Table of Freezing Points of Alcohol Solutions

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
R

rayg

Well-Known Member
Joined
Jan 31, 2010
Messages
379
Reaction score
20
Location
Boston
There was a question in another thread about freezing points
of beer, so I thought I'd post this info for reference. I wrote
a program to spit out the numbers.

%Alc Temp (F)
by
vol

3.0 30.2
3.1 30.2
3.2 30.1
3.3 30.0
3.4 30.0
3.5 29.9
3.6 29.8
3.7 29.8
3.8 29.7
3.9 29.7
4.0 29.6
4.1 29.5
4.2 29.5
4.3 29.4
4.4 29.3
4.5 29.3
4.6 29.2
4.7 29.2
4.8 29.1
4.9 29.0
5.0 29.0
5.1 28.9
5.2 28.8
5.3 28.8
5.4 28.7
5.5 28.7
5.6 28.6
5.7 28.5
5.8 28.5
5.9 28.4
6.0 28.3
6.1 28.3
6.2 28.2
6.3 28.1
6.4 28.1
6.5 28.0
6.6 27.9
6.7 27.9
6.8 27.8
6.9 27.8
7.0 27.7
7.1 27.6
7.2 27.6
7.3 27.5
7.4 27.4
7.5 27.4
7.6 27.3
7.7 27.2
7.8 27.2
7.9 27.1
8.0 27.0
8.1 27.0
8.2 26.9
8.3 26.8
8.4 26.8
8.5 26.7
8.6 26.6
8.7 26.6
8.8 26.5
8.9 26.4
9.0 26.4
9.1 26.3
9.2 26.2
9.3 26.2
9.4 26.1
9.5 26.0
9.6 26.0
9.7 25.9
9.8 25.8
9.9 25.7
10.0 25.7
10.1 25.6
10.2 25.5
10.3 25.5
10.4 25.4
10.5 25.3
10.6 25.3
10.7 25.2
10.8 25.1
10.9 25.1
11.0 25.0
11.1 24.9
11.2 24.8
11.3 24.8
11.4 24.7
11.5 24.6
11.6 24.6
11.7 24.5
11.8 24.4
11.9 24.4
12.0 24.3


Ray
 
Cool, you got anything simple like that for boiling points/solution for making N/a beer? There's debate about boiling being a viable means of doing it, and it looks like something I'll be needing to play with for healt reasons.

:mug:
 
Revvy said:
Cool, you got anything simple like that for boiling points/solution for making N/a beer? There's debate about boiling being a viable means of doing it, and it looks like something I'll be needing to play with for healt reasons.

:mug:
Click to expand...

Do you mean what the boiling point would be at reduced pressure?
If you boil beer at atmospheric pressure not only would ruin the
taste of the beer (that's why flash pasteurization is used when it
is used) but the boiling point of any typical beer isn't going to
be much different from 100C. What would happen is you would start
to collect vapor at say, 95C, then the temperature in the pot would
slowly rise until the ethanol was all gone, at which point the temp
would read 100C, or maybe a little higher because the dissolved solids
in the beer cause a boiling point elevation, unlike liquids where the
dissolved solids cause a freezing point depression. Go to Google
and type in "ethanol water phase diagram" in the exact phrase
box to see the behaviour of ethanol/water solutions.

Ray
 
I get somewhat different answers. An 8% ABW solution contains 80 g alcohol per kG for a molality of 80/46.07 = 1.7365 mol/kg so its freezing point depression should be 1.858*1.7365 = 3.23 °C ~ 5.81 °F and the freezing point should thus be 26.19 °F. Pure ethanol has a density of 0.789 g/cc (at 20 C) and so these 80 grams correspond to a volume of 80/0.789 = 101.363 cc. An 8% ABW solution has density 0.98473 (20°C) and thus 1 kg of it a volume 1015.5 cc and the ABV is thus 9.98 percent - very close to 10. In your table for 10% ABV the freezing point is given as 25.7 - about half a degree lower.

Now I note that if the true extract of a beer is, say 5°P and I model it as sucrose, with a molecular weight of 342.1 and I assume that the low pH of the beer inverts each sucrose molecule into a fructose and a glucose that will increase the molality of the solution by 2*50/342.1 = 0.292 for an extra depression of 1.858*0.292 = 0.5 C ~ 0.98 °F and the freezing point would be 25.21°F , about half a degree lower than the figure you post. Lots of assumptions go into that last calculation because 1)many beers don't have TE as high as 5 °P 2) the TE is made up of, presumably, higher molecular weight dextrines and oligosaccharides (which leads to smaller depression than fructose/glucose would) 3)sucrose might not be fully inverted 4) a kg of 5 °P solution would contains less water than a kg of pure water.

So the questions really are: 1)Are your numbers based on a pure water ethanol solution and if so where do we differ? 2)If they account for the other stuff dissolved in beer, what was the model (how much of what)?
 
ajdelange said:
I get somewhat different answers. An 8% ABW solution contains 80 g alcohol per kG for a molality of 80/46.07 = 1.7365 mol/kg so its freezing point depression should be 1.858*1.7365 = 3.23 °C ~ 5.81 °F and the freezing point should thus be 26.19 °F. Pure ethanol has a density of 0.789 g/cc (at 20 C) and so these 80 grams correspond to a volume of 80/0.789 = 101.363 cc. An 8% ABW solution has density 0.98473 (20°C) and thus 1 kg of it a volume 1015.5 cc and the ABV is thus 9.98 percent - very close to 10. In your table for 10% ABV the freezing point is given as 25.7 - about half a degree lower.
Click to expand...

First, there is an error in calculating the molality. Molality is moles
per kg of SOLVENT, not total volume of solution. 80 gm of ethanol
leaves 920 gm of water, so the correct molality is 1.7365/.92 = 1.8875.
I used the rounded value of 1.86 for the molal freezing point depression,
so that comes to 3.51C or 6.32F, which is more than the half degree
you report but it gives the same value in the table for 10% by volume.
I didn't check all the rest of your calculation, but I just
converted volume % to weight % by dividing volume% by 1.25.


The values in the table are for pure ethanol solutions, so they
are a maximum. There is no way for a homebrewer to know the
weight% of dissolved solids in any of his beers, since that requires
laboratory analysis. Even a guess would depend upon accurate
values for initial and final gravities. There is no way
to know exactly what the average molecular weight of those
solids is either (I suspect it's pretty high, higher than sucrose)
There is no way for a homebrewer to even know what the exact percent
alcohol he has. The use of the table is so you have some general
idea how low you can go, but there is no guarantee you won't
still freeze your beer, but in general the actual freezing point
should be lower (but not by much) than that in table, so you
should be able to go at least that low.

Ray
 
This thread got me thinking

rayg said:
There is no way for a homebrewer to know the
weight% of dissolved solids in any of his beers, since that requires
laboratory analysis.

The analysis is so simple that any motivated (probably very few) home brewer should be able to carry it out. Simply evaporate the beer down to about half or a bit less of its original volume then make back up to full volume with DI water and measure the SG. Convert this to Plato. Then the TE (True Extract) is TE = Plato*SG/SG_beer (SG_beer is the SG of the beer before evaporation). The only equipment required is a volumetric flask, a beaker and a hydrometer.

Then the other half:

rayg said:
There is no way
to know exactly what the average molecular weight of those
solids is either
Click to expand...

How did the earliest chemists measure molecular weight? Freezing point depression and/or boiling point elevation and that should work here too. If p= TE/100 the weight of the solids in V mL of beer is Ws= V*dens*p
and the weight of the water is Ww= V*dens*(1-p). dens is the density of the dealcoholized beer but it doesn't need to be calculated as it cancels out. The molality, call it M, is the number of moles of solids divided by the weight of the solvent (got that right this time): M = mA*p/(1-p) with mA being the average molecular weight. With the depression (or raise) dT = k*mA*p/(1-p) it's clear that mA = dT*(1-p)/(k*p) (with k being the appropriate constant depending on whether it's freezing point or boiling point being considered).

All this is not of very much practical use, of course, but I'm intrigued enough about the surviving sugar (and protein) that I may check it out on the next beer I analyze (in the fermenter now). So watch this space (if you're interested).
Click to expand...
 
ajdelange said:
This thread got me thinking

rayg said:
There is no way for a homebrewer to know the
weight% of dissolved solids in any of his beers, since that requires
laboratory analysis.

The analysis is so simple that any motivated (probably very few) home brewer should be able to carry it out. Simply evaporate the beer down to about half or a bit less of its original volume then make back up to full volume with DI water and measure the SG. Convert this to Plato. Then the TE (True Extract) is TE = Plato*SG/SG_beer (SG_beer is the SG of the beer before evaporation). The only equipment required is a volumetric flask, a beaker and a hydrometer.
Click to expand...

I don't see how this would be any more accurate for
homebrewers than simply estimating the true extract by
calculation. Most homebrewers don't have precision equipment.


ajdelange said:
How did the earliest chemists measure molecular weight? Freezing point depression and/or boiling point elevation and that should work here too. If p= TE/100 the weight of the solids in V mL of beer is Ws= V*dens*p
and the weight of the water is Ww= V*dens*(1-p). dens is the density of the dealcoholized beer but it doesn't need to be calculated as it cancels out. The molality, call it M, is the number of moles of solids divided by the weight of the solvent (got that right this time): M = mA*p/(1-p) with mA being the average molecular weight. With the depression (or raise) dT = k*mA*p/(1-p) it's clear that mA = dT*(1-p)/(k*p) (with k being the appropriate constant depending on whether it's freezing point or boiling point being considered).

All this is not of very much practical use, of course, but I'm intrigued enough about the surviving sugar (and protein) that I may check it out on the next beer I analyze (in the fermenter now). So watch this space (if you're interested).
Click to expand...

How are you going to measure the freezing point accurately?

I know that I have seen analyses of beer with a "percent
residual solids" reported in one of my brewing books.
I would then just assume that the average molecular weight
is that of maltose, and calculate a depression for
that, then do the calculation again assuming that
the molecular weight is higher, say that of maltoheptose,
and use those numbers as a range.

All I need to know is how low I can go before it freezes,
for lagering purposes. There is really no other use
for the freezing point table, I put it there mostly
just to satisfy anyone's curiosity. If I wanted to make
ice beer, for example, I'd have to watch it freeze because you
can't let it all freeze, you have to collect the
unfrozen portion after only some of it freezes.

Ray
Click to expand...
 
rayg said:
I don't see how this would be any more accurate for
homebrewers than simply estimating the true extract by
calculation. Most homebrewers don't have precision equipment.[?QUOTE]

Well, as I said it's more of academic interest than practical significance. But it doesn't take precision equipment (unless you consider the hydrometers sold in sets of three by B^3 "precision"). Obviously from the practical consideration standpoint the best way to see what temperature your beer freezes at is to put it in the freezer and measure the temp. when slush starts to form.


rayg said:
How are you going to measure the freezing point accurately?
Click to expand...
With a thermometer or RTD. Actually, now that I think of it ebulliometer's thermometers are narrow range so that boiling point elevation might be the way to go.


rayg said:
I know that I have seen analyses of beer with a "percent
residual solids" reported in one of my brewing books.
Click to expand...

That's true extract. Your books might have used the term "residual" because TE is usually measured on the residue left behind in the distillation flask when alcohol is measured. Brewers (not homebrewers) use TE and alcohol content to calculate effective original gravity (using the same equation that homebrewers use to calculate alcohol from OE and apparent extract.


rayg said:
I would then just assume that the average molecular weight
is that of maltose, and calculate a depression for
that, then do the calculation again assuming that
the molecular weight is higher, say that of maltoheptose,
and use those numbers as a range.
Click to expand...

I could do that but that wouldn't be much fun. Given I've got the distillation apparatus set up and the dealcoholized, reconstituted beer it would be a snap to take the portion of that left over from residue gravity determination, put it back in the distilling flask and measure the boiling point. And no, I don't expect that this has much relevance to the average homebrewer.

rayg said:
All I need to know is how low I can go before it freezes, for lagering purposes.
Click to expand...

Understood. But this is a Brewing Science forum so I have no qualms about discussing brewing science (the TE method comes right out of the ASBC's MOAs).
Click to expand...
 
I found an analysis of a stout in a book I have,
the alcohol content of which was 6.37% by w, which
is 7.96% by v.

The amount of other material dissolved is pretty small,
the most being protein 5.8g/L and glycerol 2 g/L. Other
things include:

K .468g/L
P .227 g/L
So4 .130 g/L
Cl .250 g/L
Pyruvate .128 g/L
citrate .187 g/L
Acetate .219 g/L
Polyphenols .199 g/L

There are lots of other things present in the microgram/L range
but I'm not including them here.

I used the exact molecular weights for everything
except protein (I just used 115, which is the weight
of proline, as an average) and polyphenols (I used
250, which is about the weight of a biphenyl with
4 hydroxy groups). The total number of moles comes
out to be tiny (0.107 moles/L) compared to the ethanol
(1.38 moles/L). So the biggest change is just in the
reduced mass of the solvent which goes from
0.936 kg to 0.927 kg. The number of moles increases from
1.384 to 1.491 moles, and the molality increases
from 1.384/.936 = 1.479 to 1.491/.927 = 1.608.
That in turn means the freezing point decreases
from 27.0F to 26.6F.

Ray
 
As I like things simple I calculated a polynomial which approximates the freezing point of pure ethanol in pure water solutions it is:

Fp = ((0.0075275*ABV + 0.054922)*ABV + 31.947 °F

You can copy this and paste it into Excel. It's good from 0 to 25% ABV. As you can see it calls the freezing point of water 31.947 - an error of 0.053 °F and at 25% ABV the error is -0.065 °F. It's better away from the ends of the range. rms error is 0.02°

It is assumed that ABV is specified at 20 °C and this temperature was used in conversion from ABV to ABW using the OIML polynomial. The water constant used to calculate the freezing point depression was 1.855.
 
rayg said:
I found an analysis of a stout in a book I have,
the alcohol content of which was 6.37% by w, which
is 7.96% by v.

The amount of other material dissolved is pretty small,
the most being protein 5.8g/L and glycerol 2 g/L. Other
things include:

K .468g/L
P .227 g/L
So4 .130 g/L
Cl .250 g/L
Pyruvate .128 g/L
citrate .187 g/L
Acetate .219 g/L
Polyphenols .199 g/L
Click to expand...

Those sum up to a TE of about 1°P. Noticeably missing from this list is sugar! HBS&Y have some limited data on sugar contents of beer. For a "Danish Lager" sugars contribute about 2.4 °P to the TE. For "German Lager" its 2.83 and even for "Diabetic Lager" it's 0.96.

My stouts (the only ones for which I have TE data) have TE's between 3.7 and 4. And these are dry stouts. They have the lowest TE's of anything I brew. Eighteen beers I've analyzed have an average TE of 5.2 with a standard deviation of 1. The stouts are the lowest, Kölschs next at 4.7 - 4.8, Pils and ales around 5, Fests 5-6 and Bocks 6 - 7.9. Remember that TE is everything in the beer except the alcohol (and any other volatile such as acetic acid) thus these figures would include the pyruvate and citrate to the extent that they behave like sucrose in terms of the way they effect density. But they are kind of in the noise relative to residual sugars. It's going to be alcohol, glycerol (makes great antifreeze), protein and sugars that are going to be responsible for most of the FP depression.
 
ajdelange said:
Those sum up to a TE of about 1°P. Noticeably missing from this list is sugar!
Click to expand...

I'm sure I just overlooked it in the table. I'll have to check again
tonight when I get home.

Don't understand why you are using TE, it can't be converted to moles
so you can't use it for calculating a freezing point depression.

Ray
 
In order to compute molality I have to know both how much stuff is in solution and what it's average molecular weight is. TE answers the first question. The average molecular weight must be determined by experiment i.e. by measuring freezing point depression or boiling point elevation. As FP depression is 3.7 times more sensitive than boiling point elevation, I think the former is the way to go. The intent, of course, would be to measure average mol. wt on a couple of beers in the hopes that the grouping would be tight enough that one "average average" value could be used for general purpose estimation or that one could at least say the AMW is one thing for Pils and another for Bock, for example.

Another approach to getting AMW is to look at the spectrum of sugars and other things reported in beer in brewing texts and compute an average that way.
 
ajdelange said:
In order to compute molality I have to know both how much stuff is in solution and what it's average molecular weight is. TE answers the first question. The average molecular weight must be determined by experiment i.e. by measuring freezing point depression or boiling point elevation. As FP depression is 3.7 times more sensitive than boiling point elevation, I think the former is the way to go. The intent, of course, would be to measure average mol. wt on a couple of beers in the hopes that the grouping would be tight enough that one "average average" value could be used for general purpose estimation or that one could at least say the AMW is one thing for Pils and another for Bock, for example.

Another approach to getting AMW is to look at the spectrum of sugars and other things reported in beer in brewing texts and compute an average that way.
Click to expand...

It seems to me that calculating the mw from reported values of
dissolved substances is the way to go. If you are going to
go to the trouble of actually measuring the freezing point of the beer,
you don't need a table of freezing points. But I don't see how you
can accurately measure the freezing point. In a lab you use a capillary
melting point apparatus, but how are you going to get an accurate
melting point of something the freezes so low at home? Are you going
to put a container of beer in a cold room and watch it melt (and
do it several times to get an average)? You have be there right at
the instant it starts to melt, and you need a really small sample
to eliminate heat imbalance across the sample (that's why it's
done in a capillary tube and viewed with a magnifying glass).

Ray
 
I have some experience in using an ebulliometer to measure BP depression (for alcohol estimation) but none with FP depression. What I was thinking of doing is put the sample in a small tube and put that in the freezer and record the temperature as a function of time. The temperature should fall as the solution cools and then stay more or less constant while freezing is taking place until the tube is frozen solid at which time it should start to decline again. I'd probably grab the middle point of that flatish region and call that the freezing point. Yes, I understand the problems you mention but I'm not shooting for extreme accuracy here. Obviously I'd do this with calibrated strength solutions to see how close I could get.

In trying to estimate the molecular weight of sucrose by by boiling point elevation this afternoon I observed a 0.2° rise when I should have seen 0.1° so I'd estimate sucrose at 171 Daltons. It's clear that a thermometer readable to 0.1°C doesn't have enough precision.
 
While brushing the dog it occurred to me that I have a device in which a capillary tube is affixed to a Peltier attemperator with a PRT that reads to 0.001°C and a camera for monitoring what goes on it the tube. Perfect, right? Well, alas no. For one thing, it only goes down to 0° C. For another it's designed to measure density - not freezing point and I'd have to commit ceremonial sepuku if I burst that tube by letting something freeze in it.
 
The real extract of that beer was 6.6. If subtract the
~1 degree of extract that the other substances
contribute, that leaves 5.6. That's going to bring the
mass of water remaining to ~0.88 kg. But the sugars must
be at least as big as maltotriose (~504.5 daltons), which
means a further reduction in the freezing point of about
0.42F. But the average mol wt is probably closer to
maltoheptose (~1153 daltons), so that would mean
a reduction of only about 0.18 degree. The low mol
wt of the alchohol means it has a much larger effect.

Ray
 
I'm sure other factors contribute to this, but I was reading an article recently about the importer for Corona having problems with beer freezing. They said it's only an issue at 17 degrees or less. Maybe this has to do with it being in a train boxcar and it would take a while for that amount of liquid to freeze?
 
ajdelange said:
I have some experience in using an ebulliometer to measure BP depression (for alcohol estimation) but none with FP depression. What I was thinking of doing is put the sample in a small tube and put that in the freezer and record the temperature as a function of time. The temperature should fall as the solution cools and then stay more or less constant while freezing is taking place until the tube is frozen solid at which time it should start to decline again. I'd probably grab the middle point of that flatish region and call that the freezing point. Yes, I understand the problems you mention but I'm not shooting for extreme accuracy here. Obviously I'd do this with calibrated strength solutions to see how close I could get.

In trying to estimate the molecular weight of sucrose by by boiling point elevation this afternoon I observed a 0.2° rise when I should have seen 0.1° so I'd estimate sucrose at 171 Daltons. It's clear that a thermometer readable to 0.1°C doesn't have enough precision.
Click to expand...


I still don't understand why you would want do this yourself. You can just look at a table online of water/ethanol freezing point equilibria and get an estimate. If you really want to get accurate use a CRC handbook. Those calculations are done using lab calibrated equipment. It won't account for the other molecules (proteins, salts, acids, etc..) in solution but you can get estimate. From the data presented here you can see that even accounting for it you have about a 1-2 degree differential.


Source: Chem Ph.D
 
BrewSwag 300

Latest posts

Back
Top