The presence of gases like hydrogen sulphide and carbon dioxide renders petroleum and natural gas products “sour”. These acidic gases cause corrosion; hence, their presence in the final gas products is unacceptable. In natural gas plants and refineries the removal of acid gases was done using alkanolamines. Amines, being weak bases, react with these acid gases to form salts. This salt forming reaction between amines and acids is called gas sweetening.
The amine content in gas sweetening solutions needs to be monitored to optimize the amine content and to profile waste, which minimizes the plant maintenance effort. The type of amine used for gas sweetening varies for each plant. The two main factors that affect the choice of amines are the acid being treated in the gas and the extent to which the acid gas dissolves in the amine.
This article looks into the use of cation exchange chromatography with nonsuppressed conductivity detection for separating several types of amine that can be used in gas sweetening solutions. The amines that are analyzed are: diethanolamine (DEA), 3-methoxypropylamine (MOPA), monoethanolamine (MEA), triethanolamine (TEA), methyldiethanolamine (MDEA), diglycolamine (DGA) and cyclohexylamine. Additionally the separation of amines in the presence of standard cations is also explorede.
The following reagents were used for this experiment:
- Oxalic Acid, CAS 144-62-7
- Nitric Acid, CAS 7697-37-2
- Dipicolinic Acid, CAS 499-83-2
- Ultrapure water, resistivity >18 MO•cm (25 °C)
- Acetone, HPLC grade, CAS 67-64-1
- Monoethanolamine, CAS 141-43-5
- Diethanolamine, CAS 111-42-2
- Triethanolamine, CAS 102-71-6
- 3-methoxypropylamine, CAS 5332-73-0
- Cyclohexylamine, CAS 108-91-8
- Diglycolamine, CAS 929-06-6
- Methyldiethanolamine, CAS 105-59-9
- Custom Cation Mix 2 (REAIC1230, MUSA)
The instrumental setup is given in the table below. Figure 1 shows the experimental setup.
Figure 1. Instrumentation Setup
|940 Professional IC Vario ONE
|IC conductivity detector
|858 Professional Sample Processor – Pump
|MagIC Net TM 3.0
|Metrosep C 6 250/4.0
|10 µL loop
The solutions used were Eluent - 1mM HNO3 + 1.5mM Oxalic acid + 0.75mM Dipicolinic acid + 1% acetone.
Samples and Standards
The sample used was ultrapure water which was spiked with amines and cations. The standards were prepared from certified stock solutions or from bulk chemicals.
The table below gives the composition of ultrapure water.
Samples were injected into the IC directly. The table below shows the IC parameters.
Calculations were done using the peak area of all the analytes via automatic integration with the MagIC Net 3.0 software.
The tables and graphs that follow depict the calibration standards for various cations.
Figure 2. Overlay of calibration standards
Figure 3. Lithium
Figure 4. Sodium
Figure 5. Ammonium
Figure 6. MEA
Figure 7. DEA
Figure 8. Potassium
Figure 9. DGA
Figure 10. TEA
Figure 11. MDEA
Figure 12. MOPA
Figure 13. Magnesium
Figure 14. Calcium
Figure 15. Cyclohexylamine
Among the conditions used, acetone showed a profound effect on separation. The stacked chromatograms depicted in Figure 16 represent 1.25% acetone in eluent (top/red) and 1% acetone in eluent (bottom/black). When the acetone concentration is greater, cyclohexylamine will elute in less than 40min. However co-elution of magnesium, calcium, DEA and MEA is also observed.
Figure 16. Stacked chromatograms
The retention of amines is influenced using other organic modifiers. The comparison of the eluent that is prepared with 1% acetone (black) and the eluent prepared with 1% acetonitrile (red) is shown Figure 17. Co-elution of morpholine and MOPA is observed when acetone is used as an organic modifier. However morpholine and MOPA are separated when acetonitrile is used. An overlap of morpholine over MDEA is observed.
Figure 17. Comparison of eluent
The work presented in this article has shown the possibility of separating amines based on the provided chemistry. Additionally, the effect of an organic modifier on separation has also been demonstrated. The calibration curves presented in this article are based on quadratic regressions. For smaller calibration concentration ranges linear regression may be used.
This information has been sourced, reviewed and adapted from materials provided by Metrohm AG.
For more information on this source, please visit Metrohm AG.