NMR studies of mixed amines

Available online at www.sciencedirect.com

   

Energy Procedia 4 (2011) 291–298

Energy Procedia 00 (2010) 000–000

www.elsevier.com/locate/procedia www.elsevier.com/locate/XXX

GHGT-10

NMR studies of mixed amines. Mat Ballard*1, Mark Bown, Susan James and QiYang Energy transformed Flagship, CSIRO Materials Science and Engineering, Bayview Ave, Clayton, VIC 3168, Australia Elsevier use only: Received date here; revised date here; accepted date here

Abstract 1

H and quantitative 13C NMR spectroscopy was used to study the reaction of CO2 with mixed aqueous amines. This technique permits the measurement of carbamates of each amine and bicarbonate concentrations as a function of time (and loading). The ratio of the amount of each carbamate formed (for primary and secondary amines) is shown to be thermodynamically rather than kinetically controlled, due a rapid exchange reaction between amine and carbamate. This exchange reaction is the likely basis of the "shuttle" mechanism of rate enhancement in mixed amines.

c© 2010 ⃝ 2011 Elsevier Published Elsevier Ltd. Open access under CC BY-NC-ND license. Ltd.byAll rights reserved Keywords: carbon dioxide; amine; absorption; NMR; 13C; kinetics; competition

1. Introduction There is currently strong concern that CO2 emission from human activities, especially from industrial consumption of fossil fuels, is driving climate change1. Scientists and engineers have investigated the development of more efficient technologies and methods to reduce CO2 emission from human activities. Post combustion CO2 capture2 (PCC) is regarded as one of the more promising processes to reduce CO2 emission from industry due to the ability to retrofit PCC to existing industrial facilities and processes3. Using alkanolamines in a PCC procedure to capture CO2 is one of the most efficient techniques. Although monoethanolamine (MEA) has been used in oil and natural gas industries to capture CO2 gas for many years, the high energy consumption of CO2 desorption makes it less satisfactory for CO2 capture in the energy industry. The flue gases emitted from power stations have low CO2 partial pressures and, in many cases, contain other acidic, corrosive and oxidative gases or components, such as NOx, SOx and unused O2 gases4. Therefore, the development of new amine solutions which have the desired properties is a key issue for the common and widespread application of amine solutions to CO2 capture. While monoethanolamine (MEA) has long been used to capture CO2, recent work has focused on alternative amines with better performance. In particular, blends of amines are attractive5: not only do they offer the possibility of combining the best characteristics of different amines, there is also evidence that certain amines, in particular

* Corresponding author. Tel.: +61-3-9545-2425; fax: +61-3-9545 2446. E-mail address: [email protected].

doi:10.1016/j.egypro.2011.01.054

2292

M. Ballard / EnergyProcedia Procedia (2011) 291–298 Mat Ballard etetalal./ Energy 00 4(2010) 000–000

piperazine (Pz), can act as “promoters” and lead to increase in the rate of absorption above that of the individual amines6, 7, 8. 13

C NMR spectroscopy has been used previously to study the absorption of CO2 by amines9, 10, 11, and in previous work12 we reported on the use of 13C NMR spectroscopy to study the speciation between carbamate and bicarbonate that occurs in the absorption and desorption of CO2 by MEA. It has also been used recently to study mixed amines in a continuous absorption/desorption system under steady-state conditions13, but only one of three amines in the mixtures could form a stable carbamate. In this work we report the use of 1H and 13C NMR spectroscopy to study the absorption of CO2 by mixed amines. In these experiments CO2 is absorbed by a mixed solution of two different amines (eg: MEA and Pz), and the relative amounts of their carbamates, bicarbonate (and other species) determined by NMR, as a function of time / CO2 dose.

2. Experimental We employed methods similar to our previous work12, using Bruker 400 and 500 MHz NMR spectrometers. Solutions of amines of known concentration were made up by weighing two amines into one 25 ml volumetric flask, and making up to volume with deionised water containing 10% D2O to act as a frequency lock. In runs 1-5 the first amine was MEA, at a concentration of 2.00 mol dm-3. In run 6 the first amine was N-methyl monoethanolamine (MMEA), and in run 7 the first amine was methyldiethylanolamine (MDEA), both at a concentration of 2.00 mol dm-3. MMEA, Pz, and MDEA were employed as second amines. These solutions (15 ml) were then transferred to a 50 ml jacketed pear-shaped flask thermostatted to 40° C, and stirred at 900 rpm. CO2 was bubbled into the solution at a known rate (usually 5 ml per minute through a PTFE tube of ID 0.71 mm), and samples (~ 0.5 ml) taken at suitable times and placed in a sealed NMR tube with a capillary of 1,4-dioxane as an external standard. Quantitative 13 C NMR was performed on these samples within 12-18 hours. 1H NMR was also performed on some samples within 15 minutes of sampling to check the results of the 13C analysis and ensure that further post-sampling reaction was minimal. The identity of some of the 13C NMR signals was determined via a long-range 1H-13C correlation experiment (HMBC) on a sample from Run 7 (MDEA + Pz) at 60 minutes, on a Bruker 500 MHz NMR spectrometer. Correlations observed between the NCH2 hydrogens and the carbonyl carbon identified the covalently bonded carbamate species. The following table summarizes the concentrations of all amines used in these experiments. Table 1: Amines and concentrations First Amine Run Name / Conc. (mol dm-3) MEA / 2.00 1

Name / Conc. (mol dm-3) MMEA / 2.00

Second Amine Structure N H

OH

Conc. amino groups (mol dm-3) 2.00

2

MEA / 2.00

Pz / 1.20

3 4 5

MEA / 2.00 MEA / 2.00 MEA / 2.00

Pz / 0.60 Pz / 0.30 MDEA / 2.00

see (2) see (2)

1.20 0.60 2.00

6

MMEA / 2.00

Pz / 1.20

see (2)

2.40

2.40

293 3

M. Ballard al. //Energy Procedia00 4 (2011) 291–298 Author et name Energy Procedia (2010) 000–000

MDEA / 2.00

7

Pz / 1.20

see (2)

2.40

In runs 1-5 MEA was employed because its kinetics are well-understood and its rate coefficients well-known14. In runs 6 and 7 MMEA and MDEA were employed to examine the behaviour of secondary and tertiary amines. Pz was used in five of the runs because it is commonly used as a “promoter”8. Two additional experiments were performed: a solution of MEA (4.0 mol dm-3, 15 ml) was bubbled with CO2 for 2 hours to give a carbamate loading of approximately 0.44 mol/mol (by 1H NMR). In run 8 a small quantity (0.5 ml) of this loaded MEA solution was mixed with a second amine MMEA (4.0 mol dm-3; 0.5 ml), and 1H NMR spectra taken at various times. In run 9 MDEA was used instead of MMEA.

3. Results 1

H and 13C NMR stack plots for run 1 are shown in Figures 1 and 2 below.

Fig 1: 1H stack plot for Run 1, MEA + MMEA:

Fig 2: 13C stack plot for Run 1, MEA + MMEA:

240min

240min

180min

180min

120min

120min

60min

60min

30min

30min

10min

10min

3.9

3.8

3.7

3.6

3.5

3.4

3.3

3.2

3.1

3.0

2.9

2.8

2.7

2.6

2.5

2.4

2.3

2.2

ppm

166

164

162 ppm

65

60

55

50

45

40

35

ppm

As can be seen, the chemical shifts change with time, particularly in 1H NMR (the signals at 3.75 ppm and 67.18 ppm in Figures 1 and 2 respectively are the external 1,4-dioxan standard). As well, there is considerable broadening and overlap of the signals after about 60 minutes in 1H NMR, making it difficult to both assign peaks and determine peak areas. While there is occasionally uncertainty in the assignment of the aliphatic carbons in 13C, particularly with Pz, this makes no difference to the calculated carbamate and bicarbonate concentrations. These changes are likely to be due to the decrease of pH with increasing CO2 loading, as seen in Figures 3 and 4 below. Similar results were observed for all runs: the 13C results showed far less change and overlap with pH than the 1H results. Accordingly, the 1H NMR results have only been used to check the 13C NMR results for some of run 1.

4294

M. Ballard / EnergyProcedia Procedia (2011) 291–298 Mat Ballard etetalal./ Energy 00 4(2010) 000–000

Fig 3: change of 1H chemical shift with pH: 3.50

Fig 4 : change of 13C chemical shift with pH: 100

a. NH2-CH2-CH2-OH

165 164

90

b. CH3-NH2-CH2-CH2-OH

3.00 d. CO2-NH-CH2-CH2-OH e. CH3-NHCOO-CH2-CH2OH 2.50

f. NH2-CH2-CH2-OH

Chemical Shift (ppm)

Chemical Shift (ppm)

c. CH3-NCOO-CH2-CH2-OH

10

11

12

70 60 50

MMEA - C1 - species 2 MMEA - C2 - species 1

160

MMEA - C2 - species 2

159

MEA - C2- species 1

158

MEA - C2 - species 2 MMEA - C3 - species 2 MMEA - C3 - species 1, 3

155 8

b2

MMEA - C1 - species 1

161

156

30

2.00 9

162

40

h. CH3-NH-CH2-CH2-OH 8

80

157

g. CH3-NH-CH2-CH2-OH

MEA - C1 - species 1 MEA - C1 - species 2

163

9

10

11

12

MEA Carbamate MMEA Carbamate Bicarbonate

pH

pH

The following graphs show the concentrations of carbamates of both amines, and bicarbonate, as a function of time, in runs 1-7. Fig 5: MEA + MMEA:

Fig 6: MEA + Pz (1.2 M):

2.0

1.5

1.5

Conc (M)

Conc (M)

1.0

1.0

0.5

0.5

0.0

0.0

0

60

120

180

240

0

60

Tim e (m in)

120

180

240

180

240

Tim e (m in)

Fig 7: MEA + Pz (0.6 M):

Fig 8: MEA + Pz (0.3 M): 1.5

1.0

1.0 Conc (M)

Conc (M)

1.5

0.5

0.5

0.0

0.0 0

60

120 Tim e (m in)

180

240

0

60

120 Tim e (m in)

295 5

M. Ballard al. //Energy Procedia00 4 (2011) 291–298 Author et name Energy Procedia (2010) 000–000

Fig 9: MEA + MDEA:

Fig 10: MMEA + Pz (1.2 M): 2.0

2.5

2.0

Conc (M)

Conc (M)

1.5 1.5

1.0

1.0

0.5 0.5

0.0

0.0 0

60

120

180

0

240

60

120

180

240

Tim e (m in)

Tim e (m in)

Fig 11: MDEA + Pz (1.2 M):

Fig 12: HMBC experiment on Pz:

2.0 ppm

161.5

1.5

Conc (M)

162.0

1.0

162.5

163.0

0.5 163.5

164.0

0.0 0

60

120

180

240

164.5

Tim e (m in) 3.20

3.15

3.10

3.05

3.00

2.95

2.90

2.85

165.0 ppm

1st Amine (usually MEA) Carbamate; 1st Amine Carbamate by 1H; 2nd Amine Legend: nd nd Carbamate, species 1; 2 Amine Carbamate, species 2; 2 Amine Carbamate, Total; 2nd 1 12 Amine Carbamate by H; Bicarbonate; MEA Carbamate (4M MEA, previous work ); Bicarbonate (4M MEA); Unknown. All runs follow a similar overall pattern: the total amount of CO2 absorbed increases linearly for the first 120 minutes (ie: the rate is constant), then increases more slowly and finally reaches a plateau. The rate of formation of amine carbamate is constant up to about 60 minutes, around which time bicarbonate starts to form. When two carbamates can form, one amine then continues at the same constant rate, while the other slows down, and “accommodates” bicarbonate formation, until about 120 minutes. In the run 1 (Figure 5), the rate of formation of MEA carbamate is about 3 times faster than MMEA. In runs 2-4 (Figures 6-8) MEA is again about 3 times faster than Pz (taking concentration into account). In run 6 (Figure 10), MMEA and Pz carbamates form at about the same rate, which is broadly consistent with those results for runs 1-4. These results for primary and secondary amines suggest that up to about 60 minutes the two amines engage in a kinetic competition15, 16 for a limited amount of CO2, and that the rates of reaction reflect the relative rate coefficients for carbamate formation of each amine. There is the possibility that this is not the case: if carbamates are unstable they might break down rapidly (to form amine and bicarbonate) and the relative amounts of each carbamate will come under thermodynamic rather than kinetic control: that is, be at, or close to equilibrium. We initially considered this unlikely as we had earlier observed MEA carbamate to be stable up to 90º C 12. On the other hand, evidence from the literature17 suggests that Pz reacts far more rapidly than MEA with CO2, so one would expect far more Pz carbamate than MEA carbamate, which is not the case in Figures 6-8. We therefore carried out runs 8 and 9 to check this possibility.

6296

M. Ballard / EnergyProcedia Procedia (2011) 291–298 Mat Ballard etetalal./ Energy 00 4(2010) 000–000

In run 5 (Figure 9), MEA (2M) + MDEA (2M), in which results from our previous work on MEA (4M) alone are plotted for comparison, MEA carbamate again reaches a maximum concentration at about the same time and mol fraction loading. What is different is that bicarbonate forms earlier – but its final concentration is no higher than in pure MEA, and the total CO2 loading at long times is lower. This suggests that MDEA is not very effective in combination with MEA. In run 7 (Figure 11), MDEA (2M) and Pz (1.2 M), we see that significant amounts of a second Pz carbamate forms at moderate times and loadings, but speciates in favour of bicarbonate at long times / high loadings (2.5 M CO2 dm-3). This is in contrast to Pz alone (at 2 M), where the formation of carbamate is favoured over bicarbonate by about 3:1, and the overall loading is lower (2.0 M CO2 dm-3)18. Given the propensity for bicarbonate to rapidly transform to CO2 gas at moderate desorption temperatures, this means that MDEA appears to be effective in combination with Pz. In run 8, in which an MEA carbamate solution and a MMEA solution were mixed, the first 1H NMR spectra was taken 2 minutes after mixing. It showed that MMEA carbamate had rapidly formed, and composed 15% of the total carbamate signal present. At 74 minutes the ratio was 16%, so the mixture was close to equilibrium after 2 minutes. These compare with a value of 24% (by 13C NMR) at 60 minutes in run 1. This means that MEA carbamate can rapidly react with MMEA to effectively “exchange” carbamate groups. In contrast, run 9, in which MEA carbamate and MDEA were mixed, 1H NMR spectra showed that there was no significant decrease in MEA carbamate concentration up to 9 hours after mixing. Further, 1H spectra on the mixture at that time gave a MEA carbamate mol fraction (loading) of 0.45, compared to 0.44 in the original CO2 loaded MEA solution. 13C analyses of the same samples yielded loadings of 0.54 and 0.58 respectively (note that the quantitative accuracy of 1H was about ± 20%, whilst that for 13C was about ± 3%, due to the former being optimized for sample throughput). In other words, the MEA carbamate did not break down. Finally, in Figure 12, the HMBC 3-bond coupling experiment on Pz (with MDEA) shows that the second 13C signal at 163 ppm originated from a carbon atom close to the protons of a N-CH2 group: that is, this is a second Pz carbamate signal originating from a Pz bis-carbamate, as observed by Bishnoi and Rochelle17. Note that this second carbamate only appears at higher Pz concentrations and moderate loadings.

4. Discussion These results show that while two amines, primary and/or secondary, may compete for CO2, carbamate formation can be followed rapidly by a further exchange reaction leading to an equilibrium mixture of the two amine carbamates. This is most likely due to carbamic acid breaking down to reform the original amine plus CO2: MEA-COOH

MEA + CO2 (aq)

MMEA + CO2 (aq)

MMEA-COOH

(1) (2)

and of course very similar reactions involving H2CO3, or carbamate and bicarbonate. The timescale of reaction (1) is of order 30 s 14, and that of reaction (2) is < 1s, so effectively, this amounts to a rapid exchange reaction: MEA-COOH + MMEA

MMEA-COOH + MEA

(3)

When only one amine is present, this is a zero-sum reaction. This resolves the apparent paradox of MEA carbamate being stable at 90º C 12, and yet reacting with MMEA on a timescale of a minute or so (run 8) at room temperature.

M. Ballard al. //Energy Procedia00 4 (2011) 291–298 Author et name Energy Procedia (2010) 000–000

297 7

When two primary and/or secondary amines are present, this leads to a rapid equilibration in the carbamate concentrations since the reaction is symmetrical and reversible. This reaction is likely to be the chemical basis of the “shuttle” mechanism first proposed by Astarita et al.19, as it applies to mixed amines8, 20, 21, 22, 23. This means that the relative rates of reaction of each amine in these experiments are not just a measure of the rate coefficients for carbamate formation, but a more a reflection of the relative thermodynamic stabilities (equilibrium constants) of each amine carbamate. However, when only one amine can form a carbamate, for example when the other amine is tertiary, this exchange reaction does not take place, and so the “shuttle” mechanism as it is cannot account for any rate enhancements in those cases. While the bis-carbamate of Pz has been observed previously17, it has not been seen at the relatively high concentrations observed in run 7. Modeling of the absorption of CO2 by Pz alone or in mixtures must therefore take into account the likely differing reactivities of Pz and its mono-carbamate towards CO2 and its sub-species.

5. Conclusions 13

C and 1H NMR are powerful tools for investigating the reactions of mixed amines with CO2. In this work we have used it to determine speciation in mixtures in two different amines. Primary and secondary amines can compete with each other for a limited amount of CO2, but then rapidly equilibrate, most likely via a rapid exchange reaction between carbamate and free amine. Tertiary amines such as MDEA cannot equilibrate rapidly via this mechanism. In future work, we will attempt to obtain relative rate coefficients by fitting these results to a complete chemical model along the lines of McCann et al14. We will also use the model to check that this “exchange” reaction is consistent with the known rate coefficients for MEA, and likely rate coefficients for MMEA.

6. Acknowledgments We acknowledge the assistance of Jo Cosgriff and Roger Mulder for their constructive criticism and invaluable help in methodology development. This project was carried out within the Research stream on Post-Combustion Capture in CSIRO’s Coal Technology Portfolio. It received funding from the Australian Government as part of the Asia-Pacific Partnership on Clean Development and Climate. The views expressed herein are not necessarily the views of the Commonwealth, and the Commonwealth does not accept responsibility for any information or advice contained herein.

7. References (1) IPCC report on Climate Change; Cambridge University Press, 2007. (2) IPCC report on Carbon Dioxide Capture and Storage; Cambridge University Press, 2005. (3) IEA Greenhouse Gas R&D Programme CO2 Capture and Storage. http://www.co2captureandstorage.info/co2db.php (accessed March 23, 2010). (4) Hakka, L. E.; Ouimet, M. A. Method for recovery of CO2 from gas streams. US 7056482, March 2006. (5) Appl, M.; Wagner, U.; Henrici, H. J.; Kuessner, K.; Volkamer, K.; Fuerst, E. Removal of CO2 and/or H2S and/or COS from gases containing these constituents. US 4336233, 1982. (6) Xu, G. W.; Zhang, C. F.; Qin, S. J.; Wang, Y. W. Kinetics study on absorption of carbon-dioxide into solutions of activated methyldiethanolamine. Ind. Eng. Chem. Res. 1992, 31, 921-927. (7) Mandal, B. P.; Biswas, A. K.; Bandyopadhyay, S. S. Absorption of carbon dioxide into aqueous blends of 2-amino-2-methyl-1-propanol and diethanolamine. Chem.Eng.Sci. 2003, 58, 4137-4144. (8) Weiland, R. H.; Sivasubramanian, M. S.; Dingman, J. C. In Effective Amine Technology: Controlling Selectivity, Increasing Slip, and Reducing Sulfur, 53rd Annual Laurance Reid Gas Conditioning Conference, Norman, OK, February 2003.

8298

M. Ballard / EnergyProcedia Procedia (2011) 291–298 Mat Ballard etetalal./ Energy 00 4(2010) 000–000

(9) Suda, T.; Iwaki, T.; Mimura, T. Facile determination of dissolved species in CO2-amine-H2O system by NMR spectroscopy. Chem. Lett. 1996, 777-778. (10) Jakobsen, J. P.; Krane, J.; Svendsen, H. F. Liquid-phase composition determination in CO2-H2O-alkanolamine systems: An NMR study. Ind. Eng. Chem. Res. 2005, 44, 9894-9903. (11) Barzagli, F.; Mani, F.; Peruzzini, M. A C-13 NMR study of the carbon dioxide absorption and desorption equilibria by aqueous 2aminoethanol and N-methyl-substituted 2-aminoethanol. Ener. Env. Sci. 2009, 2, 322-330. (12) Yang, Q.; Bown, M.; Ali, A.; Winkler, D.; Puxty, G.; Attalla, M.; CSIRO A Carbon-13 NMR Study of Carbon Dioxide Absorption and Desorption with Aqueous Amine Solutions. Ener. Proc. 2009, 1, 955-962. (13) Barzagli, F.; Mani, F.; Peruzzini, M. Continuous cycles of CO2 absorption and amine regeneration with aqueous alkanolamines: a comparison of the efficiency between pure and blended DEA, MDEA and AMP solutions by 13C NMR spectroscopy. Ener. Env. Sci. 2010, 3, 772-779. (14) McCann, N.; Phan, D.; Wang, X.; Conway, W.; Burns, R.; Attalla, M.; Puxty, G.; Maeder, M. Kinetics and Mechanism of Carbamate Formation from CO2(aq), Carbonate Species, and Monoethanolamine in Aqueous Solution. J. Phys. Chem. 2009, 113, 5022-5029. (15) Mullins, R. J.; Vedernikov, A.; Viswanathan, R. Competition Experiments as a Means of Evaluating Linear Free Energy Relationships. An Experiment for the Advanced Undergraduate Organic Chemistry Lab. J. Chem. Ed. 2004, 81, 1357-1361. (16) Holm, T. Use of Competition Kinetics with Fast Reactions of Grignard Reagents. J. Org. Chem. 2000, 65, 1188-1192. (17) Bishnoi, S.; Rochelle, G. T. Absorption of carbon dioxide into aqueous piperazine: reaction kinetics, mass transfer and solubility. Chem.Eng.Sci. 2000, 55, 5531-5543. (18) James, S.; Yang, Q.; Ballard, M.; Bown, M. A Quantitative 13C NMR Study of CO2 Absorption and Desorption with Aqueous Amines Part 3 – Ring Amines. in preparation. 2010. (19) Astarita, G.; Savage, D.; Longo, J. Promotion of CO2 mass transfer in carbonate solutions. Chem.Eng.Sci. 1981, 36, 581-588. (20) Lu, J.; Wang, L.; Sun, X.; Li, J.; Liu, X.; School of Chemical Engineering, Nanjing University of Science and Technology, Absorption of CO2 into Aqueous Solutions of Methyldiethanolamine and Activated Methyldiethanolamine from a Gas Mixture in a Hollow Fiber Contactor. Ind. Eng. Chem. Res. 2005, 44, 9230-9238. (21) Bonenfant, D.; Mimeault, M.; Hausler, R. Comparative analysis of the carbon dioxide absorption and recuperation capacities in aqueous 2(2-aminoethylamino)ethanol (AEE) and blends of aqueous AEE and N-methyldiethanolamine solutions. Ind. Eng. Chem. Res. 2005, 44, 3720-3725. (22) Idem, R.; Wilson, M.; Tontiwachwuthikul, P.; Chakma, A.; Veawab, A.; Aroonwilas, A.; Gelowitz, D. Pilot plant studies of the CO2 capture performance of aqueous MEA and mixed MEA/MDEA solvents at the University of Regina CO2 capture technology development plant and the Boundary Dam CO2 capture demonstration. Ind. Eng. Chem. Res. 2006, 45, 2414-2420. (23) Choi, W.; Cho, K.; Lee, S.; Shim, J.; Hwang, H.; Park, S.; Oh, K. Removal of carbon dioxide by absorption into blended amines: kinetics of absorption into aqueous AMP/HMDA, AMP/MDEA, and AMP/piperazine solutions. Green Chem. 2007, 9, 594-598.