Cocktail science

[Alchemist]

Can any one explain the scientific reason why ice forms on the outside if a swizzled drink?

What are the molucules doing differently in a very cold, viscous Manhattan, and a very cold. effervesent shaken Margarita. I understand that there is oxogen introduced into a shaken cocktail. But is that the only difference?

Any other questions and answers would be welcomed.

[slkinsey]

Can any one explain the scientific reason why ice forms on the outside if a swizzled drink?

I can answer many or most of these (one advantage of having chemists as parents) but I'll start with this one:

When you mix a cold solid (in this case, ice) with a warm liquid (in this case, water and alcohol), heat flows from the warmer liquid into the colder solid until thermal equilibrium is reached. Thermal equilibrium in this example would be a relatively stable temperature for the solid and the liquid to coexist at which all the solid hasn't melted and all the liquid hasn't frozen.

In the case of pure water ice mixed with pure liquid water, the temperature of thermal equilibrium would be the freezing/melting point of water -- zero degrees C. Mix together some distilled water ice and distilled water, wait a minute and measure the temperature. Zero degrees C.

What happens when we introduce alcohol into the mix? Well, alcohol has a lower freezing point than water. As a result, when we mix water and alcohol with ice, the temperature of thermal equilibrium is lower than the freezing point of water. Mix together some distilled water ice and some 100 proof alcohol, wait a minute and measure the temperature. Lower than zero degrees C.

When the liquid inside the glass is lower than the freezing point of water, condensation that forms on the outside of the glass is likely to freeze. This effect is magnified in drinks like a swizzle or a julep because the ice is crushed, resulting in increased surface area for transfer of thermal energy, and because the liquid addition is likely to be fairly high proof, lowering the temperature of thermal equilibrium. The effect is further magnified when a julep is served in a traditional silver cup, because silver has excellent thermal conductivity.

[eje]

Does the added energy from shaking alter the system?

Stirring seems to be a much more gentle way to reach thermal equilibrium.

Be interesting to do some time/temperature tests with the two methods.

[slkinsey]

In tests that have been done by friends, stirring with cracked ice seemed to produce the coldest cocktail.

IMO, as it relates to chilling cocktails, I believe that the most important variables are the temperature of the ice, the size and configuration of the ice, and the composition and temperature of the shaker. This last one may be a little confusing to some, but there is a reason that a lot of the better cocktail bars in NYC are using metal-on-metal Boston shakers instead of the more common metal-and-pint-glass ones. A glass mixing cup has a much higher thermal mass than a metal mixing cup of equivalent size. So, assuming that the shaker and mixing cup are at room temperature, the metal mixing cup will transfer less thermal energy into the drink than the glass mixing cup. If the mixing cup is frozen, however, the advantage is reversed and the drink should turn out colder from the metal-and-glass shaker. Pegu Club uses metal-on-metal for shaken drinks and frozen glass for stirred drinks to good effect. I remember when they went over to using frozen glass for the stirred drinks, shortly after the opening parties, and there was a very noticeable drop in the temperature of the stirred drinks.

[jsolomon]

Don't forget, there are a lot of bars out there that store their glasses in a freezer, so even sans-drink, they would form ice crystals due to the temperature of the glass.

It really pisses me off when I have to strain the slush from my drink through my teeth. I let my bartender know after the first time that I expect a room-temperature glass.

[jsolomon]

Does the added energy from shaking alter the system?

Stirring seems to be a much more gentle way to reach thermal equilibrium.

Be interesting to do some time/temperature tests with the two methods.

←

It's not at thermal equilibrium til the ice is mostly melted. Ice doesn't have a high thermal conductivity.

[slkinsey]

Does the added energy from shaking alter the system?

Stirring seems to be a much more gentle way to reach thermal equilibrium.

Be interesting to do some time/temperature tests with the two methods.

It's not at thermal equilibrium til the ice is mostly melted. Ice doesn't have a high thermal conductivity.

Thermal equilibrium is also not an applicable concept when one is speaking of shaking or stirring a cocktail. You don't reach thermal equilibrium when you are shaking a cocktail, you shake fast and strain the drink. When you are making a cocktail where a lot of crushed ice is sitting in the glass with the liquid, like a swizzle or a julep, is when the concept of thermal equilibrium applies.

It really pisses me off when I have to strain the slush from my drink through my teeth.  I let my bartender know after the first time that I expect a room-temperature glass.

Hmmm. I definitely prefer a chilled glass. It makes a notable difference in the temperature of the drink, especially over time. If you're finding a lot of ice shards in the bartender's drinks, that's likely due to bad ice or overzealous shaking or a combination of the two.

[Alchemist]

Thanks Sam. That all makes sense.

Now when you stir a Manhattan in thick frozen glass, with big, cold ice the texture becomes (this is a technical term) gooey, I think Dale uses silken. Are the molecules slow and staggering about like an drunken frat-boy because of the temperture? Or is it something else?

[jsolomon]

Hmmm.  I definitely prefer a chilled glass.  It makes a notable difference in the temperature of the drink, especially over time.  If you're finding a lot of ice shards in the bartender's drinks, that's likely due to bad ice or overzealous shaking or a combination of the two.

←

No, this isn't ice shards, it's frozen, slushy drink... like a beer you've left just a bit too long in the freezer.

[slkinsey]

Now when you stir a Manhattan in thick frozen glass, with big, cold ice the texture becomes (this is a technical term) gooey, I think Dale uses silken.  Are the molecules slow and staggering about like an drunken frat-boy because of the temperture?  Or is it something else?

All liquids have increased viscosity when the temperature is lowered. Viscosity may be simply described as a "fluid's internal resistance to flow" (more detailed info avaible here from Wikipedia). Think about how pourable simple syrup is when it is heated up, and how thick and slow it is when you take it out of the refrigerator. The difference is viscosity as a function of temperature.

The same thing is happening to that Manhattan when you stir it with ice. Liquors like bourbon and rye already tend to have more viscosity than gin or vodka, and when you chill the liquid the temperature-dependent increase in viscosity is magnified. This is why, although a stirred Martini is also silky, it's not quite as silky as a stirred Manhattan. The dilution of the drink by water is also important, because water has higher viscosity than alcohol at low temperature (alcohol remains fairly low viscosity at low temperature). This is why a drink that is chilled by stirring with ice is more viscous than one made with pre-chilled undiluted spirits.

The reason a stirred drink is silky and viscous compared to a shaken drink (assuming it is the same drink) is that the act of shaking incorporates air into the liquid, which decreases the sensation of viscosity and contributes a certain "lively" impression. But, as we all like to say about shaken drinks, you have to drink them quickly while they're still laughing at you (thanks to Harry Craddock for that bon mot). This is because the aeration provided by shaking will dissipate if you wait too long. If you make a stirred drink and a shaken drink and let them sit on the bar for a minute before sampling, you'll find that they're more or less the same by that time. Yet another argument for the three ounce drink, IMO.

This is a little more complicated than the idea that the molecules are slower when the drink is chilled because, as explained above, a Manhattan made by stirring with ice will be more viscous than a Manhattan made with pre-chilled spirits by virtue of the added water content, even though the latter might in fact be colder and therefore have "slower moving molecules."

[thebartrainer]

The reason a stirred drink is silky and viscous compared to a shaken drink (assuming it is the same drink) is that the act of shaking incorporates air into the liquid, which decreases the sensation of viscosity and contributes a certain "lively" impression.  But, as we all like to say about shaken drinks, you have to drink them quickly while they're still laughing at you (thanks to Harry Craddock for that bon mot).

Doesn't the presence of the bubbles contribute to a sensation of viscosity, due to the interuption of the flow of the liquid, rather than detract from it?

When one shakes a drink, especially one containing fruit juice, the texture is noticably more mouth coating than the same drink built or stirred. Indeed we have recently listed a creamless pina colada on the very principle that a creamy texture can be achieved by shaking both chunks of fresh pinepple and commercially produced pineapple juice.

On another point... Could the impression of relative smoothness of a martini vs a manhattan also be attributed to the human perception of the flavours in the spirit. The vanilla of the whisky being inherantly linked with a perception of dairy smoothness and the floral and herbal qualities of Gin being more readily associated with freshness.

Just a thought,

Cheers

Ian

[slkinsey]

Doesn't the presence of the bubbles contribute to a sensation of viscosity, due to the interuption of the flow of the liquid, rather than detract from it?

No... I would say that, everything else being equal, bubbles detract from perceived and actual viscosity (the two being related, but not exactly the same thing). Try this: Open a nice heavy ale and remove the carbonation to the greatest extent possible. You want a flat beer. Then open a fresh bottle of the same beer. Sample them side by side for mouthfeel. The flat beer is more viscous. Or, again... try making two identical Manhattans, one stirred and the other shaken. Sample them immediately. The stirred drink is more viscous. All other things being equal, a liquid full of bubbles should pour faster than the same liquid without bubbles. This means it is less viscous.

When one shakes a drink,  especially one containing fruit juice, the texture is noticably more mouth coating than the same drink built or stirred.  Indeed we have recently listed a creamless pina colada on the very principle that a creamy texture can be achieved by shaking both chunks of fresh pinepple and commercially produced pineapple juice.

There are many things that shaking can do to affect the texture of a drink, depending on the ingredients. Emulsification is certainly one, which is why shaking a drink with egg white or cream will often increase the viscosity. That said, I am not sure that I would always equate "creamy" with perceived viscosity.

I also don't have a hard time believing that shaking drinks with fruit juice improves the mouthfeel. Indeed, this is standard practice. But, again... I wouldn't say that too many drinks with fruit juice have a mouthfeel I would describe as "viscous." Experimental results have shown that acidity acts to suppress perceived viscosity, so this only makes sense. In general, fruit juice drinks go for "lively" rather than "silky."

On another point... Could the impression of relative smoothness of a martini vs a manhattan also be attributed to the human perception of the flavours in the spirit.

Yes, that could have something to do with it in terms of perceived viscosity. But I also believe whiskey is higher than gin in terms of real viscosity to begin with.

Mouthfeel is a complex perception, so I believe it is important not to mix descriptors too much. Viscosity, both real and perceived, is one discrete element of mouthfeel. Viscosity can be related to some of the other qualities you have used here (creamyness, smoothness, mouth-coating, etc.) but I think it's important to understand that it is not the same thing. A drink can be more smooth, creamy, mouth-coating, etc. and yet less viscous compared to another drink (for example, a Ramos Fizz is more creamy and mouth-coating compared to a stirred Manhattan, but less viscous)

[jsolomon]

Doesn't the presence of the bubbles contribute to a sensation of viscosity, due to the interuption of the flow of the liquid, rather than detract from it?

No... I would say that, everything else being equal, bubbles detract from perceived and actual viscosity (the two being related, but not exactly the same thing). Try this: Open a nice heavy ale and remove the carbonation to the greatest extent possible. You want a flat beer. Then open a fresh bottle of the same beer. Sample them side by side for mouthfeel. The flat beer is more viscous. Or, again... try making two identical Manhattans, one stirred and the other shaken. Sample them immediately. The stirred drink is more viscous. All other things being equal, a liquid full of bubbles should pour faster than the same liquid without bubbles. This means it is less viscous.

←

Sam, I'm not sure I agree with you on the mouthfeel of the beer, simply because of the difference of carbonation vs aeration in the mouthfeel category. I'm not going to argue any of the other points (though I disagree).

[slkinsey]

Well, you have a point about the beer thing. Aeration and carbonation aren't the same thing, and of course I wouldn't want nitpicking of the beer example the obscure the main point I am making about cocktails. I was just trying to find a common and easy to understand example, but the beer one perhaps isn't best. With respect to perceived viscosity, however, I think it's perfectly easy to tell that a stirred Manhattan is more viscous than a shaken one.

It also seems to be a widely accepted fact, for example, that lubricants that are aerated have reduced viscosity. So I'm not sure how one can disagree with the premise that an aerated drink is less viscous than a non-aerated drink (assuming a drink without eggs, cream, etc.). But if you have information to the contrary, or a better explanation of the OP's observed phenomenon, by all means offer it.

[jsolomon]

Tell you what, Sam, I'll try with a Boddington's.

Then, I might see if I can do some horse-trading with a P-chem lab T.A. to see if I can borrow a viscometer. I don't know about mouth-feel vs actual viscosity, but I could ask the other members of Fond Of Omaha Dining to see if they're in for some testing.

Edit to add: Jackpot

Edited January 24, 2006 by jsolomon (log)

[thebartrainer]

No... I would say that, everything else being equal, bubbles detract from perceived and actual viscosity (the two being related, but not exactly the same thing). 

I disagree with the ale example...

Hal McGee, "On food & Cooking" p.638 (revised edition)

Like emulsions, foams are a dispersion of one fluid in another.  In the case of foams the fluid is not a liquid but a gas, and the dispersed particles are not droplets, but bubbles.  Still the bubbles do the same thing that droplets do in a sauce: they get in the way of water molecules in the sauce and prevent them from flowing easily, and thus give the sauce as a whole a thic

ker body.

Or am I making an elementary error in mistaking viscosity with thickness?

Cheers

Ian

[jsolomon]

I think we're talking different talks. There is aeration where you have lots of bubbles, and there is aeration where you change the amount of gas dissolved in the liquid. I am not sure which we are talking about between Ian, Sam, and myself. We seem to be muddling the lines between all of them.

I can assure you that dissolving gas and creating bubbles do drastically different things to liquids.

[eje]

Quoting Splificator from the Classic Cocktails for Beginners topic:

...Removes mixing glass from freezer. Fills it with gin and vermouth in just proportion and dashes in orange bitters. Cracks ice, adds it to glass and stirs using circular motion of the wrist. Removes cocktail glass from freezer, strains liquid into in, cuts twist of lemon peel, twists it over drink...

Does anyone else keep mixing glasses in the freezer?

Tried this last night with the Blue Bird cocktail, and found I didn't get much (if any) dilution.

When I poured the gin into the 0 degrees Fahrenheit glass, it seemed to immediately chill to below freezing. Adding ice and stirring didn't do much. When I strained it out, I got basically the same amount of cold gin out of the glass as I had put in.

It does seem to be an efficient way of cooling a small amount of alcohol quickly. Very stiff cocktail, though.

On the other hand, I've chilled my larger glass martini pitcher, and had very good luck when making cocktails for 6. They seemed appropriately diluted and chilled.

[ThinkingBartender]

Does anyone else keep mixing glasses in the freezer?

Nope. Not in any bar I have worked in, nor in anyones house.

Quite frankly I have never had any trouble using a "room temperature" mixing glass. The only time in my experience that cocktails start being made badly, due to heat, is when the AC packs in and the ambient temperature of the bar melts everything.

It just seems like another layer of pomposity, and unfortunately some people are attracted to such things.

Cheers!

George

[Splificator]

I like my Martinis very cold. If that's pompous, then I'm pompous. But I still like a cold, cold Martini.

Edited after much visualizing of fluffy kittens.

Edited January 29, 2007 by Splificator (log)

[johnder]

I keep 2-3 mixing glasses and a few cocktail glasses in my freeze. I don't think I am pompous -- I just have the room to store it in my freezer.

When I make a stirred drink I will add the ice and liquor(s) into the glass and stir for 30 seconds, then take my time getting my garnish and glass out of the freezer. During this time I let the drink sit, then right before serving give it another couple of stirs.

John

[ThinkingBartender]

I like my Martinis very cold. If that's pompous, then I'm pompous. But I still like a cold, cold Martini.

←

So rather it is about the cold gin, than the actual cocktail (read: dilution).

Also, I am not finding any literature that says that the Martini was the first gin and vermouth cocktail, and that all other gin and vermouth cocktails are but variants of this primigenes cocktail.

Cheers!

George

"I never have more than one drink before dinner. I do like that one to be large and very strong and very cold, and very well-made. I hate small portions of anything, particularly when they taste bad"

[Splificator]

So rather it is about the cold gin, than the actual cocktail (read: dilution).

No, it is not about the cold gin rather than the actual cocktail. Be nice, George--who knows, you might enjoy it.

I've only been storing a mixing glass in the freezer for a few weeks, mostly because I found myself with an extra one and, like Johnder, I had some space there (and nowhere else in the kitchen, especially once I broke down and bought the huge Ra Chand juicer). But I haven't noticed a conspicuous lack of dilution in the drinks made with the cold glass, and I've made a good sampling of sours, vermouth drinks and old-style Cocktails with it and they seem to have come out properly balanced, to my palate anyway.

Then again, when it comes to drinks I do rather like them the way Bertie Wooster does: "stiffish, Jeeves, stiffish."

Also, I am not finding any literature that says that the Martini was the first gin and vermouth cocktail, and that all other gin and vermouth cocktails are but variants of this primigenes cocktail.

Then look harder.

[slkinsey]

Does anyone else keep mixing glasses in the freezer?

Nope. Not in any bar I have worked in, nor in anyones house.

You need to hang around in a better class of bar, or make better friends. :smile:'

They have been freezing the stirring glasses at Pegu Club since the day they opened.

There are many things that can effect dilution. The most important would be the temperature of the ice, the surface area to volume ratio of the ice (i.e., shape and size) the temperature of the spirits and the contact time with the ice. I would be happy to delve into this in (significantly) greater detail, but suffice it to say that it is possible to have a Martini that is both extremely cold and properly diluted.

The point of freezing the mixing glass is simply that a glass mixing glass has a fairly large thermal capacity in this context. When ice and liquid are introduced into a room temperature mixing glass, this creates a thermal disequilibrium. Thermal energy is therefore conducted from the mixing glass into the contents of the glass, resulting in a higher overall temperature. This is the reason shaking in all metal will produce a colder drink when the equipment is at room temperature -- the thermal capacity of as thin metal cup is much lower than the thermal capacity of a thick pint glass, and therefore less thermal energy is conducted into the liquid.

If the glass is cold, there is no disequilibrium, and no thermal energy is conducted into the liquid. In practice, what we would like to have is a disequilibrium that goes in the other direction, so that thermal energy is conducted from the liquid into the glass, resulting in a lower overall temperature. Given the temperature of the ice, this particular effect may be minimal. Mainly what we want is for the mixing glass to not donate thermal energy to the liquid and for the only meaningful exchange of thermal energy to take place between the ice and the liquid.

So... inside the mixing glass thermal energy is still being conducted from the liquid into the ice. This has two results. First, the temperature of the liquid is lowered as thermal energy is transferred to the ice. Second, the transfer of thermal energy from the liquid to the ice creates a phase change and there is some melting of ice into the liquid. The colder the ice, the colder the liquid can be. The greater the surface area to volume ratio of the ice, the more efficient and rapid the exchange of thermal energy can be. This can also result in a colder drink and, if the ratio is too large, excess dilution. Regardless, so long as the liquid is at room temperature, there will be sufficient exchange of thermal energy to create the proper amount of dilution. Want more dilution? Crack the ice into smaller pieces.

So, what do we learn from this? We learn that freezing the mixing glass is good, resulting in a colder drink that can still have sufficient dilution. We also learn that freezing the gin is bad, resulting in a cold drink but one that is unlikely to have sufficient dilution.

Edited January 29, 2007 by slkinsey (log)

[Splificator]

Thanks, Sam! And to think that all that brainpower could have been lost to mixology.

I always use cracked ice, BTW, FWIW and any other initials that apply.