Why's my drink greener?

[evo-lution]

At the weekend I worked some bartending shifts at one of my consultancies, Mim in Aberdeen, and found myself puzzled by the difference between two drinks that were practically identical save for the fact one was alcoholic and the other non-alcoholic.

A guest at the bar ordered two Mediterranean Fizzes, a cocktail on the menu that is available as both an alcoholic and non-alcoholic beverage, requesting both variants.

At the point of serving I noticed that the non-alcoholic variant had taken on a considerably darker colour of green compared to the drink with alcohol. I tried both to ensure I'd made them right and found no problem whatsoever and didn't really think any more of it. Luckily these guests ordered the same round a further three times and on each occasion I found the colour to be exactly the same, darker for the non-alcoholic drink, lighter for the alcoholic version.

The full text can be read over on my blog.

Would love to hear some thoughts on the colour difference and also on the darker colour gained when using metal-on-metal shakers.

[KD1191]

Well...I'm no scientist, but I did find these two links, which suggest to me that the non-alcoholic drink is greener because the chlorophyll in the plant matter has not been broken down or otherwise affected by ethanol.

ETA: There's a lot of talk at that first link about photons, "lower-energy" color and stuff that's way over my head, but it generally sounds like it could explain why chlorophyll dissolved in alcohol is 'less green'.

Edited July 26, 2011 by KD1191 (log)

[Sunny&Rummy]

I am a scientist, but alas, I am late to the conversation. I think KD1191 has hit this one right on the head. Provided that the green color in your cocktail is from a plant and not an artificial coloring in an ingredient, I think the solubility of chlorophyll in alcohol explains the visual difference between the two drinks.

Chlorophyll is highly nonpolar, and as such it is soluble in nonpolar solvents like ethanol. Laboratories take advantage of the solubility of chlorophyll in solvents like ethanol, hexane, acetone, etc., when they have to perform chlorophyll extractions.

[slkinsey]

Basil is especially affected by enzymatic browning and oxidation reactions. This is why fresh-cut basil will almost immediately darken along the cut line whereas this effect is not as pronounced with many other herbs. I suspect it's also especially noticeable in basil because the initial color tends to be quite light compared to, e.g., parsley or mint.

So there are a number of questions you raise here, and there are a number of variables I think you'd want to tease out.

For example, I'd start out shaking out samples that contain only basil and either water or water and ethanol. Is one sample different from the other? This then suggests that it's the ethanol alone that's responsible for the difference. If you are finding that one sample has a darker color, then this suggests that the enzymatic browning/oxidation reactions (or whatever it is that turns bruised basil dark) are more pronounced in this sample. It could be that this sample is more effective at bringing the various substances into solution where the reactions happen, or it could be that something in the other sample (presumably the ethanol) has some action that mitigates these effects.

The potential effect of fruit juice would then be another layer of complexity, which has two parts: the chemical effect of the acid and the optical effect of the juice. You could start off by shaking the samples of water versus water/ethanol together with measured doses of citric acid. This would separate out any optical color effect of the juice itself. Then you could shake out samples with water versus water/ethanol and citrus juice together with citrus juice and measured amounts of food coloring. This would let you see whether the citrus tended to produce a different color effect just optically when combined with water versus water/ethanol.

These things would give a pretty good idea of the relevant variables, I'd think.

As for why a basil drink would turn out darker green when shaken in all metal versus metal and glass... that's an interesting question. First, I guess I'd ask what the metal is made of. Are they using silver-plated shakers, for example? If they're using (presumably) nonreactive stainless steel, are they able to shake harder (or less hard) with the metal shakers? Is the headspace the same? Is the amount of ice used the same? Is it possible that the stainless shakers don't rinse quite the same as the glass ones? I can't think of any chemical reason, if everything else is equal, why there would be a stronger browning reaction using all metal shakers.

[Panaderia Canadiense]

SLKinsey - the oxidation reaction in basil is strongly reduced when it's torn as opposed to cut with a knife, so the metal shakers may be having more of an effect than you think.... It would be worthwhile to perform the shaking with water vs water-methanol in both the glass and metal shakers to see if it's significant or not.

[slkinsey]

. . . I think the solubility of chlorophyll in alcohol explains the visual difference between the two drinks.

Chlorophyll is highly nonpolar, and as such it is soluble in nonpolar solvents like ethanol. Laboratories take advantage of the solubility of chlorophyll in solvents like ethanol, hexane, acetone, etc., when they have to perform chlorophyll extractions.

Except, if the color difference were the result of more chlorophyll being extracted into solution by the ethanol, wouldn't one expect the color of the drinks with ethanol to be darker? The opposite is what was observed.

SLKinsey - the oxidation reaction in basil is strongly reduced when it's torn as opposed to cut with a knife, so the metal shakers may be having more of an effect than you think.... It would be worthwhile to perform the shaking with water vs water-methanol in both the glass and metal shakers to see if it's significant or not.

I don't believe this is particularly true. Try comparing basil leaves that have been cut with a really sharp knife to basil leaves that have been torn into similarly-sized pieces and see how great a difference you observe. Not as much as you might think. As a generality, people tend to cut basil into many more pieces than they might tear basil into, which I believe explains much of this purported effect. I think the belief that "torn basil doesn't darken" is an often-repeated but rarely tested "kitchen truism" similar to the belief that seasoned cast iron is nonstick.

That said, to the extent that tearing does have a mitigating effect on browning in basil leaves versus cutting, this is believed to be because tearing has a tendency to separate the leaves along the boundaries of cell walls whereas cutting has a tendency to cut and crush cells, thereby exposing a greater amount of cell contents to oxidation and enzymatic browning. Notably, it is not because the knife is made of stainless steel (I imagine knives made of reactive carbon steel may have an effect). You could cut basil leaves with a ceramic knife and observe the same effect as you would with a similarly sharp stainless steel knife.

Anyway, yes... doing straight-up comparisons of the results in glass-and-metal versus all metal shaker sets to see the difference would be step #1. But this has apparently already been done by Joerg Meyer and there is a difference. So now the question is what caused the difference. It could be that there is some fundamental difference between glass and metal that causes the observed difference, which I view as unlikely, or it could be that some associated variable is what causes the observed difference.

[Panaderia Canadiense]

Well, seasoned cast iron definitely isn't nonstick, it's just less sticky than unseasoned pans. However, I have tried the chop with wicked-sharp steel vs. chop with the plastic lettuce knife vs the wicked sharp ceramic knife vs tear into very small bits thing with basil, and in my kitchen at least the torn stuff darkens about 10 minutes more slowly and about 15% less than the cut does.

[KD1191]

. . . I think the solubility of chlorophyll in alcohol explains the visual difference between the two drinks.

Chlorophyll is highly nonpolar, and as such it is soluble in nonpolar solvents like ethanol. Laboratories take advantage of the solubility of chlorophyll in solvents like ethanol, hexane, acetone, etc., when they have to perform chlorophyll extractions.

Except, if the color difference were the result of more chlorophyll being extracted into solution by the ethanol, wouldn't one expect the color of the drinks with ethanol to be darker? The opposite is what was observed.

There's some sciencey language in the first of the two links I posted above that talks about the different ways in which chlorophyll absorbs light when it's surrounded by other chlorophyll vs. when it's dissolved in ethanol...

[haresfur]

I think Sam is on the right track. Controlled experiments with simplified combinations of ingredients (as you started with your tests) would be enlightening. Possible causes I can think of are:

- differential extraction of the clorophyl in the basil or some compounds in the lemon juice due to the less polar ethanol than the more polar water. Try leaving out one or the other (using citric acid as Sam suggested). But if the alcohol increases the extraction of the clorophyl, wouldn't the vodka drink be darker, not the other way around? Oh, Sam already said that.

- pH effects since the alcohol could muck with the hydrogen ion activity (maybe the acid, H+, is bound to the -OH of the ethanol). Seems unlikely to me because that is a fair bit of lemon juice. You could try mixing without the lemon and then adding it or citric acid by drops and look for a colour change. Can you make the colour match by adding more acid to the alcoholic one?

- oxidation effects. Try letting the drinks sit for a couple of hours and see if they look the same. Maybe one is oxidizing more rapidly. This could also be affected by the metal shaker. Even if steel is very inert it could catalzye the reactions.

- spectral effects. The alcohol might just shift the way the coloured compounds absorb & re-emit light. Does the effect look the same in day light and other light sources? I suppose borrowing a spectrophotometer is out of the question.

I think the first thing would be to establish for sure if it is the basil or the lemon that is causing the effect. Either that or say stuff it and use Goslings rum

- Dr haresfur, PhD

Edited July 26, 2011 by haresfur (log)

[KD1191]

- spectral effects. The alcohol might just shift the way the coloured compounds absorb & re-emit light. Does the effect look the same in day light and other light sources? I suppose borrowing a spectrophotometer is out of the question.

This is my bet. As I linked above, dissolving chlorophyll in ethanol causes it glow under UV light (while it doesn't when contained in plant tissue). Absent further data, it seems to me the shortest leap to suppose that it might change how it interacts with the visible light spectrum as well.

[evo-lution]

Basil is especially affected by enzymatic browning and oxidation reactions. This is why fresh-cut basil will almost immediately darken along the cut line whereas this effect is not as pronounced with many other herbs. I suspect it's also especially noticeable in basil because the initial color tends to be quite light compared to, e.g., parsley or mint.

The repeated rounds of drinks I made on Saturday contained basil, all my 'tests' at home (as written on the blog) have been with mint, and each person that's tried it has reported back the same thing with either basil and/or mint. I'm also interested to know if the the ABV would have an affect; would the drink be lighter with a stronger base spirit (more ethanol, less water) compared to a lower ABV gin.

The interest to me would be the ability to affect/control the colour using, let's say, red basil dependent on the shakers I used and the strength of my base spirits.

Edited July 27, 2011 by evo-lution (log)

[slkinsey]

So that's another thing to try: Compare samples at a wide variety of ABVs and see what the differences are.