-ascorbic acid (vitamin C) from a non-carbohydrate source" (Banwell et al, 1998). It's behind a paywall, but there's a blurry snapshot of its first page. They start with cis-1,2-dihydrocatechol, obtained by microbial oxidation of chlorobenzene. I'm pretty sure the compound 1,2,-dihydrocatechol is more commonly known as 1,2-cyclohexanedione, and we don't need bacteria to make it. So that's weird. *squints for more info* They convert it into 3,5-O-benzylidene-L-gulonolactone. And 2,3-O-isopropylidene-L-gulono-1,4-lactone also makes an appearance. All the famous L-gulonolactones are here. Okay, they mention chemo-enzymatic synthesis. If that's just the enzymes of the bacteria that do the initial oxidation of chlorobenzene, we might be alright. We could just make the cyclohexanedione another way. For now, until I've read the whole paper, I'll say that I don't know a way to make vitamin C chemically, but humanity at large might. Also, let's just fix our stupid broken genes.
If the Banwell paper turns out to use enzymes in another place, then not all hope is lost. Someone has probably done a vitamin C synthesis from sugar without the bacteria. And we do have Kiliani Fischer synthesis for making sugars. I wouldn't want to use KF to make a carbohydrate that's a large part of someone's diet, both because of its inefficiency and the toxicity of attendant cyanide, but maybe for vitamin C, which we need much less of, the cost of inefficiency and the work-up to food-grade purity would be worth the value of producing small essential amounts of an essential nutrient.
Also if the Banwell paper ends up not being great, I've also seen an interesting Ph.D. thesis by Donald Deardorff from 1979 , "Synthetic Applications Of Carbanions Derived From Glycolates - Studies Directed Toward The Synthesis Of Vitamin-C". He proposed a synthesis of vitamin C from tartaric acid. And then he does the synthesis. And he never declares victory. But it's an interesting paper. Maybe that's another possible route.
: Vitamin D
Your body produces D3 (cholecalciferol) from 7-dehydrocholesterol in your epidermis when you receive UV-B radiation. It's a hormone. Then it gets hydroxylated in the kidneys to the active form, calcitriol, another hormone. I'm not sure that vitamin D should be called a vitamin, if we can make it ourselves and if it's not biologically active. But it's a good idea to get it from dietary sources, and it's traditionally called a vitamin, so let's talk about it. The pathway by which it works has a lot of steps, but ultimately vitamin D helps you to absorb calcium in your small intestine, which is especially important in space to fight the increased rate of bone loss in low gravity (disuse osteoporosis). It's also an especially good idea to get your vitamin D as a dietary supplement, because, well, you can't sunbathe on Mars. Or can you? Apparently it only takes like 15 minutes a day under a high-intensity UV-B lamp to get your D3. Maybe we should just bring some lamps to mars and forget the synthetic D. Ah, but what if your lamps break? Let's learn about synthetic vitamin D. For great prudence.
D3 is produced industrially by washing sheep wool to extract lanolin, which contains 7-dehydrocholesterol, and then you irradiate it with UV-B. Not sure if there's another way. I'll look into it. There are other vitamers of vitamin D besides D3 (cholecalciferol) that your body can also convert into the biologically active calcitriol. They're all fat-soluble. There are at least five chemicals on the market called synthetic vitamin D, but they're all derivatives of natural vitamers D. So let's see if we can synthesize the natural ones. Ergocalciferol is known as vitamin D2. It's found in a few culinary mushrooms and also in the alfalfa plant, but that might be because of fungal contamination. Another chemical, ergosterol, that's found in fungi, including the cell membranes of yeasts, can be converted into D2 by UV radiation, so it's called provitamin D2. D2 might be a little less effective than D3, but they'll both prevent bone problems (i.e. rickets). In the worst-case scenario, we can irradiate yeast instead of people. I saw a D2 synthesis paper, but I've lost it.
But wait, I sought again and I found some D3 synthesis papers:
"An asymmetric total synthesis of vitamin d3 (cholecalciferol)" (Wilson and Haque, 1984)
"Synthetic approaches to vitamin D" (Dai and Posner, 1994)
"Synthesis of Vitamin D (Calciferol)" (Zhu and Okamura, 1995)
"New vitamin D analogues" (Posner, 1996)
"An effcient synthesis of the 25-hydroxy Windaus-Grundmann ketone" (Fall et al, 2000)
: Vitamin E
Paul Karrer reported the first synthesis of vitamin E in 1938 in the form of DL-α-tocopherol (DL referring to a racemic mixture) with his team of Hans Heinrich Fritzsche, Beat Heinrich Ringier, and Harry R. Salomon. They reacted trimethyl-hydroquinone (TMHQ) and phytyl-bromide over a zinc chloride catalyst. I've seen a US patent application granted in 1946, "DL-tocopherols and process for the manufacture of same", to Paul Karrer and Otto Isler for the synthesis of vitamin E, assigning the patent rights to Hoffman-La Roche Inc. The patent application actually lists 10 different synthesis routes for vitamin E. Very thorough. I don't particularly want to read them all. We can make vitamin E though, one way or another. There are total syntheses for TMHQ and for phytyl-bromide. I think modern synthesis switched from using phytol at one point in the reaction to using isophytol, but we can make both. It's fine.
What does vitamin E do? Seemingly not much, despite being essential. Some antioxidant stuff. There isn't any famous disease associated with Vitamin E deficiency. It's pretty hard to not get enough vitamin E while eating a Terran diet. Deficiency eventually results in neurological and muscular degeneration, and then death. So let's make some in space. Moving on.
: Vitamin K
Vitamin K is important for blood clotting and is perhaps useful for bone development too, and both of these functions are tied to vitamin K being necessary for proteins to bind to calcium ions. Four independent groups made one vitamer or the other in 1939, with Paul Karrer among them, I believe. The chemical isophytol, used in vitamin E synthesis, is used to make vitamin K also.
: B vitamins
I don't know why the B vitamins all have the same letter, as if they were equivalent vitamers that did the same thing. They're not and they don't. We need a bunch of different ones. As best as I can tell, Elmer McCollum found that there was a fat-soluble essential nutrient, which he called A, and a water-soluble essential nutrient, which he called B, and then better scientists figured out that B was a bunch of things. Vitamin C is also water-soluble though. Why isn't that called B0 or whatever? And why do the fat-soluble vitamins get different letters? I blame Elmer McCollum. The names are dumb, but the chemicals are important. Let's have a look.
: Vitamin B1
Structure: B1 is thiamine and nothing else. Thiamine, as the name hints, contains sulfur ("thio-" in chemistry) and an amine group. More specifically the sulfur is in a thiazole ring, the amine is hanging off a pyrimidine ring, and the two rings are connected by a methylene bridge. Ok, there's a little more junk: there are some CH3 methyl groups hanging off randomly, and an ethanol chain on the thiazole ring.
Function: Thiamine does a ton. One thiamine derivative, TPP, is part of the citric acid cycle and is a cofactor for at least six enzymes. And there are a bunch of other thiamine derivatives that do their own things. Generally, they help to break down sugars and amino acids for energy - destructive metabolism or catabolism (as opposed to anabolism). There are multiple diseases associated with thiamine deficiency, but beriberi is the famous one that was involved in the discovery of thiamine. Thiamine is depleted pretty quickly from the body, and more quickly if you're more active.
Synthesis: Thiamine was synthesized in 1935 by Robert Runnels Williams and his group, and he published about in 1936 with J. K. Cline as a coauthor. The first paper I looked at mentioned using ergosterol, a vitamin D2 provitamin (functionally, it's like the fungus version of cholesterol), and I wasted a lot of time trying to figure out how to make ergosterol from the carbon skeleton that's common to all sterol compounds, 1,2-cyclopentanophenanthrene. But I've since seen multiple papers describing the Williams-Cline synthesis without ergosterol, and that makes total sense because ergosterol isn't even that close to thiamine. Page 8 of "Vitamins B1, C and E" describes the Williams and Cline synthesis in detail. You start with ethyl 3-ethoxy propionate and then add an aldehyde group from ethyl formate. Then you close that up into a pyrimidine-like ring and swap out some of the bits dangling off with better bits. Finally, you just directly add in the thiazole ring using 4-methyl 5-hydroxyethyl thiazole. Ethyl 3-ethoxy propanoate is made by adding ethanol to ethyl acrylate. The 4-methyl 5-hydroxyethyl thiazole is a little trickier. It can be made by reacting thioformamide with 2,3-dichlorotetrahydro-2-methylfuran.
The first thing, thioformamide, is made by reacting formamide with phosphorus pentasulfide. Formamide is made by heating up ammonium formate, which is made from formic acid and ammonia,.
This isn't a good way to make that second thing, but it's the first method I found, pieced together from a depth-first search: 2,3-dichlorotetrahydro-2-methylfuran can be made by chlorinating 4,5-dihydro-2-methylfuran. 4,5-dihydro-2-methylfuran can be formed by cyclization of 5-hydroxy-2-pentanone. 5-hydroxy-2-pentanone can be made by hydrolyzing 2-acetyl-gamma-butyrolactone with sulfuric acid. 2-acetyl-gamma-butyrolactone is made by condensing gamma-butyrolactone with any old acetic acid ester, like ethyl acetate or butyl acetate. Gamma-butyrolactone is made by hydrogenation of maleic anhydride. Maleic anhydride is made by the oxidation of benzene or butane. Ouch.
: Vitamin B2
B2 is riboflavin and nothing else.
Structure: Riboflavin is a flavin molecule with an attached ribose sugar, you might not be surprised to learn. Flavins are all structurally based on a molecule called isoalloxazine, which has three rings right next to each other. Isoalloxazine kind of reminds me of an ant with its three body segments. And the molecule even has some legs and two antennae sticking off. It's a pretty good likeness. All the flavin molecules have the same ant body, and there's just a different R-group sticking off of the middle ring. In riboflavin, there's a ribose group, but you kind of have to squint to see it. The ribose is in its reduced radical form, ribityl. What does that mean, "reduced radical"? It sounds pretty cool. Ribose has a carbon with a double bond to an oxygen atom on one end. If you make it a single bond, then the carbon and the oxygen can each bond to another hydrogen atom. That new molecule with two hydrogen atoms added is the reduced or hydrogenated sugar, ribitol. If you pull off the OH hydroxyl group that you just made, then the molecule is called ribityl, and it has a free valence electron, i.e. it's a radical, and so it can be an R group sticking off of an ant's thorax.
Synthesis: Riboflavin was first synthesized in 1935 by both Paul Karrer and Richard Kuhn. I'm having a little difficulty figuring out which papers of theirs from 1935 on the synthesis of riboflavin (called lactoflavin at the time) were their respective first announcements. Not only are the papers in German, which I don't really understand, but they're also chemistry, which I don't really understand, and I'm just looking at the abstracts. Richard Kuhn coauthored with Friedrich Weygand, and maybe Karl Reinemund, Rudolf Ströbele, and Heinrich Trischmann, depending on the paper. Paul Karrer coauthored with Fritz Benz and Kurt Schöpp, I think. All of the papers just use initials for the first names, or at least that's how they're reported online, which made for kind of an interesting historical investigation (and not just for this vitamin). But I think people should know, or be able to easily find out, the full names of the chemists who figured out how to make essential nutrients. I still don't know which group did it first or published first. And maybe that's okay. They can all get credit.
Riboflavin is currently made industrially by bacterial fermentation, but Karrer and Kuhn did it without bacteria. You need a sugar to do it their way, which is a little hiccup for total synthesis. Karrer and Kuhn both used ribose, as you'd expect, but Karrer at least seems to have synthesized his D-ribose from D-arabinose. Weygand did a later synthesis directly from arabinose in 1940, rearranging it in the middle of the synthesis by an Amadori rearrangement (adding a little acid and heating the thing up to induce isomerization of the sugar). Arabinose was (and probably still is) much cheaper on earth than ribose. In space, though, if you're making sugars by Kiliani-Fischer synthesis, I think ribose and arabinose would have similar production costs, as they're both 5-carbon aldehyde sugars.
Anyway, to make riboflavin in a similar way to Karrer and Kuhn, you first condense 3,4-xyline with d-ribose. Condensation means the reaction produces water when two molecules join up. And then you reduce the new thing over a metal catalyst. And then you get tired of writing and you and pick it up on another day. //
: Vitamin B3
Vitamin B3 has two vitamers. The first is called niacin or nicotinic acid. It's structurally very simple; take a pyridine ring and then add on the carboxylic acid group COOH in the right spot in place of a hydrogen atom. If you pull off the OH from the COOH and stick on an NH2 amine group, then you've got nicotinamide, the second vitamer of B3. It's incorporated into a dinucleotide (nicotinamide adenine dinucleotide or NAD) that's very important in cellular metabolism. If you don't get enough niacin, you develop pellagra and then die.
Synthesis: Niacin is made by the oxidation of 2-methyl-5-ethylpyridine with nitric acid. 2-Methyl-5-ethylpyridine is made from paraldehyde and ammonia over an alumina catalyst. Paraldehyde is made by mixing acetaldehyde with sulfuric or nitric acid. Pinacol Rearrangement is the name for the acid-catalyzed dehydration of glycols to make aldehydes and ketones, and ethylene glycol can be Pinacol-rearranged into acetaldehyde using dilute sulfuric acid (H2SO4). Finally, ethylene glycol is made from water and ethylene oxide.
Is B3 really a vitamin? There is a metabolic pathway that breaks down the essential amino acid tryptophan to form the B3 nucleotide NAD. It's called the kynurenine pathway, because kynurenine is an intermediate chemical. The pathway relies on two other B vitamins, namely B2 (riboflavin, in the form of the nucleotide FAD) and B6 (pyridoxal, in the form of the nucleotide PLP), and you also need to be getting enough iron to do it, because at least two of the enzymes involved, TDO and IDO, are made with the iron-containing compound heme, which you might also know as a component of hemoglobin. The pathway is also pretty inefficient: you need about 60 mg of tryptophan to get the same level of B3 metabolites in your urine as if you'd taken 1 mg of niacin. I think that poor ratio alone - which helps to demonstrate that your body can't make B3 fast enough from other dietary sources - is grounds to call B3 essential as a vitamin. Also, excess tryptophan that isn't being used to make proteins in your body normally gets used for making the neurotransmitter serotonin, so niacin deficiency can mess up your brain chemistry by using up serotonin's precursor chemical. That might also be part of why B3 is called a vitamin.
: Vitamin B5
B5 is pantothenic acid. It's made from pantoic acid and beta-alanine. It makes up two of the five parts of Coenzyme A, which is involved in the metabolism of basically everything. The first synthesis was reported by scientists at Merck & Co. in 1940, including among them Stanton Harris and Karl Folkers, who are also known for the first synthesis of vitamins B6 and B7.
Here is their paper, which I haven't read yet: https://pubs.acs.org/doi/abs/10.1021/ja01864a039
And here is a modern paper on pantothenic acid synthesis, just in case we've gotten way better at it since then: https://link.springer.com/article/10.1007/BF00182178
: Vitamin B6
There are a bunch of vitamers of B6. They/re all pyridine derivates, meaning they have a 6-sided ring made of five carbons and one nitrogen. The first B6 vitamer synthesized was pyridoxine in 1939 by Stanton Harris and Karl Folkers. Going around the pyridine rings of pyridoxine, the carbon atoms have some dangly bits. First there's a CH3 methyl group, then an OH hydroxyl group, then two CH2OH hydroxymethyl groups, and the last carbon just has a hydrogen atom. So the full name of pyridoxine is 2-methyl-3-hydroxy-4,5-di-(hydroxy-methyl)-pyridine.
Here's the Harris-Folkers method. You first condense ethoxy acetylacetone with cyanoacetamide, forming a certain ring structure, 3-cyano-4-ethoxymethyl-6-methyl-2-pyridone. The condensation is performed in a solution of ethanol with a piperidine catalyst. Piperidine is just pyridine with extra hydrogen atoms, and it's made by reacting pyridine with hydrogen. You then take the ring molecule with the long name and you nitrate it with nitric acid (HNO3), then chlorinate it with phosphorous pentachloride (PCl5). Then you reduce it (react it with hydrogen gas) twice (using a platinum catalyst in ethanol first, and then a palladium catalyst in acetic acid). Then you treat it in sequence with hydrochloric acid, nitrous acid, and hydrobromic acid. Finally, you throw in some silver chloride, just for good luck, and out pops pyridoxine in perfect 100 mg tablets.
: Vitamin B7
B7 is biotin. It's a weird-looking molecule. It has two pentagons, some nitrogen, a sulfur atom, and a little rat tail.
The first synthesis of biotin was reported in 1943 by Stanton Harris, Donald Wolf, Ralph Mozingo, and Karl Folkers. In '44, they published another paper with all the details of the synthesis: https://pubs.acs.org/doi/pdfplus/10.1021/ja01238a041
They start from L-cystine, chloroacetic acid, and glutaric acid. We can make those. We'll be alright.
: Vitamin B9
The first B9 vitamin synthesized was folic acid, called pteroglutamic acid at the time. It has three parts. There is 1) a group with two hexagonal rings and lots of nitrogen called a pteridine, 2) a molecule of para-amino-benzoic acid (called PABA for short. Para-amino-benzoic acid is just a benzene ring with an NH2 amine group sticking off one end and a COOH carboxylic acid group sticking off the opposite end), and finally a molecule of glutamic acid, one of the inessential alpha-amino acids used to make proteins. The paper announcing the synthesis, "Synthesis of a Compound Identical with the L. Casei Factor Isolated from Liver", was published in 1945 by Robert B. Angier and 15 coauthors, including among them the great Yellapragada Subbarow. Most of the paper is about how they figured out the structure of folic acid from degradation products. The synthesis description is very short: combine 2,4,5-triamino-6-hydroxypyrimidine, p-aminobenzoyl-L-glutamic acid, and 2,3-dibromopropionaldehyde in the presence of an acetate buffer. And we all know how to make those things.
: Vitamin B12
B12 is the largest and most complex of the vitamins. The non-enzymatic synthesis of B12 was a very impressive project in the history of chemistry. It took many years and many people to make almost nothing. We can not make it at an industrial scale. We can not make it at a personal scale. Practically, we can not make it at all. We can harvest it from bacteria, and that's it. It's not made by plants or fungi. It's only ever been made by bacteria and one species of animal - the human chemists who participated in the Woodward-Eschenmoser project. Fortunately, we need very little of it and the body is great at recycling it through enterohepatic circulation. Even if your bacterial culture for B12 synthesis dies, it wouldn't be hard to bring enough with us for a long, long time.
I do have a silly idea for how we could get our bodies to make B12. It relies on an expanded definition of "our bodies". Some B12 is produced by bacteria in our large intestines, whereas we absorb it in the ileum, the last segment of our small intestines. Wouldn't it be cool if we could extract B12 from feces? Not good for the public image of the space program, but cool. Supposedly, most of the B12 in feces isn't even bioactive in humans, but still. Alternatively, and less gross, we could make some edits to our genes so that our large intestines could absorb B12, where it's produced. B12 can't be absorbed without a glycoprotein called "intrinsic factor" (awesome name) which is made in the stomach. But what if we changed it so it was made in the colon too? And also why are there no B12 producing bacteria in the ileum? Maybe we could fix whatever's getting in the way of that. All of these workarounds require a healthy B12-producing culture of gut bacteria, which is not something you want to stake your life on. I think we just have to admit that the chemical diet is technically inadequate without some bacteria for this one thing. Even though humans theoretically made it non-enzymatically once. Sad.
A bunch of bacteria can make it though! Most of them are associated with diseases and food spoilage. I was wrong about this for a long time, but now I think only bacteria (and archaebacteria) can make B12 from scratch, but they all make it in a form that isn't bioavailable to humans (pseudocobalamins) and then various non-bacterial microorganisms can transform pseudo-B12 into a form that they and humans can use, and the differences are basically just what molecule gets stuck on the lower ligand Like there's a microalgae called Chlorella that makes bioavailable B12, but I think it need a bacterium to make pseudo-cobalamin for it first. Chlorella's really cool. It is itself a food: it's a photosynthetic microalgae that's already used on earth as a nutritional supplement and protein source. Some other nice non-disease causing things can help make B12, like some lactic acid bacteria, but I think it's a form that isn't bioavailable to humans. Like Lactobacillus coryniformis subsp. coryniformis can make a version of B12 that has adenine in the lower ligand instead of dimethlbenzimidazole, and then Chlorella is good at putting the dimethlbenzimidazole on. Stick around for more in the microbial food synthesis section.
Choline:
Sources vary in calling choline an essential nutrient or a conditionally essential nutrient. The human body makes some. Does it make enough? The (U.S. based) National Academy of Medicine, the Linus Pauling Institute, and the National Institutes of Health all call it essential. Maybe the sources calling it conditionally essential are repeating the pre-1998 institutional opinion. I think infant rats fed low-choline diets die pretty soon, while adults live, but with some serious health problems, like liver cancer. I've never heard of a case of a human dying from diagnosed choline deficiency, but they definitely develop clinical illnesses. Whether it's a commonly called a vitamin or not, it's a great idea to get some in your diet (though it's also quite pssible to get too much).
Choline does a bunch of things in the body. 1) It's combined with acetate to form the neurotransmitter acetylcholine. Acetylcholine, in addition to being a neurotransmitter, is also a vasodilator, and that might be related to why choline deficiency eventually causes kidney problems. Acetylcholine, in addition to doing brain stuff and vein stuff, is also what allows your muscles to contract - skeletal and smooth muscles both. 2) A metabolite of choline, trimethylglycine, is used as a methyl group donator to recycle the essential amino acid methionine from homocysteine (an intermediate in cysteine synthesis). 3) Cell membranes are composed of phospholipids, and at more than 50%, the most abundant phospholipids, the phosphatidylcholines, incorporate choline (and so do their derivatives, the lysolecithins). One symptom of choline deficiency is an increase of certain enzymes in the blood, probably leaking out of cells with compromised structural integrity, especially muscle cells. This kills the cell. 4) The liver also needs phosphatidylcholines to export VLDLs (very-low-density lipoproteins) which transport fatty acids through blood vessels to other parts of the body. One of the first symptoms of a choline-deficient diet is the accumulation of fat in the liver, known as fatty liver disease or hepatic steatosis, which actually isn't particularly serious at first, but it can progress to cirrhosis. So let's make dietary choline.
Choline is a pretty small molecule. It has a nitrogen atom with some CH3 methyl groups sticking off and also a little ethanol chain. It's an ion, and it's usually manufactured as choline chloride or choline hydroxide. Choline hydroxide is made by mixing ethylene oxide and trimethylamine in water. Trimethylamine and the other two methylamines (mono- and di-) are made simultaneously by reacting methanol and ammonia over a catalyst, usually amorphous silica-alumina. And then you have to extract the TMA to make choline. But that's fine. Choline is on the menu.
:: Essential chemical elements, water, and gaseous oxygen
Finally, there are some elements that we need. Most of them we just need in trace amounts. Lots of nutritionists call them minerals instead of elements. As someone with a love of mineralogy, please don't ever do that. I don't think there will be any problems related to essential dietary chemical elements on Mars. We can recycle what we bring, and we can acquire more in place to support growing infants.
Carbon, hydrogen, and oxygen make up 98% of the atoms in the body. Those can be gotten on Mars from CO2 in the atmosphere and H2O, which probably exists as underground ice in lots of places on Mars, and definitely exists in hydrate minerals. A further 1.1% of the human body is nitrogen, which exists as both nitrate minerals in Martian regolith and in small amounts as N2 gas in the Martian atmosphere. Almost all the other essential elements are needed in such small amounts, while Mars seems to have such a great mineral diversity - by which I mean real minerals, the ones that form rocks, the ones with definite crystalline structure - such a great mineral diversity that I struggle to imagine trace element acquisition being a problem for astronauts. It would be prudent to think a little bit about how to get enough nitrogen, phosphorous, and sulfur from Mars. And here's how I think we'll get them: nitrates, phosphates, and sulfates. They're all over there, just waiting to be plucked. We'll be fine. Don't even worry about it, fam.
When it comes to essential trace dietary elements, the thing to focus our attention on is less the difficulty of acquiring more of the elements from Mars, and more the difficulty of recycling those elements from our waste and putting them into a form that's bio-available. For example, we know that sodium and chlorine are bio-available in the form of sodium chloride. That's an easy one. But what about iron? Iron supplements sold in stores contain salts like iron(II) sulfate or iron(III) citrate. There's also an iron(III) sulfate that isn't used as a supplement, and I don't know why one is used and not the other. Maybe there's a reason. Maybe we need to be careful about the oxidization state of the iron in our multivitamins depending on the anion to which it is bound. It's possible. On earth, I think people mostly get their phosphorous from the phosphate groups in nucleotides, like the ones that are polymerized together to form RNA and DNA. Fortunately, I don't think we don't need to make nucleotides from scratch to get phosphorous into a chemical diet: the common dietary supplements of phosphorous just contain dipotassium phosphate or disodium phosphate - quite simple salts. Some pyrophosphate salts are also used as food additives, such as to stabilize whipped cream. I wonder though if we'd be getting too much sodium or potassium if all our phosphorous was all coming from sodium or potassium salts. Probably not, but it's something to look into. In the same way that the value of a little vitamin C might be worth the cost of a Kiliani-Fischer sugar synthesis and purification, it might be worthwhile to synthesize sugar to make some sugar-phosphate molecule, like glucose 6-phosphate. There's a beautiful sugar-phosphate molecule called phytic acid that plants make, but humans can't digest it due to a lack of the enzyme phytase. Sad. For a simpler sugar-phosphate, there's also a metabolic intermediate called dihydroxyacetone phosphate or DHAP that's highly edible (LD50: 14,500 mg/kg). DHAP can be made from dihydroxyacetone, which was the 3-carbon sugar we talked about earlier that's used in sunless tanning lotions. Ooh, and there's another edible phosphorous source: a "glycerophospholipid" is a triglyceride with a phosphate group stuck on in place of one of the three fatty acids. That could work. Lots of options for edible phosphorous.
As for the bio-available forms of nitrogen, it's in all of the amino acids, and if you're getting your amino acids, my guess is that you're getting enough nitrogen too. But there's also nitrogen in nucleotides, which are a part of normal earth diets. I don't exactly know what typical fraction of earth diets nucleotides comprise. My guess is that it's a small amount, and a chemical diet without nucleotides can compensate by including more amino acids. Still, I should get some hard numbers on how small a portion of our diets they are before throwing them out with the bathwater. For dietary nitrogen supplements, I once saw a paper on supplementing sheep food with urea. Don't eat urea. Just eat your amino acids. It's fine.
As for bio-available sulfur, sulfur is found in two amino acids (methionine and cysteine) and also in some vitamins that we need in much smaller amounts (B1, B5, B7). My guess is that if you're getting those, then you're getting enough sulfur. There are two weird sulfurous chemicals marketed as dietary supplements, more for like joint paint than directly as a source of sulfur, and the public scientific consensus on both seems to be "don't", so let's go with that. One of them is cool for non-dietary reasons though: dimethyl sulfoxide is absorbed right through the skin and circulated pretty fast through the body. Some people taste garlic in their mouths shortly after getting it on their skin. How trippy is that? It's like if you flushed an airplane toilet and the pilot was ejected. Those systems just shouldn't be coupled.
There are a bunch of metals that we need. Calcium can be digested in the form of calcium carbonate, which is the main component of ash. And potassium salts were also first discovered as the water-soluble portion of ashes and are digestible in that form, I believe. At worst, you just have to treat the potassium salts in ash with hydrochloric acid to get potassium chloride, which is definitely bioavailable. So maybe incinerating your waste and eating the ash is how you get calcium and potassium. That doesn't sound right, but go on and tell me why I'm wrong. Magnesium is another light and highly reactive metal like calcium and sodium and potassium. If it's not precipitating spontaneously out of your sewage as struvite, I'm not sure what form it's in or how to recover it. In geology, the process by which magnesium atoms can slowly replace calcium atoms in calcium carbonate is called dolomitization. Maybe if you mix some of the ashes back in with the sewage before eating it, that's even better for element recovery. No, don't do that. Magnesium's actually pretty important. I should look into this. Magnesium supplements include the oxide, the hydroxide, and the chloride, among more complicated organic anions. I don't know why I was so worried about the bioavailability of these things.
Those are the 11 main elements in the human body. Arranged by decreasing percentage of body mass it's (O, C, H, N, Ca, P, K, S, Na, Cl, Mg). Hydrogen is the most plentiful by atom count, but also really light.
Many other elements are needed in tiny, tiny, trace amounts. There are so many needed in such small amounts that it's almost more of a victory for science to say definitely that we don't need an element than to say that we do. There are some transition metals we need (nickel, zinc, copper, chromium, manganese, molybdenum, cobalt). The cobalt we need is in the essential vitamin B12, and if you're getting enough B12, then it's taken care of. So worry about B12 and forget cobalt, basically. We also need some halides besides chlorine (fluorine, iodine, maybe bromine). For the remaining things to the right of the transition metals we need (boron and selenium). Selenium, essential, is found in one of the inessential amino acids, selenocysteine. I think that's probably a good way to get it in your diet, but I don't know how to make it. Other bioavailable forms of selenium that are found in dietary supplements are selenomethionine, which I also don't know how to make, sodium selenite (Na2SeO3), and sodium selenate (Na2SeO4). The selenite salt can be made by reacting sodium hydroxide with selenous acid (H2SeO3) or the acid's anhydride, selenium dioxide (SeO2), which are made by oxidizing selenium with nitric acid. The selenate is made by oxidizing the selenite with hydrogen peroxide, but one salt doesn't appear to be absorbed more readily than the other, so stick with the simpler sodium selenite. To round out the essential alkali and alkaline metals (Na, K, Ca, Mg), we also seem to need a little lithium. Salts like lithium carbonate, lithium chloride, and lithium acetate are used medically and are easy to make.
I do not know how to get most of those elements out of sewage and into people. Electrowinning, chelation, mineral acid dissolution, ... no. It's just too much to recover simultaneously. It would be nice and lucky if the remaining trace elements just came for free in the recycled water. It probably depends on how you're recycling your water. I'm not too worried about it, honestly. If those elements tend to get lost in the recycling machinery, we can bring some multivitamins with trace elements for the trip and acquire more elements when we get to Mars. Mixing sand with water is not a hard thing on Mars. The hard part is making sure that the fortified water is not also fortified with toxic things like perchlorate salts and reactive oxygen species. // Hah, a few paragraphs down, something I wrote last week, you can see that I was super excited about recovering trace elements from sewage. I guess I'll let you read that too for now. It's too late to do any more editing tonight. I'll consolidate the content of both paragraphs another night.
I'm not sure whether to call gaseous oxygen and liquid water essential nutrients, but regardless of whether they get their own named section in this essay, we do need them, and we should plan to recycle them. I'm not sure what the best way is to get O2 gas from waste CO2. I don't think it works to heat it up: it takes a lot of heat for the CO2 just to dissociate into CO and O2 (2 CO2 = 2 CO + O2), and the CO doesn't undergo thermal decomposition readily either. Electrolysis of CO2 might work better than heat though, with the same overall reaction. And then the Boudouard reaction (2 CO = CO2 + C), performed over certain metal catalysts, turns CO into carbon dioxide and solid carbon, as little graphite flakes. You could keep alternating electrolysis and Boudouard to pull carbon out of CO2 as graphite and thereby free some O2. If you're on Mars, instead of on the way to Mars, I suppose there's less need to recycle all the oxygen from your waste CO2. You could electrolyze CO2 to generate CO and O2, then vent the CO, and then take in new CO2 from the thin local atmosphere to compensate. That's the design used in NASA's MOXIE experiment. For every input molecule of CO2, Moxie wastes half of the oxygen and all of the carbon. So it's not materially efficient, but maybe it's economical in terms of energy use. Thinking back more along the lines of recycling, artificial photosynthesis would work nicely, at the cost of some consumed water, but we're not there yet technologically. I think a good method to recycle O2 from CO2 might be to turn the CO2 into methane by reacting it with hydrogen gas, H2. That's called the Sabatier reaction: (CO2 + 4 H2 = CH4 + 2 H2O). It's done at high temperatures over a nickel catalyst. After that, the water can be split by electrolysis to make O2, and in the process, we get back some of the initially supplied H2. If you want to get the carbon out of the methane and into a solid form, and regenerate the rest of the input H2, I think that the thermal decomposition of methane is much, much easier than thermal decomposition of CO2. You might even be able to do it by just increasing the temperature in the Sabatier reaction vessel. If you'd rather bind the carbon up into useful hydrocarbons right away, methane has some direct uses: It's natural gas fuel of course, but burning it would release the carbon that we just sequestered. For applications that don't release the carbon, methane can apparently be fed to a methanotrophic bacterium called Methylococcus capsulatus to make animal feed. But that's a topic for the microbial synthesis section. If you want larger hydrocarbons, you can start with the same reagents as the Sabatier reaction and instead do a water-gas shift reaction (CO2 + H2 = CO + H2O) followed by a Fischer-Tropsch synthesis, which will give you saturated alkanes that are the starting materials for most of the chemical recipes I've given for nutrients. So like you start with hydrocarbons from Fischer-Tropsch and make them into carbohydrates, which contain oxygen, but less oxygen per carbon atom than CO2, leaving some oxygen behind by mass balance. That's how you get O2 from CO2, if you're limited to physicochemical methods. Maybe. I don't actually know much chemistry.
As long as we're talking about gasses, what about nitrogen? How do we recycle waste nitrogen? Well, if it's nitrogen in the form of ammonia, or urea (which decomposes to ammonia), it's just about ready for use in Strecker synthesis of amino acids. If it's nitrogen gas, N2, the Haber process will turn it into ammonia. I hear that before we had the Haber process, we generated some of our ammonia by dry-distillation of nitrogen-containing biosolids. I think feces contains like 30% of the nitrogen we consume. Maybe we could pyrolyze feces, extract some ammonia from the effluent gases, and leave most of the solids in the form of charcoal, which has many uses. That would be cool. You need gas separation tech for all this. But we can do that.
What about water? Water recycling is a huge necessity in space. Even for a little two year Mars mission where you can bring your own food, you have to recycle your water, with a very high rate of reclamation. Now, dewatering sewage and then purifying the water for consumption is a problem that every wastewater treatment facility on earth partially solves. They use physicochemical methods like settling tanks, centrifuges, and press filters alongside microbial digesters. Water recycling in space has higher standards though, and I should write a lot more about how it's done. I think osmotic filters and vacuum stills are particularly important. The Yuegong-1 (Lunar Palace 1) closed ecosystem research program also uses activated charcoal filters which are a good idea.
...
I'm not sure about recycling sulfur and phosphorous. There are a few phosphate minerals that precipitate fairly readily from sewage solutions: the main two are struvite (NH4MgPO4 . 6H2O) if there's lots of magnesium and vivianite (Fe(II)3(PO4)2) if there's lots of iron and a reducing chemistry. Also concentrated urine brine contains lots of phosphorous, possibly in the form of stercorite (Na(NH4)HPO4). That's actually where phosphorous was discovered - the first modern discovery of a chemical element happened when the alchemist Hennig Brand boiled his urine. I think that probably dewatering and phosphorous recycling will be intimately connected. Phosphorous will be removed from osmotic filters to prevent fouling, or it will be left behind after distillation and evaporation, or something like that. I have even less idea about sulfur. I don't even know if it's more likely to be dissolved in water or to be escaping as gasses like H2S and SO2. In wastewater streams, I think chlorination can reduce H2S to solid elemental sulfur. I guess I have some reading to do. In the petrochemical industry, flue gasses are desulfurized in a few different ways. You can burn H2S to form SO2, and then oxidize SO2 to SO3 over a vanadium catalyst, and then bubble that through water to make sulfuric acid. That's the wet sulfuric acid process. You'd still need to get the sulfur into the form of sulfurous amino acids: methionine and cysteine. Okay, sulfur recycling still needs some significant attention. //
Besides sulfur and phosphorous, I'm also interested in the problem of getting metals and inorganic salts out of sewage by physicochemical means. Some metals can be extracted from sewage solutions by applying an electric current and letting the metals accumulate on the electrodes. That's called electrowinning. I'm not sure what else to do. Incineration of dried sewage sludge leaves you with some metal in the ashes and some metal in the flue gasses. That could be a useful step toward metal extraction. Another possible idea: Robert Zubrin, an aerospace engineer who has perhaps done more work advancing Martian colonization than anyone else, has an interesting proposal for recycling waste by steam reforming. It breaks complex molecules of carbon, hydrogen, and oxygen down into simple gasses. The proposal doesn't mention it, but it seems to me that separating out the vast bulk of C, H, and O from sewage sounds like a great step toward isolating the small remaining fraction of heavier elements that we need in small amounts. Steam reforming seems like a big, big waste of energy to me when compared to microbial digestion of wastes, but it's still an interesting proposal as physicochemical methods are concerned. It might have some benefits over incineration.
Update: It's not mentioned in Zubrin's proposal, but what he's doing is usually called gasification, while steam reformation refers specifically to gasification of methane. Gasification leaves some sludge behind - it's not a perfect process - but it really is a pretty cool thing. Zubrin's not wrong, he's just bad at sharing credit for ideas by calling them by their standard names I guess? Don't really know what's going on there.
: Inessential food chemicals (nucleotides, carnitine, inositol, taurine, cholesterol, fiber, solid binders, flavorings, PABA, lipoic acid, no flavonoids, no Coenzyme Q10)
The goal of this diet is to sustain astronauts while they do cool things on Mars, not to test whether the essential nutrients named by biologists are actually fully adequate for humans. If normal diets include nucleotides, for example, and we can make nucleotides non-enzymatically, and we have reason to believe that they're beneficial to humans when ingested, then it makes sense to include them in the diet, even if they're not technically essential. It's prudent to not test the absolute metabolic limits of the people you're trying to sustain. It might even be economical: if the body makes nucleotides out of amino acids (which it does), and, hypothetically, it's cheaper to make nucleotides than amino acids, then we could save ourselves some expensive amino acid manufacturing. Hypothetically. So how about it? Which inessential food chemicals can we make? Not nucleotides, actually, because they have a sugar subunit, and we're bad at making sugars. But they also have a nucleobase subunit and some phosphate groups. We've already tackled the problem of bioavailable dietary phosphates. How about nucleobases? Can we make those from scratch? I don't know. I'll look into it soon and put the answer here. And other things that might be conditionally essential. There aren't many. We're almost done.
Nucleobases (adenine, guanine, cyotsine, thymine, uracil):
Adenine: There is no vitamin B4 right now, but the nucleobase adenine is one of a few things that used to be called vitamin B4. Our bodies can synthesize it in requisite amounts, so it's not called a vitamin anymore. The nucleobases adenine and guanine are both variants on a molecule with lots of nitrogen called purine, while the other three nucleobases are pyrimidine derivatives. Actually, all five nucleobases contain pyrimidine, because the purine substructure of adenine and guanine is a pyrimidine ring bound to an imidazole ring. Adenine in particular has the purine structure (pyrimidine + imidazole) and an NH2 side chain. A chemist named Joan Oró, one of the giants in the field of prebiotic chemistry which studies the origin of life, reported in 1960 that adenine could be made by heating ammonia, hydrogen cyanide, and water for a day or two ("Synthesis of adenine from ammonium cyanide"). He kept researching the reaction with his colleagues and was soon making adenine in decent quantities. An insoluble black crust of polymerized hydrogen cyanide is formed in the reaction and that can be removed from the liquid using a centrifuge. The same HCN polymer might be why Haley's comet is black! Adenine is then isolated from the remaining liquid by chromatography. Easy peasy. Should we be eating it? I'm not sure. Seems kind of toxic. In a study on rats ("The toxicological effects of pure dietary adenine base in the rat"), the oral LD50 was set at just 227 mg/kg. Below a tenth of a percent of their diet, the rats appeared fine, and then above that their kidneys suddenly became swollen and the level of uric acid in their blood spiked. That's as far as I got in the paper, but it's not hard to guess what happened. The enzyme that recycled adenine into nucleotides is adenine phosphoribosyltransferase. The enzyme also takes a molecule with ribose and lots of phosphate groups, phosphoribosyl pyrophosphate, and produces from them a nucleotide called AMP, adenosine monophosphate. When people are deficient in the enzyme, adenine is degraded by different metabolic pathways rather than being recycled, and the result is the accumulation of uric acid salts in the body and the eventual breakdown of the kidneys. So an excess of adenine symptomatically looks just like a deficiency of the adenine-recycling enzyme. Okay, so we shouldn't make adenine a big part of our diets. Should it be a small part? Most vitamins are a small part. Is there a quantity that would be beneficial? I will look into it. //
Guanine: Emil Fischer won a Nobel prize in 1902 for the synthesis of sugars and purines. Kiliani-Fischer synthesis is the classic method of non-enzymatic synthesis of sugars, but I haven't yet found his work on synthesizing or interconverting purines. However, there are only two more purines that interest me for this essay, namely guanine and caffeine, and we have methods for those that don't have Fischer's name on them, even if perhaps he deserves first credit. Here's a short, one-pot synthesis of guanine: https://patents.google.com/patent/CN1966504A/en . It doesn't get much better than that.
Cyotsine:
...
Thymine: Thymine sounds like the vitamin thiamine, but it's a differen thing. There's no sulfur in this one. Rather thymine is a pyrimidine nucelobases that was first discovered in an animal's thymus, which gland is shaped like the bud of a thyme plant, reportedly. Structurally, thymine is uracil with an extra CH3 methyl group, and I've seen a synthesis from uracil: treat uracil with formaldehyde and hydrochloric acid to make uracil-5-methyl chloride, then reduced that with tin and hydrochloric acid to give thymine. That's a pretty old synthesis, but I think the first reported sythesis was "Synthese des Uracils, Thymins und Phenyluracils" by Emil Fischer And Georg Roeder in 1901, possibly using cyanoacetic acid and urea. Almost all the non-enzymatic methods use urea at some point. In the body, thymine is made from the amino acids serine and glycine. The most common modern industrial synthesis route for thymine uses ethyl alpha-formylpropionate (EFP). EFP is combined with urea under an acid catalyst (an example of Claisen condensation), and then you add in some sodium methylate to induce ring closure, and thymine is formed. The EFP is made from ethyl formate and ethyl propionate (over sodium metal in an N2 atmosphere). If you don't want to do chemistry over sodium metal in an oxygen-free atmosphere, Xianghai Guo and Jiaxiang Shen have developed a new route for thymine ("An environmentally benign approach to the synthesis of thymine via hydroformylation of methyl acrylate", 2014) that combines methyl alpha-formylpropionate with urea and then introduces sodium methanolate for ring closure. The methyl alpha-formylpropionate can be made by hydroformylation of methyl acrylate, without sodium metal. Easy.
Uracil: We can make uracil by combining malic acid with urea under a catalyst of fuming sulfuric acid. Malic acid is made by hydrating fumaric acid, which is made by an isomerization of maleic acid, which is made by hydrating maleic anhydride, which is made by oxidizing benzene or butane.
That's it for nucleobases. Let's address some other inessential or conditionally essential nutrients.
Carnitine has a carboxylic acid group and an amine group. I vacillate on whether to call it an amino acid. If it's not an amino acid, it's a damn fine imitator. It's almost always drawn in a form called a "zwitterion", with a COO- instead of COOH, but lots of organic molecules including amino acids are zwitterions at biological pH. That's not what would keep carnitine from being an amino acid, I don't think. Derivationally, carnitine is a dipeptide (an organic molecule made from the leftover residues of two amino acids). But you can derive some amino acids from other amino acids. Carnitine's status as an amino acid derivative doesn't tell you whether it's an amino acid structurally. Carnitine is usually not called an amino acid, maybe because of the quaternary structure of the amine group. I dunno. I'm sticking with convention and talking about carnitine here, instead of as a conditionally essential amino acid in the amino acid section. L-carnitine is needed for fatty acid metabolism by cellular mitochondria. Healthy adults can make enough of it in their liver and kidneys from the amino acids lysine and methionine, provided they're getting enough iron and vitamin C, but infants born preterm sometimes can't make enough of it quickly enough. I can't help but wonder if preterm infants would be at less risk of carnitine deficiency if we fixed the broken mutated human genes that should code for the enzyme that makes vitamin C from glucose, but that's a conversation for when I become the gene therapy research czar of the solar system. Anyway, right now the only non-enzymatic carnitine synthesis I can find is in "Asymmetric synthesis of L-carnitine from (R)-3-chloro-1,2-propanediol" (Li et al, 2011). I thought I'd seen a different one, but oops, lost it. That chloro-propanediol thing mentioned is just glycerol with one of its OH groups swapped for a chlorine atom. It's also called 3-MCPD. And while glycerol is not chiral, 3-MCPD is. Sometimes chemists use an (R) or an (S) to distinguish chiral enantiomers. It's a more modern and less ambiguous system of notation than the L- and D- prefixes commonly used for amino acids and sugars, and it has lots of rules, but the important point here is that we need the (R) enantiomer of 3-MCPD to make L-carnitine. In the paper, they got the pure (R)-enantiomer of 3-MCPD by starting with pure (R)-epichlorohydrin, and doing a base-catalyzed hydrolysis. They even tell us how they got the pure epichlorohydrin: kinetic resolution. That's a technique for separating chiral molecules in a racemic mixture based on their reaction rates with a separate chiral catalyst. Anyway, I only read the abstract, but I have every reason to believe it was a simple and abiological synthesis. So carnitine is on the chemical food menu, if you want it. You know, to help preterm infants with the mitochondrial oxidation of fatty acids. It's actually a big deal; that's how your heart gets most of its energy, I think. You wouldn't break a premature Martian baby's heart, would you?
Inositol: Inositol is 6-carbon (hexose) sugar alcohol first extracted from muscles. Your body makes about a gram of it per day from glucose. In addition to whatever it does in muscles (I should find out), it's used to make some neurotransmitters. It might help to treat OCD. It was one of the chemicals once called vitamin B8, but the body can make enough of it, so it's not called a vitamin anymore. It can probably be made from a normal non-alcohol sugar some way. I don't know. If I find out, I'll add it here. But sugars are hard to make, so I'm glad inositol is inessential. It's probably more important than I'm giving it credit for here. More research needed.
Taurine: Taurine is weird to me. It's derived from cysteine and it contains cysteine's sulfur atom. Sulfur molecules in the body are weird enough as it is. It's made in the pancreas, but it's more associated with steroidal bile acids made in the liver (specifically taurocholic acid). It's widely distributed throughout the body and it's involved in cellular and mitochondrial signaling with calcium ions, particularly in neurons and muscle fibers, and maybe that's why it's also important for muscle development and maybe retina health. Vegans get by just fine with none of it in their diets, most people get like 50 to 300 mg per day from meat, and people who consume energy drinks might get 1,000 grams per serving or more, despite there being no apparent reason to put any of it in, except perhaps that it was first discovered in ox bile, and oxen are sometimes aggressive and that feels edgy to people? It's a really small molecule, and it's easy to synthesize: I've seen three abiotic synthesis routes. The toxicity reports all say > 5 g/kg or something like that, but no one seems to have pinned down how much of this weird sulfur-ammonia acid rats can actually stomach. It's added to baby formula because prematurely born babies sometimes can't adequately make cysteine from methionine, and taurine is a cysteine derivative, but the amount added is based on reasoning about as slim as that for the amount added to energy drinks. You can make taurine from ethanolamine, sulfuric acid and sodium sulfite. Ethanolamine is made from ethylene and aqueous ammonia.
Cholesterol: Cholesterol is not essential to the human diet. Some dietary guidelines recommend eating as little as possible. Your body makes what it needs. But we can make it non-enzymatically! Maybe not at a large scale, but we've done it more than once. It's not an easy synthesis, but it was a big achievement in the history of organic chemistry. Vitamin D is a sterol derivative, and I imagine the routes we have for vitamin D synthesis owe a lot to the synthesis of cholesterol.
Fiber, solid binders, and fillers:
When I started this essay, the chemical menu as far as I'd thought it through was just a bunch of non-toxic liquid alcohols, and I was worried that it would be hard to find something solid to eat. I'm not as worried about that now, since amino acid crystals are solid, and long-chain triglycerides are semi-solid, and even one of the alcohols I've found since (pentaerythritol) is solid. Pentaerythritol probably isn't food, but maybe another solid alcohol is.
Back then, I thought that clay and chalk (or powdered limestone) might have some limited use as dietary fiber. All the material safety data sheets I've seen on different clay minerals just give LD50s like >2,000 mg/kg or >5,000 mg/kg. Sidewalk chalk (aka gypsum or calcium sulfate dihydrate) is also listed imprecisely at >2,000 or >3,000 mg/kg, while real lime chalk (calcium carbonate) has been pinned down more precisely, at 6,450 mg/kg. Mars has clay and limestone. Those are options if you really want to eat rocks. Also, ashes are mostly calcium carbonate. If you're not on a rocky planet, you could technically still eat calcium carbonate in the form of ash. I've never seen the toxicities of wax or petroleum jelly pinned down precisely either, but those might work as non-toxic indigestible semi-solid binders.
There are some other indigestible materials with higher LD50s. The artificial sweetener saccharin is solid at standard temperature and pressure, and it has an oral median lethal dose in rats of 14,200 mg/kg. Aspartame and neotame are two other solid artificial sweeteners with high LD50s, but they're made from amino acids (which are already solid) and they're digestible as amino acids (so they're not useful as fiber). I've seen both silica sand and titanium dioxide listed at >10,000 mg/kg, which is kind of impressive. Some silicone putty compounds have LD50s around 20,000 mg/kg, if you're looking for that meat-like chewy mouthfeel. Polyvinyl alcohol (PVA) is a non-toxic but indigestible polymer that can thicken liquids, perhaps to a pudding-like consistency or thicker, depending on the concentration and the molecular weight of the polymer. Its LD50 is 14,700 mg/kg. I've seen another polymer, poly-N-vinyl pyrrolidone, listed at 100,000 mg/kg, which is really hard to believe because water's oral toxicity is only 90,000 mg/kg, but I still wouldn't be surprised if it were highly non-toxic at least. I think someone just typed an extra zero, and the typo got repeated in a few places.
Do we actually need fiber or solids? I don't know the current state of public reproducible evidence on whether dietary fiber is protective against colorectal cancer, but in grade school, they taught me that it was. Also, I think solidity is somewhat important, whether the solid ingredient is digestible or indigestible fiber, because chewing is good for mouth muscles, I think. I don't want anyone on a chemical diet to suffer disuse atrophy of their jaw muscles.
I'm not recommending eating sand and wax and silly putty, for anyone on any hypothetical diet. I'm not even recommending that Martian astronauts eat artificial sweeteners like saccharin and acesulfame-K. But while amino acids give us solid food, they don't give us fiber, and if dietary fiber turns out to be important to human health, then these are unfortunately the best ideas I currently have that can be manufactured by physicochemical methods. If we get better at polysaccharide manufacturing, then we'll have some more natural options.
Finally, charcoal is not digested, and its toxicity is quite low, with an oral LD50 of 15,400 mg/kg. I'm not sure which materials that we can make at scale are able to be charred, but I wouldn't be surprised if formose syrup was among them. Charcoal has a nice crunch, but it will also pull nutrients out of your digestive tract, which isn't ideal.
Oh hey, antidiarrheals and laxatives! Fiber does something like one of those, right? We can make those from scratch. Polyethylene glycol (PEG) is a hydrophilic polymer used as an osmotic laxative. It's indigestible, highly non-toxic (which is crazy, because ethylene glycol is very toxic), and made by combining ethylene oxide with water or with ethylene glycol, over an acidic or basic catalyst. If, instead of a laxative, you want a total synthesis of an antidiarrheal, bismuth subsalicylate is on the menu: it's made by the hydrolysis of bismuth salicylate, and salicylate is the conjugate base of salicylic acid, which is made in the Kolbe–Schmitt reaction which combines CO2 and sodium phenolate under high temperature and pressure, and then acidifying the resulting sodium salicylate salt with sulfuric acid. I wonder if you could use bismuth phenolate instead of sodium phenolate and save yourself from having to convert from salt to acid to salt again. Phenol/phenolate can be made simultaneously with acetone in the cumene process (the reaction of benzene and propylene under oxygen). There are also some attempts to make phenol more directly by oxidizing benzene or toluene, but they come with harsher reaction conditions or lower yields. We also used to make phenol from benzenesulfonic acid, but again, lower yields and more waste. All three of those ways of making phenol need benzene. Benzene and other simple aromatics (toluene, xylene) can be made by catalytic reforming of hydrocarbon liquids of 6 to 8 carbons, which I think just means heating liquid petroleum with hydrogen gas over a platinum catalyst. And liquid petroleum can be made in the Fisher-Tropsch process from carbon monoxide and hydrogen.
It seems like we can cover for most of the possible useful functions of dietary fiber, and maybe we just need to focus on finding a chemical we can synthesize that can feed colonocytes? Maybe colonocyte feeding is the last function of dietary fiber for which we need a functional substitute. And the main result of colonocytes feeding on dietary fiber is the release of butyrate in the large intestine. So maybe we could skip the fiber and the colonocytes and go straight to the butyrate? If it has a medical benefit, that is. I really don't know much about the health impact of fiber. It seems to improve the immune function of the intestinal mucosal membranes among other things, not all of which are good?
Oh hey, what about bio-degrable bio-plastics? Could we use those as fiber? Most of them are made from starch, which is hard to manufacture, but not all of them. Proteins can be used be used as a cross-linking agent, for example. PBAT (Polybutylene adipate terephthalate) is a biodegradable plastic that we can make from scratch. It's made from 1,4-butanediol, adipic acid, and dimethyl terephthalate. I don't really want to eat it. You can't make me eat it. PLA (polylactic acid) is another biodegradable plastic. It can be made by polymerizing lactic acid, which is made by reacting acetaldehyde with HCN and then hydrolyzing the resulting lactonitrile, for example with hydrochloric acid. That sounds much more edible, doesn't it? Delicious lactic acid. I could eat that, if I had to, maybe. PBS (polybutylene succinate) is another biodegradable plastic we can make from scratch. I don't know if these are specifically degradable by colonocytes. But that seems like a pretty easy experiment, right? Just poop on some plastic. I guess you'd also want to verify that the plastic will survive the digestive tract mostly intact until they get to the large intestine and that they have safe degradation products.
Flavorings: A diet with varied food flavors is both nice and normal. I think it would be worthwhile to see which flavoring compounds have very simple synthesis routes and at least moderately low toxicities. I'm going to look for things with LD50s no lower than 3,000 mg/kg. If you're making something inessential, it might as well be moderately non-toxic. We've already mentioned some organic acids in the carbohydrate section, some fruity and floral esters in the fatty acid section, and MSG n the amino acid section. Quick recap?
...
Now for some new chemicals. Manzanate (LD50 > 5,000 mg/kg) tastes like apple. It's more formally known as ethyl-2-methylpentanoate. Acetoin (LD50 > 5000 mg/kg) tastes and smells buttery. A good addition to Imhausen margarine or an amino acid shakes. Isoamyl acetate (LD50 = 16,600 mg/kg) tastes like banana. Look at that figure. What would we do without esters? Ethyl propionate (LD50 = 8,732 mg/kg). It's fruity and a bit like pineapple. Ethyl decadienoate, (LD50 > 5,000 mg/kg) tastes like pear. Benzaldehyde (cherry and almond flavor) is too toxic to include here, despite being used in lots of cooking. Sad, but I've got standards, yo. Cinnamaldehyde's LD50 is also too low. Methyl anthranilate, tastes like grape, not in the club. Capsaicin in chili peppers and piperine in black pepper and allicin in garlic are all too low. Table salt just makes the cut (LD50 = 3,000 mg/kg). Limonene (5,300 mg/kg) tastes like citrus. Delicious. Vanillin is too low. Methyl salicylate tastes like wintergreen, too low. L-menthol could work though (LD50 = 3300 mg/kg). Similar to peppermint. Zingiberene is in ginger, and I'm seeing wildly different numbers. We'll come back to it. Raspberry ketone, too low. Eugenol is spicy like cloves, too low. Eucalyptol is like mint, too low. Nerol (4,500 mg/kg) smells sweet, a bit like roses and greenery. The taste is a little more bitter than the smell, but it's used to make lots of things taste more natural, less chemical. Wine lactone is too low. Carvone, too low. A lot of these are terpenes, which might be hard to make. What else is there? Ooh, lots of median lethal doses for flavorants in this PDF: https://efsa.onlinelibrary.wiley.com/doi/pdf/10.2903/j.efsa.2011.2164. I'll go through that, find some good things, and then focus on synthesis. Who knows?; there might even be a new carbohydrate in there.
PABA: There is a chemical called para-aminobenzoic acid (PABA for short) that people normally get in their diets and it's also synthesized by bacteria in the colon. It used to be called vitamin B10. It's actually part of vitamin B9, folic acid. If the gut bacteria that make it are wiped out by antibiotics and you're not getting any PABA in your diet, there might be some small associated health problems. But we can synthesize it from scratch non-enzmatically, and in more than one way actually. If, for example, you nitrate toluene (like when you make TNT), then you're most of the way there. PABA is also produced by yeast, which will come up in the microbial food synthesis section. PABA as a dietary supplement is often sold by New Age folks on earth with exaggerated claims of medical efficacy, but there's nothing wrong in principle with being able to synthesize it in your space habitat, just in case.
Lipoic acid is another nutritional supplement touted as a cure-all by people who seem to latch onto beliefs precisely because they're treated with doubt by public scientific consensus. But again I think perhaps we should be prepared to make lipoic acid at an extraterrestrial habitat, especially one with a large population. Lipoic acid doesn't cure cancer, but it plays a big enough role in metabolism that I wouldn't be surprised if it was a conditionally essential nutrient for people with some metabolic disorder. And we're able to make it from scratch, so why not plan for it? The R-entantiomer is the form we need, and the body makes enough. R-lipoic acid is most notably a cofactor for one of the three enzymes that helps convert pyruvate into acetyl-CoA, thereby linking glycolysis to the citric acid cycle (which feeds NADH into another pathways which makes ATP). This patent (https://patents.google.com/patent/US2792406A/en) mentions a bunch of ways to make it in passing. It's not a fun read, but I'll pick an easy synthesis pathway and draw it up in a nice diagram. And here's another one.
Next up, the microbial space diet!
