HSP Application note #22
Hansen Solubility Parameters (HSP) and Carboxylic acid2010.3.26
HSPiP Team Senior Developer, Dr. Hiroshi Yamamoto
I got the HPLC chart of Carboxylic Acids from HPLC vender technical data sheets.
| No | name | RT |
| 1 | oxalic acid | 3.068 |
| 2 | L-tartaric acid | 3.579 |
| 3 | formic acid | 3.831 |
| 4 | L-malic acid | 4.597 |
| 5 | L-ascorbic acid | 5.002 |
| 6 | lactic acid | 5.622 |
| 7 | acetic acid | 5.967 |
| 8 | maleic acid | 7.191 |
| 9 | citric acid | 8.008 |
| 10 | succinic acid | 9.279 |
| 11 | fumaric acid | 10.475 |
| 12 | acrylic acid | 12.471 |
| 13 | propionic acid | 14.53 |
| 14 | glutaric acid | 19.278 |
| 15 | itaconic acid | 23.037 |
ChromatographyThe most popular column for HPLC is ODS column and this column contains Silica-gel covered by Octadecyl. When we insert some chemicals into this column, some chemicals dissolve to octadecyl alkyl chain deeply and some do not. So, highly interacted chemicals will delay to elute. Or some chemicals which are very easily dissolve to carrier liquid, elute very early.
We can evaluate these solubility with Hansen Solubility Parameters (HSP). Molecular size also play important role. Pirika Java Demo Applet design Carrier Solvent. HPLCDemo is available here. |
So I calculated these compounds HSP and simulate HPLC.
I use only calculated HSP only.
Please refer to HPLC article about basic theory.
If I plot Distance from Octadecane(OD) to carboxylic acids divided by Volume to retention time, I got very clear correlation.
The Distance from OD is long, that means hard to dissolve to OD, and elute first.
This result is very reasonable.
If you draw several molecules and calculate each molecules' properties, program will simulate Retention Time (RT) of OSD column for HPLC. If you want to know how to draw molecules, please refer to PowerTools applications. I have full version of this HPLC RT simulation program at PowerTools+ Applications.
But I can not understand this result.
I want to explain the reason.
| Official Value | Calculated Value | ||||||
| No | Name | dD | dP | dH | dDMR | dPMR | dHMR |
| 1 | oxalic acid | 17 | 17 | 26 | 17 | 12.7 | 28.7 |
| 2 | L-tartaric acid | 17.3 | 15.5 | 40.3 | |||
| 3 | formic acid | 14.6 | 10 | 14 | 16 | 12 | 22 |
| 4 | L-malic acid | 17.1 | 12 | 30.1 | |||
| 5 | L-ascorbic acid | 18 | 11.7 | 25.5 | 18.6 | 21.6 | 41 |
| 6 | lactic acid | 17 | 8.3 | 28.4 | 16.3 | 10.7 | 25.9 |
| 7 | acetic acid | 14.5 | 8 | 13.5 | 15.9 | 8 | 16.2 |
| 8 | maleic acid | 16.9 | 10 | 22 | |||
| 9 | citric acid | 17.9 | 9.3 | 31.3 | |||
| 10 | succinic acid | 16.9 | 8.3 | 19.3 | |||
| 11 | fumaric acid | 16.9 | 10 | 22 | |||
| 12 | acrylic acid | 15.63 | 8.72 | 14.5 | 15.7 | 7.7 | 14.6 |
| 13 | propionic acid | 14.7 | 5.3 | 12.4 | 15.8 | 6.4 | 12.8 |
| 14 | glutaric acid | 16.6 | 8.9 | 16.5 | |||
| 15 | itaconic acid | 16.6 | 8.4 | 18.1 | |||
HSP estimation accuracy of Carboxylic Acids is not so high.
HSP technology is based on Heat of Vaporization.
(Please refer to Heat of Vaporization calculation more detail.)
And heat of vaporization and boiling point, there is a correlation shown with below chart.
From the rule of Trouton,
Lb/Tb=21 cal/g-mol
for none polar solvents.
This rule can be expanded to Heat of vaporization at 25 C.
And expanded to if functional group thread is identical, it make parallel line.
Only exception is carboxylic acid.
The compounds boiling point become larger than 480K, then it become almost parallel to hydroxy group compounds. But below 480C compounds, Heat of vaporization sharply dropped and it goes down even hydrocarbon compounds.
These phenomena is very popular for chemical engineering area especially distillation.
Very small Carboxylic Acids vaporize with dimer.
2011.6.11
|
In distillation area, they believe that carbon number below 3 make dimmer.
But I compiled several kinds of molecules that have almost same boiling point and plot of temperature dependency of heat of vaporization.
I can say even isobutyric acid, it is abnormal.
I have strong temptation introduce donar/aceptor terms into HSP to increase accuracy.
But before introducing such terms, I have so a lot to do for these hydrogen bonding terms.
Anyway, now I can understand why CALCULATED HSP of carboxylic acids have good correlations with HPLC retention time.
Calculated HSP ignore the dimer effect, and in HPLC they use 0.1% of H3PO4/aq as mobile solvent, so no dimer exist.
