why ice melt faster in regular soda than in diet soda

I found two webpages you may find interesting, it's not written up as
an answer though because I couldn't find one heh.

http://www.madsci.org/posts/archives/jan2001/980622126.Ph.r.html
http://www.nsta.org/recommends/product.asp?id=13073

skermit-ga

A great question, connie1428, but I think you have it backwards. Ice
melts faster in diet soda than regular soda, not the other way around
as you have phrased it. Skermit's first link clarifies that is is the
other way around. Unfortunately, neither of skermit's links answers
the question, and that is too bad because it is a fantastic question.

The first link rambles on about standardizing experiments and blah,
blah, blah. Well, none of that matters because ice melts so much
faster in diet soda than regular soda that you would catch the
difference even in the sloppiest experiments. It is a big difference;
too big to miss.

The first link also misses the reason for the phenomenon completely!
You can probably figure it out from the following list of ingredients
in Coke (tm) and Diet Coke (tm). Here they are:

Coca Cola (R) Classic: High Fructose Corn Syrup and/or Sucrose, Water,
Caramel Color, Phosphoric Acid, Natural Flavors, Caffeine.
Diet Coke (R): Water,  Caramel Color, Phosphoric Acid, Sodium
Saccharin, Potassium Benzoate, Natural Flavors, Citric Acid, Caffeine,
Potassium Citrate, Aspartame, Dimethylpolysiloxane.

These are taken from http://www.rense.com/general50/class.htm

Did you catch it? It helps if you know that the FDA mandates that
ingredients must be listed in DECREASING order by content. Let me show
you the first part of each again, with corresponding ingredients one
above the other:

Coca Cola (R) Classic: High Fructose Corn Syrup and/or Sucrose, Water,
Caramel Color...
Diet Coke (R): ------------- NOTHING! ------------------------- Water,
 Caramel Color...

Why does Diet Coke (R) have nothing (actually, water) instead of the
sugary stuff? It's because Saccharin and Aspartame are VERY much
sweeter tasting than sugar, so you need a lot less of them. The folks
at Coca Cola hope that they taste about the same sweetness (and they
more or less do) but from the point of view of chemistry, all we care
about is the number of molecules. That is what makes the difference!

Here's the surprising fact: An ice cube dropped into a solution of
sugar or salt will melt MUCH more slowly than an ice cube of the same
size dropped into plain water. That sounds very wierd, because we are
used to melting ice with salt, and we know that salt lowers the
melting point of ice, and we put salt into ice to make it very cold
when we make ice cream. The salt must be melting the ice fast to keep
the ice cream so cold, right? Well, actually no.

Each of these situations, first of all, is actually very different
from dropping an ice cube into salty or sugary water. In the first
situation, salt is added to ice (say, on the roads in winter) and a
little bit of the salt mixes with the ice. That leaves us with
something that is liquid at normal winter temperatures: a little bit
of salt water. In the frigid north of Canada, they don't use salt on
the roads because it won't melt ice at temperatures below about 10
degrees celsius (about 15 degrees F). They use sand instead. (That is
why cars rust more in Toronto and Buffalo than either Ft. Lauderdale
or Timmins.)

Putting salt into a mixture of water and ice cubes when we're making
ice cream is closer to what we're talking about, but it is still not
right on. In that case , we're adding salt after the water and ice are
mixed. We observe a temperature drop for exactly the same reason that
our ice cubes take longer to melt in Coke Classic, but it is harder to
think about and get the right answer. In the ice cream-making
situation, we add the salt, the temperature drops, and we make our ice
cream. We don't really notice that the ice takes longer to melt
(though it does). If it lasts long enough for us to freeze our ice
cream, that's good enough.

The easiest way to think about our problem, then, is to imagine two
glasses of water side by side at room temperature. Once contains tap
water, and the other contains tap water with a tablespoon of sugar
added. Drop identical ice cubes into the water at the same time, and
watch them melt. The ice cube in the sugar water will be there long
after the ice cube in the regular water is gone. In this case, the
sugar water is standing in for the Coke Classic, and the tap water is
replacing the Diet Coke. You can do this because, to a first
approximation, the two are the same except for the absence of large
amounts of sugar in the Diet Coke. If you use an equivalent amount of
salt instead of sugar, the ice lasts even longer. Up to about twice as
long, with 1/6 the amount of salt by weight added to the water. (I'll
explain why later.)

So why does the ice last longer in the sugar or salt water? There are
many ways to think about it. One simple explanation is that the sugar
(or salt) lowers the temperature of the water. In a state-change,
where a substance is going from a solid to a liquid, there will always
be some molecules hopping out of the solid to the liquid (melting),
and some molecules hopping back into the solid FROM the liquid
(freezing), and the difference between them determines the rate of
melting (or freezing). If you lower the temperature of the water, you
increase the rate of freezing compared to melting, and the ice takes
longer to melt.

That is the correct explanation, but I find it deeply unsatisfying
because it doesn't really explain what is going on. Here's my
favourite:

A state change (melting, boiling, freezing) isn't really a chemical
reaction, but it has much in common with one. State change reactions,
like chemical reactions, either release energy (in which case we say
they are exothermic) or they consume energy (in which case we say they
are endothermic). They proceed at a rate that depends upon the energy
required for the state change, and upon the available energy. In
chemical reaction terms, this is called the reaction rate. Chemical
bonds are broken, and others are formed, although not quite the same
kind of bonds as in chemical reaction. These bonds are weaker "ionic"
bonds, or bonds formed by the electrostatic attraction between
molecules rather than the strong bonds of chemistry which are mediated
by the outer (called valence) electrons. And finally, like in normal
chemical reactions, other substances may affect the reaction rate
without being changed themselves by the reaction. In chemical
reactions, we call these catalysts. In our ice cube melting
"reaction", this is the very important part played by sugar or salt.

So here is what happens in our state-change "reaction" (when our ice
cube melts). Let's start with the situation of the ice cube melting in
pure water.

In the water as well as the ice, the water molecules, consisting of
one oxygen and two hydrogens are attracted to one another by
electrostatic forces. The oxygens have a bit of a negative charge, and
the hydrogens have a bit of a positive charge. Naturally, since
opposites attract, the molecules line up with their oxygen next to one
of their neighbor's hydrogens, and their hydrogens next to each of two
neighbor's oxygens. This is exactly like arranging the guests
boy-girl-boy-girl at a birthday party. The attachments between the
molecules are not very strong in the water, so they can break easily,
and the water can flow. In the ice, they are a bit stronger, but still
not terribly strong compared to normal chemical bonds. The bonds are
strong enough in the ice that it is solid and can't flow like water
but they are still weaker than the bonds between the hydrogen and the
oxygen. That's why it is a lot easier to melt ice or make water flow
or make steam than it is to separate water into hydrogen and oxygen
gases. The first things all involve breaking weak "ionic" or molecular
bonds, and the latter involves breaking the chemical bonds.

Now to make ice into water, you have to put in some energy in the form
of heat to break those bonds, but how much energy you put in depends
upon the state of the water as well. A molecule has to leave its pals
in the ice, and connect up with three other water molecules to form
the regular structure I described before, H-O-H-O-H, except in three
dimensions. How much energy you put in determines how fast the ice
melts. The energy comes from the water, which is warmer than the ice.
The ice melts, the water cools, and pretty soon the ice cube is gone.

Now let's take the situation with the salt already in the water when
you drop in the ice cube. The starting temperature is the same. Just
adding salt doesn't lower the temperature of the water by itself (or
not enough for us to measure, anyway).

The structure of the water is very different with the salt present.
Sodium chloride (NaCl), made up of a positively charged sodium atom
and a negatively charged chlorine atom, gets pulled apart by the
water. Instead of lining up H-O-H-O-H like before, the water lines up
something more like H-O-Na-Cl-H-O etcetera, again in three dimensions.
This means that, in order for a molecule to come out of the ice and
line up with three water molecules, the water molecules now have to be
pulled away not only from each other, but from all the sodiums and
chlorines as well. In addition, the sodiums and chlorines pull a
little more strongly than the water molecules pulled on each other.
This means that the salt acts as a "catalyst" to change the reaction
energy of the state change (in this case negatively, so more energy is
required). Since more energy is required, and the water starts out at
the same temperature, the "reaction" (melting) is slower and the ice
cube takes longer to melt.

It takes a long time to say it that way, but it is easier to
understand than waving your hands and saying that the salt lowers the
temperature of the water. (Between us, it doesn't really lower the
temperature of the water, anyway. It increases the energy needed for
the state change, which transfers more energy out of the water and
into the melting process, which in turn lowers the temperature of the
water. You can assure yourself of this by dumping some Epsom Salts in
a nice, warm bath before a long soak. The water doesn't get cold. If
it did, it would defeat the whole purpose of the bath. But if any high
school teachers tell you otherwise, just play along. It'll all get
sorted out when you get to college.)

What about sugar, you ask? Well the sugar works just like the salt,
only weaker. It doesn't have the strong positive and negative of the
sodium and chlorine, but it has hydrogens (weakly positive) and some
oxygen (weakly negative) surrounding a ring of (mostly) carbon (which
is electrically neutral), so it does the same thing, only not as
strongly. Also sugar doesn't split into two like salt does, so that
means it is half as good as an equivalent number of molecules of salt
right off the bat.

Remember I said that salt was six times better by weight at keeping
ice from melting than sugar? that is because the effect depends upon
the number of molecules (like everything in chemistry) rather than the
actual weight of the substance. Well, one molecule of salt weighs
about 58 (in molecular units, based on the weight of hydrogen being
one, approximately) and glucose (the main ingredient in Coca Cola once
the corn syrup stuff gets broken down to its components) weighs about
180. That is about a 3 to one ratio, meaning that in a sample of the
same weight, there are 3 times as many salt molecules as sugar
molecules, so the salt can do three times more work keeping ice from
melting. BUT the salt splits into two! Two times three is six, and
that is the number I gave before. (Actually the number six applies
more to the boiling point of water than the freezing point, but that's
another story. You get the general idea.)

If you have made it this far, you now know why you should order Coke
Classic rather than Diet Coke if you want your ice to last a long,
long time!

Alan Kali

P.S. Here are some web pages on the subject:

Anomalies of hydrochemistry: http://www.martin.chaplin.btinternet.co.uk/explan.html

Chemistry of artificial sweeteners: http://chemcases.com/nutra/nutra1e.htm

Contents of Coke and Diet Coke: http://www.rense.com/general50/class.htm

What happens to corn syrup: http://www.elmhurst.edu/~chm/vchembook/548HFsyrup.html

Salt really does make ice cubes melt slower:
http://antoine.frostburg.edu/chem/senese/101/solutions/faq/why-salt-cools-icewater.shtml

Road salt ate my Pinto: http://science.howstuffworks.com/question58.htm

What the heck is molecular weight:
http://users.rcn.com/jkimball.ma.ultranet/BiologyPages/M/Mole.html

What is salt: http://www.saltinstitute.org/15.html

Hi Alan,

I agree with some of your analysis, but I don't buy the ending part
about the "catalyst" effect.  Salt will accelerate the melting of ice
in most cases by lowering the melting point.  The "solution" (heh,
heh) to the Diet Coke effect is more simple.  I have actually studied
this exact phenomena some with my chemistry classes.

If you take a glass of salt water and a glass of tap water and place
an ice cube in each, you will observe that initially the cube in the
salt water melts rapidly, then appears to almost stop melting.  It
melts rapidly at first due to the ion interference, but as it melts,
it creates a layer of very cold water around itself.  If the salt
water is sufficiently salty, the newly melted fresh water from the
cube floats on top of it.  When enough of the ice has melted, it will
become fully enclosed in a layer of very cold fresh water, that helps
insulate the ice cube and prevent further melting.  So in the regular
Coke, you are actually getting a layer of cold fresh water floating on
the surface of the much more dense syrupy drink below.  The ice cube
melts very slowly in this cold water bath.

In the case of the fresh water, as the ice cube melts, the cold water
formed is actually MORE dense that the room temp fresh water, and it
sinks.  Thus, in the case of the Diet Coke, the cold newly melting
water from the cube merely sinks to the bottom.  It does not form an
insulating layer, and the original ice cube continues to melt at a
relatively constant rate.

You can verify the above effect with a glass of tap water, a glass of
salt or sugar water, some ice cubes, and some food coloring.  You can
use the food coloring to track the convection currents in the water.

Neat effect!


Dave West

Dave,

You have a very interesting hypothesis. The "inadequate mixing"
argument has probably been leveled against experimenters since the
neolithic period. It was the first shot fired against Fleischman &
Pons, although it turned out to be wrong (there was plenty of other
methodology to question).

You are absolutely correct, fresh water is less dense than salt water.
In an undisturbed container, it is possible for the fresh water to
form a a colder layer at the top and slow the melting of the ice.
Since ice is less dense than water, it floats in the fresh water layer
and melts more slowly.

There are a couple of problems with the mixing argument, however.
First, people who dring soft drinks don't usually leave them
undisturbed for very long. There's lots of mixing, which should
circulate the layers and ensure that convection is not the dominant
form of heat transfer.

Another more significant problem is that if your hypothesis is
correct, the ice at the bottom of the container should melt first in a
sugar or salt solution. The less dense and colder fresh water from
melting ice should move to the top, resulting in a temperature
gradient from cooler at the top to warmer at the bottom.

My first experiments after I read this question involved three
identical containers. I added nothing to the first one, salt to the
second one, and sugar to the third one. Then I filled each to the top
with tap water from a pitcher that had been standing until it was at
room temperature and I mixed each one equally. I selected three
identical ice cubes and simultaneously dropped one into each
container. I repeated it twice, and each time the results had the ice
cube in water melting first, the sugar second, and the salt third. The
differences were very significant, and did not appear to involve just
the melting of the last little bit of ice. The melting of the ice cube
in plain water pulled way into the lead from the beginning.

I admit, however, that I did not mix the water. This was an oversight
in my experimental method, and a colder layer or gradient may
certainly have affected the results.

I repeated my experiment after reading your response, but this time I
mixed each container vigorously. The results were the same. The ice
cube in the tap water went fastest, and it certainly looked like it
was in the lead from the start (by visually comparing the ice cube
sizes).

Next, having obtained high quality reagents (Coca Cola and Diet Coke)
that I did not have at my disposal originally, I repeated the
experiment, but this time I used three ice cubes in each container.
The diameter of the containers was such that two ice cubes floated
side-by-side at the top, and one ice cube was underneath, way down in
the liquid. I did not mix.

When about half the volume of the ice cubes had melted, before the
lower ice cube had room to float toward the top, I stopped the
experiment, removed the ice cubes, and compared their sizes. I looked
at the Coca Cola first, and to my surprise, the bottom cube was
smaller! I thought, "Aha, Dave was right!". Then I looked at the Diet
Coke ice cubes, and the bottom one was smaller over there, too.

Then I realized that because the ice floats, one side of each of the
top cubes was in air. Air is a far better insulator than water, so of
course the top cubes were larger - they didn't melt as much on the
side that was exposed to the air! Since both the Coke and the Diet
Coke showed the same pattern, there was no reason to believe that the
differential melting was the result of a temperature gradient.

Alas, my final experiment allowed the melting of the ice cubes to
proceed until complete melting had occurred, and it showed no
difference between the Coke and the Diet Coke. This was a surprise,
since my salt water and sugar water experiments have showed drastic
differences. As I said in my original post, my original experiments
showed differences too large too be missed, even under less than ideal
conditions.

I would have repeated the experiment a few more times with the Coke
and the Diet Coke, but unfortunately my girlfriend drank the reagents.
I'll get some more tomorrow. One thing that may account for my
experimenal findings is that the Coke experiments were done with the
liquid at very cold temperatures, whereas the earlier experiments were
done with the liquid at room temperature. The temperature difference
may diminish or exaggerate the results.

Apparently, the thermal conductivity of salt water is lower than fresh
water, but I'm having a surprising amount of trouble getting numbers
for the values. Mixing the water during the experiment is obviously a
good idea, but my preliminary findings indicate that density and
temperature gradients alone do not explain the phenomenon. Also, the
original question involved not salt water but diet versus non-diet
drinks. Finding the data on the thermal conductivity of sugar water is
going to be very difficult. I have a feeling that sugar water is going
to be an even poorer thermal conductor that salt water, but who knows?
Maybe the salt water slows the ice melting for one reason, and the
sugar water does so for a different reason. I'm still standing by my
conclusion that the presence of salt or sugar alters the thermal
properties of the water. More work to do, I suppose.

Alan Kali

P.S.

I just thought of one other mechanism that would definitely contribute
to the slower melting of the ice in sugar or salt water. It's an
extension from my last comment:

The density of either sugar water or salt water is greater than pure
water, so the ice cube will float with more of its surface out of the
water, projecting into the air. Just like if you try to swim in the
Dead Sea (which is very salty) you end up floating on top of the water
and unable to submerge.

The thermal conductivity of water is enormously higher than air, so
even a small decrease in the amount of the ice cube that is submerged
will result in a large decrease in the rate of melting.

That doesn't account for the totally submerged ice melting faster in
pure water (if, in fact, it does) but it would certainly be part of
the answer. Interestingly, if your fresh water layer hypothesis is
correct, the ice would sink lower into the (less dense) fresh water
layer and therefore tend to melt FASTER!

Alan Kali

It doesn't look like a simple problem.  You can do all kinds of hand
waving but until someone does a scientific study and determines the
contribution of all factors, determines heat conductivity of these
solutions, their melting heats, how different densities affect heat
transfer due to partial submersion, how the density affects convection
and so on, I don't think you would be able to claim that you have
figured it out.

Actually, I have done some very careful experiments and so far the
hand waving and the hot air are the only things that consistently make
the ice melt faster.

Alan Kali