Monitoring the Preparation of Biodiesel with Benchtop NMR

Sustainability has become a main focus of innovation.  Topics of green chemistry, and renewable feedstocks, are paramount in society today.  Biodiesel is one such example!  Although Rudolf Diesel dreamed of powering his diesel engine on edible oils in the 1890’s, it didn’t become prevalent until about a hundred years later.

Biodiesel is a fuel that is acquired from oils, fats and greases.  Why are we interested in this?  Well, because unlike non-renewable fuels like diesel, oils are derived from plants that can be replanted, grown and harvested from many sources each year.  Once the oil is isolated, biodiesel can be easily afforded through a relatively simple organic reaction: transesterification.

Biodiesel is a lubricating, clean burning, less viscous, and high-heating value fuel.  Regardless, typically biodiesel isn’t used to run a car exclusively, we usually buy B5 and B20 blends that are made up biodiesel/diesel.

There are many synthetic procedures that can be followed to afford biodiesel – I choose to modify this one[1]:

1 ) Weigh 3 g of oil (the example shown here uses palm oil) into a reaction vessel charged with a stir bar.  Assume it has an average molar mass of 880 g/mol.
2 )  Prepare an NMR tube sample of your oil starting material by adding 0.1 mL to 0.5 mL d-CHCl3.  Collect the 1H NMR spectrum.
3 ) Add 6 mmol of anhydrous MeOD for every mmol of oil.  Stir well and heat to 60 oC.
4 ) While heating, prepare 10 NMR tubes by adding 0.5 mL of d-CHCl3 to each one.
5 ) Calculate the amount of anhydrous K2CO3 required so that it is 6% the mass of oil.
6 ) Add the K2CO3 to the reaction, stir for 30 seconds then take a 0.1 mL aliquot and add this to one of the prepared NMR tubes.  Acquire a 1H NMR spectrum.
7 ) Heat the reaction mixture to 60 oC for 45 minutes, taking a 0.1 mL aliquot and acquiring a 1H NMR spectrum every 3-5 minutes.
8 ) Cool the reaction mixture to room temperature and slowly add acetic acid.
9 ) Using a separatory funnel, isolate biodiesel.
10 ) Wash 3 times with H2O, collect, dry over MgSO4 & filter.
11 ) Determine the volume & weight of the biodiesel.
12 ) Measure the synthesized biodiesel’s melting point.

If you make a stacked plot of all the spectra collected at different time intervals, at first it looks quite complex.  To make this easier to understand, the first thing to do is to examine the spectrum of the small molecule components.  That is, look at proton NMR spectra of free fatty acid.  The saturated ones – e.g., palmitic (C16:0),stearic (C18:0) – aren’t very interesting so I didn’t include them here.  You know, you see a lot of the same: an alpha –CH2 , a slough of–CH2’s, and a terminal methyl group.  The unsaturated spectra are a little more informative so we’ll examine them a little more closely – see oleic (C18:1), linoleic (C18:2), and a-linolenic (C18:3)).

They all have the same chemical moieties:

1)   Carboxylic acid – peak around 11-12 ppm.
2)   Olefin (integrating to 2, 4, or 6 – depending on whether we’re looking at omega-9, 6, or 3, respectively) – observable around 5 ppm.
3)   bis-allylic peaks seen in the 2.5-3 ppm range.  These are only present in the top 2 spectra – i.e., polyunsaturated omega 3 and 6 – as there is no bis-allylics in omega-9.
4)   Alpha-CH2‘s (to the carbonyl/olefin) – in the 2.0 – 2.5 ppm range.  These are moved downfield, relative to the bulk CH2’s, due to the electronic nature of their neighbours.
5)   The chain -CH2’s – observable in the 1.0-1.5 ppm range.
6)   Terminal -CH3 in the 0.5-1.0 ppm range

Okay – now that we have the small molecules under our belt, we’re finally ready to move up to edible oils.  Oils are triglycerides of fatty acids (just think of that as a three-armed fatty acid chain with a glyceride linker), so they’re sort of oligomeric.  If you’re concerned – DON’T BE!  We already did the hard work assigning the free fatty acid spectra – so now the edible oil spectra are just as easy to understand!  We see the same main peak regions, but: (a) they’ll just integrate differently relative to each other; and (b) the carboxylic acids are gone, and replaced with the glyceride –CH2’s (at 4-4.5ppm).

Here’s 4 examples – canola, sunflower, mustard and peanut oil.  The intensity of each peak depends only on their typical compositions.

Okay – now we understand the ‘complex’ spectra, analyzing our kinetic data is just as effortless!  Everything stays the same (i.e., olefins, bis-allylic, allylic/alpha-C, chain -CH2’s and -CH3) EXCEPT the glyceride peak disappears – it, of course, becomes glycerol in the reaction – and the methyl ester peak grows in as the transesterification reaction proceeds and we convert our edible oil to biodiesel. 

Sooo……we can monitor the whole reaction in the 3-4 ppm range and pay no mind to anything else.  Again, the triglyceride alkyl peaks (4-4.5 ppm) get consumed, as does the MeOD whereas the methyl ester grows in.

If we integrate these regions, and then plot them, we can get a speciation curve showing us the components that are present during various stages of the reaction.  Plotting the disappearance of the triglyceride alkyls as a function of time versus (1) I(t); (2) ln (I(t)); and (3) 1/I(t) you can also determine the order of the reaction.

If you’re interested in attempting this or learning more please see our sample experiments here.

[1] Behnia, M.S., Emerson, D. W., Steinberg, S. M., Alwis R. M., Duenas, J. A., Serafino, J. O., J. Chem. Educ. 2011, 88, 1290