# Separating sugar from soda

I want to do a lab with my students in which they will try to prove how much sugar is in a can of coke. I know that distillation is the go-to method for separating out the syrupy sugar mix, but does anyone know of a method that goes further to separate out the pure granulated sugar and not just a syrup?

The simplest approach would be to degas a sample and measure its specific gravity with a vintner's hydrometer. It will have scales of specific gravity and Brix (sugar content) on it.The Brix (Bx) reading is the percent of sucrose, w/w, in a solution with the same density/specific gravity as the one you are testing. There are, of course, other things in coke besides sucrose (fructose, coca extract, phosphoric acid) but they should be in small enough concentration that the sugars will swamp everything else and all sugars behave amazingly the same with respect to density of their solutions).

Based on the reported sugar content of Coke it contains 108 grams of sugar per liter. Thus its specific gravity should be about 1.041 and the corresponding Bx reading about 10.4. The density of water at 20 °C is 0.998203 so the density corresponding to 1.041 SG is 1.039 and 10.4% of 1039 grams (weight of a liter of 10.4 Bx sucrose solution) is 108 grams.

If you don't have a hydrometer with the Bx (or Plato) scale or if you determine density by weighing a known volume (convert to specific gravity) you will need to be able to determine the sucrose content of a solution of given specific gravity. There is lots of opportunity here to go into the history of wine and beer making, which you male students will love (though I understand that in modern schools discussion of things related to alcohol may not be politically correct) but in any case determining the w/w sucrose content of solutions of given SG is a well studied problem. There are various tables but the easiest way to get w/w sucrose from specific gravity is a polynomial promulgated by the ASBC (American Society of Brewing Chemists). It is

°P = ((135.997*S -630.272)*S + 1111.14)*S -616.868

I believe that separating out the pure sugars at anything close to 100% recovery would be difficult (barring HPLC or something of that sort). Sugars seem to be assayed by indirect methods such as the gravimetric ones I proposed or measurement of refractive index.

• Degassing will be fast(er) with an ultrasound bath, or a foamer as «popularized» e.g., by Guinness as surger (video). Jan 21, 2022 at 19:51
• The Journal of Chemical Education has a lab experiment that fits here. The calibration goes uses a linear regression with water, and samples of 5, 10, and 15 weight% of sugar and from this infers 11.9 weight% of sugar in (then available) Minute Maid apple juice. Henderson, S. K.; Fenn, S. A.; Domijan, D. D. Determination of Sugar Content in Commercial Beverages by Density: A Novel Experiment for General Chemistry Courses. J. Chem. Educ. 1998, 75, 1122-1123; doi 10.1021/ed075p1122. Jan 21, 2022 at 20:08
• I suppose another tractable, if tedious, method would be to (i) assume nearly all of the organic dissolved solids in the Coke is sugar, (ii) to dry the degassed Coke, and (iii) to perform some sort of calorimetry on the combustion of the solid residue. Jan 21, 2022 at 23:47

If the amount of sugar is to be determined precisely in a given solution, the simplest thing is to mix it with a $$0.2$$ M $$\ce{HCl}$$ solution, to heat it up to boiling for ten minutes, and cooling it down to room temperature. This will hydrolyze the saccharose to a mixture glucose + fructose. Then, the glucose is determined by Fehling's solution, producing a $$\ce{Cu2O}$$ precipitate, which can be weighed or titrated. The procedure is described in all books of analytical chemistry.

• In North America at least, most soft drinks are sweetened with high fructose corn syrup (which doesn't suppress hunger and metabolizes into lipids, not glucose: hello obesity), so there might be no sucrose and not much glucose there. How will that affect this procedure? Jan 22, 2022 at 2:05
• @Ray Butterworth: one of the deceptive names in food is "high fructose corn syrup". Sucrose is 50%/50% fructose/glucose. HFCS is 42% or 55% fructose on a dry basis (en.wikipedia.org/wiki/High-fructose_corn_syrup). What makes the fructose high is the low water content, ~24% in the syrup. Of course, that also makes the glucose high. Coke claims to use sucrose in its Coke, but there are varieties of Coke, so maybe some use HFCS. Jan 23, 2022 at 15:18

How about precipitating the sucrose? Acetone and CaCl2 can precipitate up to 98% of the sucrose. There are other precipitation schemes on the web that use complex organic compounds of transition metals, but they seem more useful for analyses than for actual separation.

I can't do any better than to copy and paste an abstract (the whole article is available at the reference url):

"The use of methanolic CaCl2 for the non-aqueous extraction of sucrose from dried sugar beet (Beta vul­ garis L.) cossettes was investigated, with emphasis on the effects of CaCl2 concentration, time of extraction, and operating temperature. Solubility of sucrose from dried cossettes was optimal in range of 10-15% CaCl2 in methanol and increased by a factor of four as the temperature was raised from lC to 60C. Nearly complete extraction (98%) was achieved in batch experiments with four successive extractions. The methanolic CaCl2 extract was found to have both higher purity and greater sucrose stability compared with aqueous diffusion juice. The sucrose could be precipitated completely from the solution by the addition of 5 volumes of acetone at 20C, leaving 9~/o of the CaCl2 in solution. Steady state conditions could not be reached in simulated continuous column experiments, probably because of calcium exchange with other cations in the beet pectin. Pure sucrose was precipitated from these effluent solutions on cooling. The amount of sucrose recovered was pro­ portional to the decrease in calcium content on pas­ sage through the column. From this, the molar ratio of calcium to sucrose in the methanolic CaCI2­ sucrose complex was calculated to be 2:1." Ref 1.