Glucose determination
Observations on glucose content and its changes over time in an "isotonic" soft drink
variety with the help of a personal blood glucose monitor
Background: my contributions to the science-madness group admittedly became much less frequent than
they were a few years ago. I changed jobs and have much less time to pursue amateur chemistry. At least this was the case till these days. :-)
I went to a small(ish) family business two years ago where we produce soft-drinks ("functional drinks", "sports drinks"). Now I am their designated
"quality assurance man" and partly responsible for the "R&D department" too. :-) This department is composed of only two (the wife of the C.E.O
and me) but the leadership likes my "dull man's view" of things and encourage my (sometimes) experimental approach. Well, encourage may not be the
appropriate term. Let's say they don't hinder what I want to do as long as personal safety and lab (building?) safety is not seriously jeopardized.
:-) I like this freedom and started to seek opportunities to do something meaningful with a "madscientist" twist.
Some may remember that a few weeks ago I asked questions regarding the oddities of my experiences with a personal blood glucose monitor and glucose
(sugar) content determinations in a test environment.
Since then I pushed things further and now I can "report" my achievements on the subject.
The basic problem:
We produce "isotonic drinks" (among other things) as well. What is "isotonic" in the field of beverages and soft-drinks is not something with exactly
defined composition. As far as I know there are no exact limits regarding the required ionic composition and/or the resulting osmolality of the
isotonic drink. I use an EFSA guide line which basicaly says that "isotonic is everything where the net osmolality is close to physiological
osmolality (0,3 osm) with a possible deviation say within 10% of this value." I aim to be within the range of 0,27 - 0,3 osmol in the final drink
(because slightly hypotonic is better than slightly hypertonic when it comes to "drinkability" and "hydration" and general mouth feel).
As you can guess, the final 0,3 Osm can be reached in many-many ways but in general if you realistically expect people to actually drink your product
then you don't push the "salt route" too hard.. :-) Yes, a 9 g/l table salt solution would be nicely isotonic (and wonderfully cheap to produce!) but
your customer base would be very limited at best. :-)
So you use some salt, yes, but the bulk of the required osmolality comes from sugar(s)! This is why those isotonic drinks you can buy in a store or a
gym taste sweet and not like salt water.
Sugars you can use: glucose, saccharose, fructose. There are others too, but these are the most frequently used ones due to various reasons.
Of course, the cheapest one is sucrose (saccharose) so you have a strong "urge" to use as much of it as possible in your recipe but there is a caveat:
sucrose is a disaccharide and can invert! Inverting means that a sucrose molecule disintegrates into two smaller monosaccharide molecules and yields
one molecule of glucose and one of fructose. The speed of this reaction, the inverting of sucrose, depends only on two variables: the temperature and
the pH of the solution! Nothing fatal (regarding consumer health) happens as this reaction progresses except two "inconvenience": your carefully
tailored to be isotonic drink may end up being fairly hypertonic well before the declared best before date! And the change in sweetness: the drink
becomes much sweeter than it was at the time of bottling.
At least this was the theory when I started to look after this problem. I could have sent a few bottles of different age isotonic drink samples to a
commercial laboratory and request them to analyze these but in those labs that I know of they don't offer "deep sugar analysis". They can give you a
total sugars result but it won't help for obvious reasons. Other option could have been "osmolarity measurement" (freeze point depression measurement)
but again, it could only give me a half answer. They could tell me if the "isotonicity" of the drink was right or not, but the reason still could
remain uncertain.
So I decided I would use a blood sugar monitor and use it to determine the glucose content of a calibration series and if this series seems right then
I may get the real answer when I measure the actual drink samples too.
The calibration:
The blood sugar monitor is a clever little gadget: I don't want to go into details regarding how it works (amperometry) but encourage everyone to
Google it.
Because we use our own recipes and know the full composition of the drinks we produce it was possible to make a calibration series as close to the
expected composition to the "drinks in question" as it gets. Basically I made a "blank base" with all the components that go into the drink (except
the sugars) then divided it into 10 equal parts. Each part was named after how much saccharose inversion it "suffered" if it was a real drink. E.g.
the calibration sample called "20%" was composed of the fix ammount of blank base and 20% less sucrose than the 0% (unchanged, undecomposed) sample
point. Plus it contained glucose in 10% of the original weight of sucrose and another 10% as fructose.
The blood sugar monitors work with blood samples. Composition of our blood (ionic composition, pH) is fairly tightly controlled by the body. So I made
"artificial blood" too. (Really? Well, not. But I am happy to find at least a madscientist term for a buffer solution.)
This buffer was composed of 20 g/l trisodium citrate dihydrate and 1 g/l citric acid monohydrate. The resulting pH of this buffer was around 6.20 .
This buffer was needed because the glucose concentration of the isotonic drink is much higher than the expected glucose concentrations in the blood so
a dilution (1:12) was required to bring the concentration down to the measurable range. (The pH of human blood is also much higher than 6.20 but 6.20
is still good for the measurement.)
Then I made quite a lot of measurements to be able to acquire the calibration curve. Here are the final, averaged values for the various
"decomposition %-s":
0% decomposition: 12.69 mmol/l*
10% : 13.07
20% : 13.45
30% : 14.17
40% : 14.85
50% : 15.03
60% : 15.23
70% : 16.28
80% : 16.52
90% : 17.50
Ta-da! The points are at least in a nice, monotonous order! :-)
* (The blood sugar monitor reports the result in "mmol/l". But we not really care how it reports it because we dont use it in real blood so this
mmol/l reading is actually meaningless. That only means that the instrument detected the same current like it detected in a blood sample that
contained 12.69 mmol/liter glucose. Our calibration sample surely not contained 12.69 mmol/l glucose BUT IT DOES NOT MATTER because we know how much
it actually contained BECAUSE WE MEASURED IT on our scale when we put together the calibration solutions! In this experiment we are only interested in
relative values!)
We can fit a (not very nice, but still) linear calibration curve on these points: slope: 0.0496, intercept: 12,591.
Measuring actual isotonic drinks from the warehouse:
Now the momet of truth arrived. I measured three different samples of the same type (blood orange flavoured) isotonic drink, manufactured on three
different dates:
Bottled: 26.05.2025 Average: 13.23 Decomposition%: 13.3
Bottled: 03.07.2024 Average: 14.75 Decomposition%: 40.5
Bottled: 01.08.2024 Average:15.17 Decomposition%: 50.8
Discussion:
The 26.05 sample was bottled just a few days ago and already shows 13.3% decomposition?
Yes, and actually this is what I expected! 13.3% seems reasonable because our products get a heat treatment (pasteurization) before bottling. Although
the time of pasteurization is very short (around 90 seconds) but the temperature is much higher than ambient (around 80 Celsius) during this 90 sec. A
good 10% inversion may very well happen during this heat treatment.
The other two samples were made almost a year ago and show much higher glucose content (decomposition) than the few days old batch. Reasonable.
The only "problem" is the actual ratio of the decomposition is higher in the "younger" sample than in the older one. I can find a few somewhat
acceptable reasons to explain this: maybe the initial heat treatment was a bit different (different hold time, different peak temperature hence
somewhat different initial decomposition during pasteur). Or (even more plausible) the storage of these samples happened at slightly different
temperatures. These samples were "randomly" collected from a warehouse where they were stored "randomly". Maybe one was on a lower shelf, hence at
lower general temperature, the other might be stored higher (warmer). Or far from doors - close to a door opening to outside... During storage even
small temperature differences may cause relatively big decomposition differences because the storage time is long!
That's all I can say about this experiment. Of course, I'm open to suggestions, questions. After reading this, maybe someone got excited to try to
check glucose content of other things (besides isotonic drinks) with a blood glucose monitor too!
|
![[*]](images/xpblue/default_icon.gif){kind=icon}
