Experiment and model for the viscosity of carbonated piperazine-N-methyldiethanolamine aqueous solutions

Journal of Molecular Liquids 186 (2013) 81–84

Contents lists available at ScienceDirect

Journal of Molecular Liquids journal homepage: www.elsevier.com/locate/molliq

Experiment and model for the viscosity of carbonated piperazine-N-methyldiethanolamine aqueous solutions Dong Fu ⁎, LiGuang Qin, Huimin Hao School of Environmental Science and Engineering, North China Electric Power University, Baoding 071003, PR China

a r t i c l e

i n f o

Article history: Received 10 April 2013 Received in revised form 6 May 2013 Accepted 14 May 2013 Available online 20 June 2013 Keywords: Viscosity PZ promoted MDEA aqueous solution CO2 loading

a b s t r a c t The viscosities of both CO2-unloaded and CO2-loaded piperazine (PZ) promoted N-methyldiethanolamine (MDEA) aqueous solutions were measured by using a NDJ-1 rotational viscometer, with temperatures ranging from 293.15 to 323.15 K. The total mass fraction of amines was fixed as 0.5 and the mass fraction of PZ ranged from 0.025 to 0.15. The CO2 loading ranged from 0 to 0.6. The Weiland equation was used to correlate the viscosities and the calculated results agreed well with the experiments. The effects of temperature, mass fraction of amines and CO2 loading on the viscosity were demonstrated. © 2013 Elsevier B.V. All rights reserved.

1. Introduction In recent decades, atmospheric levels of CO2 have increased rapidly due to the utilization of great amount of fossil fuel. The reduction of CO2 emissions became a global issue [1,2]. Chemical absorption is one of the most effective approaches for CO2 capture because CO2 can be satisfactorily removed and the absorbents can be regenerated by heating. Currently, aqueous solutions of alkanolamines are widely used for the removal of CO2 from a variety of gas streams [3–10]. Among the alkanolamine series, Nmethyldiethanolamine (MDEA) has the advantages of high resistance to thermal and chemical degradation, low solution vapor pressure, and low enthalpy of absorption. However, MDEA has a low absorption rate. Adding a small amount of monoethanolamine (MEA) or piperazine (PZ) to the aqueous solution of MDEA has found widespread application in the removal of CO2 [11–20]. PZ is considered as the most promising additive to MDEA aqueous solution. The mixtures of PZ and MDEA preserve the high rate of the reaction of PZ with CO2, and the low enthalpy of the reaction of MDEA with CO2, hence leading to higher absorption rates in the absorber column, yet lower heat of regeneration in the stripper section. Solution viscosity is important in the mass transfer rate modeling of absorbers and regenerators because these properties significantly affect the liquid film coefficient for mass transfer. Viscosities of both CO2-unloaded and CO2-loaded MDEA-PZ aqueous solutions are required when designing or simulating an absorption column for CO2 absorption using MDEA-PZ aqueous solutions. So far, there are some

experiments concerning the viscosity of CO2-unloaded MDEA-PZ aqueous solutions [19,20]. However, the viscosity of CO2 loaded MDEA-PZ aqueous solutions has been rarely reported, and the influences of temperature, amine concentration and CO2 loading on the viscosity of carbonated MDEA-PZ aqueous solutions have not been well described due to the lack of experiments. The main purpose of this work is to investigate the viscosities of carbonated MDEA-PZ aqueous solutions in wide ranges of CO2 loading, temperature and amine concentration, and then demonstrate the temperature, mass fraction of amines and CO2 loading dependence of the viscosities on the basis of experiments and calculations. To this end, the viscosities of both CO2-unloaded and CO2-loaded PZ promoted MDEA aqueous solutions were measured, with the temperatures, mass fraction of PZ and CO2 loading respectively ranging from 293.15 to 323.15 K, 0.025 to 0.15 and 0 to 0.6. The Weiland equation [21] was used to correlate the viscosities. 2. Experimental section 2.1. Materials Both MDEA and PZ were purchased from Huaxin chemical Co. The sample description is shown in Table 1. They were used without further purification. Aqueous solutions of MDEA-PZ were prepared by adding doubly distilled water. The uncertainty of the electronic balance (FA1604A) is ± 0.1 mg. 2.2. Apparatus and procedure

⁎ Corresponding author. Tel.: +86 312 7522 037. E-mail address: [email protected] (D. Fu). 0167-7322/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.molliq.2013.05.027

The carbonated MDEA-PZ aqueous solutions were prepared according to the methods mentioned in the work of Weiland et al.

82

D. Fu et al. / Journal of Molecular Liquids 186 (2013) 81–84

Table 1 Sample description.

Table 2 Viscosities of CO2-unloaded and CO2-loaded MDEA-PZ aqueous solutions.

Chemical

CAS no.

Purity (in mass fraction %)

Molecular mass

Density (g·cm−3) at 293.15 K

wMDEA/wPZ

PZ MDEA

110-85-0 105-59-9

99.5 99.5

86.14 119.16

0.876 1.0377

0.475/0.025

[21] and Amundsen et al. [22]: CO2-unloaded MDEA-PZ aqueous solutions were put into a volumetric flask immersed in a thermostatic bath with a built-in stirrer for uniform temperature distribution. CO2 from a high-pressure tank was inlet into the volumetric flask at certain temperatures (CO2 pressure is atmosphere). Once the carbonated solution was prepared, varying proportions of the unloaded and loaded solutions were mixed together to produce a set of samples having fixed ratios of MDEA/PZ-to-water, but with varying CO2 loadings. CO2 loading is defined as α = nCO2/(nMDEA + nPZ), in which nCO2 is the mole of loaded CO2, and nMDEA and nPZ are respectively the moles of MDEA and PZ in the unloaded aqueous solutions. It is worth noting that CO2 loading is expected to be a major uncertainty in the experiment. In this work, the carbonated solution was prepared at non-equilibrium conditions, under which the thermodynamic equilibrium (saturated absorption, corresponding to maximum CO2 loading, αmax) was not achieved. A certain amount of CO2 escaped when the loaded solution was mixed with the unloaded solution and the atmospheric CO2 and humidity have some effects on the CO2 loading and solution concentration. The CO2 loadings of some diluted samples were checked by using the analysis method based on the precipitation of BaCO3 [21–26]. The estimated uncertainty in the CO2 loading was less than 2%. The viscosities of the carbonated MDEA-PZ aqueous solutions were measured from 293.15 to 323.15 K by using a NDJ-1 rotational viscometer produced by the Shanghai Hengping instrument factory. The measurement ranges for temperature and viscosity are respectively 273.15–383.15 K and 0.1–100 mPa·s. The uncertainty of temperature is ±0.05 K. Taking into account the uncertainty in the CO2 loading, the uncertainty of the viscosity in this work is about 2%. 3. Results and discussion The viscosities of CO2-unloaded and CO2-loaded MDEA-PZ aqueous solutions are shown in Table 2. Besides experiments, models that can correctly correlate the viscosities are also important. Among the popularly used equations [21,27,28], the Eyring equation [27] can only quantitatively describe the temperature dependence of viscosity. The Grunberg–Nissan equation [28] can well describe the temperature and amine concentration dependence, however, it is not applicable for CO2-loaded cases because the contribution of CO2 loading has not been taken into account. The Weiland equation [21] can simultaneously describe the temperature, amine concentration and CO2 loading dependences. When applied to carbonated MDEA-PZ aqueous solutions, the Weiland equation can be expressed as: ηmix ¼

w1 w2 η þ η w1 þ w2 1 w1 þ w2 2

ð1Þ

where ηmix is the viscosity of carbonated aqueous solution, and w1 and w2 respectively stand for the mass fractions of MDEA and PZ. η1 and η2 are expressed as:
 ½ðai w þ bi ÞT þ ðci w þ di Þw ηi =ηwater ¼ exp  f ðα; wÞ 2 T

ð2Þ

0.45/0.05

0.40/0.10

0.35/0.15

α

0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6

η/(mPa·s) 293.15 K

303.15 K

313.15 K

323.15 K

10.7 11.6 12.5 14.8 16.6 17.5 19.5 10.1 11.9 13.5 14.7 16.7 17.6 20.1 12.5 13.5 15.1 16.2 17.3 19.7 20.0 11.8 12.7 13.9 16.7 18.4 19.6 21.0

7.3 7.9 8.5 10.0 11.2 11.9 13.3 7.6 8.8 10.5 10.8 11.3 12.7 15.5 8.9 9.3 10.9 11.5 12.0 14.9 17.6 8.4 9.5 10.9 11.7 12.7 15.0 16.6

5.1 5.5 5.9 7.0 7.9 8.4 9.5 5.7 6.0 6.9 7.4 7.9 9.3 12.0 6.1 6.2 7.0 8.1 9.2 10.6 12.0 6.0 6.5 8.0 8.4 9.4 11.9 14.0

3.6 3.9 4.2 5.0 5.6 6.1 7.0 4.3 5.0 5.5 6.4 6.7 7.0 8.0 3.7 4.6 5.6 6.4 8.2 9.4 11.0 4.6 5.4 6.6 6.8 8.3 9.7 11.3

where ηwater is the viscosity of pure water, and w = w1 + w2 is the total mass fraction of amines. f(α, w) refers to the contribution of CO2 loading: f ðα; wÞ ¼ α ðei w þ f i T þ g i Þ þ 1

ð3Þ

where ai, bi, ci, di, ei, fi and gi are adjustable parameters. For MDEA, a1 = − 0.1944, b1 = 0.4315, c1 = 80.684 and d1 = 2889.1 were directly taken from the work of Weiland et al. [21]. The parameters for PZ were regressed by fitting to the viscosities of CO2-unloaded MDEA-PZ aqueous solutions from Derks et al. [20]. The objective function was expressed as: fs ¼

n h i X cal exp 1−η =η  100%=n

ð4Þ

i¼1

where the superscripts ‘exp’ and ‘cal’ respectively stand for the experimental and calculated data, and n is the data numbers. The optimized values are a2 = 0.1156, b2 = 8.444, c2 = − 9.074 and d2 = 3.224. The average relative deviation (ARD) is 6.2%. Figs. 1 and 2 show the viscosities of the CO2-unloaded MDEA-PZ aqueous solutions calculated from the Weiland equation, and the comparison with experiments [20]. The viscosity increases with the increase of wPZ at a given temperature and a given wMDEA, and exponentially decreases with the increase of temperature at a given wPZ and wMDEA. The Weiland equation correctly captures the amine concentration and temperature dependence of the viscosities, and satisfactorily fits the experimental data, except that some data at low temperatures are underestimated. To describe the viscosity of CO2-loaded MDEA-PZ aqueous solutions using the Weiland equation, one should firstly determine the adjustable parameters ei, fi and gi. For MDEA, e1 = 0.0106, f1 = 0 and g1 = 80.684 were directly taken from the work of Weiland et al. [21]. The parameters for PZ were regressed by fitting to the viscosities of CO2-loaded MDEA-PZ aqueous solutions from this work. The objective function is

D. Fu et al. / Journal of Molecular Liquids 186 (2013) 81–84

83

100

1

η/(mPa s)

η/(mPa s)

10

η/(mPa s)

10

0

0.00

0.02

0.04

0.06

0.08

0.10

PZ

η/(mPa s)

10

10

0.0

0.2

0.4

α

0.6

0.8

1 0.00

0.02

0.04

0.06

0.08

0.0

0.10

0.2

the same as that in Eq. (4). The optimized values are e2 = 0.7412, f2 = 0.0225 and g2 = −9.074. Fig. 3 shows the CO2 loading dependence of the viscosity of carbonated MDEA-PZ aqueous solutions. One finds from this figure that at a given temperature and a given wPZ and wMDEA, the viscosities of carbonated aqueous solutions increase monotonously with the increase of CO2 loading. The Weiland equation correctly captures the CO2 loading dependence of the viscosities and the calculated results matched the experiments satisfactorily. The corresponding ARD is 6.63%.

Fig. 4 shows the temperature dependence of the viscosity. One finds that at a given CO2 loading, and a given wPZ and wMDEA, the viscosity exponentially decreases with increasing temperature. The temperature dependence of the viscosity can also be well described by the Eyring model [27], η = κ1 exp(κ2T), despite the fact that both κ1 and κ2 are dependent on temperature. Fig. 5 shows the effect of wPZ and wMDEA on the viscosity of carbonated aqueous solutions. One finds from this figure that at a given total mass fraction of amines (wPZ + wMDEA = 0.5), no matter for CO2-loaded or CO2-unloaded cases, the viscosity increases with the increase of wPZ. It

100

η/(mPa s)

1

0 290

300

310 T/K

320

η/(mPa s)

η/(mPa s)

η/(mPa s)

10

10

1 290

0.8

Fig. 3. CO2 loading dependence of the viscosity of carbonated MDEA-PZ aqueous solutions at wPZ/wMDEA = 0.15/0.35 and wPZ/wMDEA = 0.05/0.45 (insert). ● T = 323.15 K, ○ T = 313.15 K, ■ 303.15 K, □ T = 293.15 K. Lines: calculated results.

100

100

0.6

α

PZ Fig. 1. wpz dependence of the viscosity of MDEA-PZ aqueous solutions at wMDEA = 0.477 and wMDEA = 0.119 (insert). Symbols: experimental data [20]. ● T = 293.15 K; ○ T = 298.15 K; ■ T = 303.15 K; □ T = 313.15 K; ▲ T = 323.15 K. Lines: calculated results.

0.4

330

1 290

10

1 290 300

310

320

330

10

300

310

320

330

T/K

300

310

320

330

T/K

T/K Fig. 2. Temperature dependence of the viscosity of MDEA-PZ aqueous solutions at wMDEA = 0.477 and wMDEA = 0.119 (insert). Symbols: experimental data [20]. ● wpz = 0.0; ○ wpz = 0.043; ■ wpz = 0.086. Lines: calculated results.

Fig. 4. Temperature dependence of the viscosity of carbonated MDEA-PZ aqueous solutions at wPZ/wMDEA = 0.05/0.45 and wPZ/wMDEA = 0.15/0.35 (insert). Symbols: experimental data from this work. ● α = 0.0; ○ α = 0. 3; ■ α = 0.6. Lines: —: calculated from the Weiland equation; —: calculated from the Eyring model.

84

D. Fu et al. / Journal of Molecular Liquids 186 (2013) 81–84

20

3. The Weiland equation can correctly capture the effects of CO2 loading, mass fraction of amines and temperature on the viscosities, and satisfactorily fitted the experimental data.

20

16

η/(mPa s)

16

Acknowledgments

12

The authors appreciate the financial support from the National Natural Science Foundation of China (Nos. 21276072 and 21076070), the Natural Science Funds for Distinguished Young Scholar of Hebei Province (No. B2012502076), the Fundamental Research Funds for the Central Universities (No. 13ZD16) and the Key Laboratory of Renewable Energy and Gas Hydrate, Chinese Academy of Sciences (Y007K8).

η/(mPa s)

8

12

4 290

300

310

320

330

T/K

References

8

4 0.0

0.2

0.4

α

0.6

0.8

Fig. 5. Effect of the mass fraction of amines on the viscosity of carbonated MDEA-PZ aqueous solutions at wPZ/wMDEA = 0.15/0.35 (●, —) and 0.05/0.45 (○, —). Main plot: T = 313.15 K; insert plot: α = 0.4. Symbols: experimental data from this work. Lines: calculated from the Weiland equation.

seems that for the viscosities of both CO2 loaded and CO2 unloaded MDEA-PZ aqueous solutions, the contribution from wPZ is more significant than that from wMDEA. 4. Summary In this work, the viscosities of carbonated PZ promoted MDEA aqueous solutions were measured in wide CO2 loading, temperature and amine concentration ranges. The Weiland equation was used to correlate the viscosities. Our results showed that: 1. For CO2-unloaded MDEA-PZ aqueous solutions, the viscosity increases with the increase of the wPZ at a given temperature and a given wMDEA, and exponentially decreases with the increase of temperature at a given wPZ and wMDEA; 2. For CO2-loaded MDEA-PZ aqueous solutions, the viscosity increases monotonously with the increase of CO2 loading and exponentially decreases with increasing temperature. When the total mass fraction of amines is fixed as 0.5, the viscosity increases with the increase of wPZ, indicating that for the viscosities of carbonated aqueous solutions, the contribution from wPZ is more significant than that from wMDEA;

[1] N. Nahicenovic, A. John, Energy 16 (1991) 1347–1377. [2] R. Steenevel, B. Berger, T.A. Torp, Chemical Engineering Research and Design 84 (A9) (2006) 739–763. [3] J. Knudsen, J.N. Jensen, P.-J. Vilhelmsen, O. Biede, Energy Procedia 1 (2009) 783–790. [4] L. Raynal, P.A. Bouillon, A. Gomez, P. Broutin, Chemical Engineering Journal 171 (3) (2011) 742–752. [5] T. Chakravarty, U.K. Phukan, R.H. Weiland, Chemical Engineering Progress 81 (4) (1985) 32–36. [6] A.L. Kohl, R. Nielsen, Gas Purification, 5th ed. Gulf Publishing, Houston, TX, 1997. [7] Akanksha, K.K. Pant, V.K. Srivastava, Chemical Engineering Journal 133 (1–3) (2007) 229–237. [8] J.M. Navaza, D.G. Diaz, M.D.L. Rubia, Chemical Engineering Journal 146 (2009) 184–188. [9] E. Alvarez, R. Rendo, B. Sanjurjo, M. Sanchez-Vilas, J.M. Navaza, Journal of Chemical and Engineering Data 43 (1998) 1027–1029. [10] M. Ghulam, M.S. Azmi, K.K. Lau, A.B. Mohamad, Journal of Chemical and Engineering Data 56 (2011) 2660–2663. [11] T.T. Teng, Y. Maham, L.G. Hepler, A.E. Mather, Journal of Chemical and Engineering Data 39 (1994) 290–293. [12] A. Samanta, S.S. Bandyopadhyay, Journal of Chemical and Engineering Data 51 (2006) 467–470. [13] S. Kadiwala, A.V. Rayer, A. Henni, Fluid Phase Equilibria 292 (2010) 20–28. [14] X. Zhang, J. Wang, C.F. Zhang, Industrial and Engineering Chemistry Research 42 (1) (2003) 118–122. [15] S. Bishnoi, G.T. Rochelle, Chemical Engineering Science 55 (2000) 5531–5543. [16] S. Bishnoi, G.T. Rochelle, Chemical Engineering Science 48 (2002) 2788–2799. [17] M. Ayyaz, I.A.M. Mohamed, M. Thanabalan, S. Amir, Journal of Chemical and Engineering Data 54 (2009) 2317–2321. [18] P.W.J. Derks, J.A. Hogendoorn, G.F. Versteeg, Journal of Chemical Thermodynamics 42 (2010) 151–163. [19] S. Paul, B. Mandal, Journal of Chemical and Engineering Data 51 (2006) 1808–1810. [20] P.W.J. Derks, E.S. Hamborg, J.A. Hogendoorn, J.P.M. Niederer, G.F. Versteeg, Journal of Chemical and Engineering Data 53 (2008) 1179–1185. [21] R.H. Weiland, J.C. Dingman, D.B. Cronin, G.J. Browning, Journal of Chemical and Engineering Data 43 (3) (1998) 378–382. [22] T.G. Amundsen, L.E. Oi, D.A. Eimer, Journal of Chemical and Engineering Data 54 (2009) 3096–3100. [23] D. Fu, Y.F. Xu, X.Y. Hua, Fluid Phase Equilibria 314 (2012) 121–127. [24] D. Fu, L.H. Chen, L.G. Qin, Fluid Phase Equilibria 319 (2012) 42–47. [25] D. Fu, H.M. Hao, L.G. Qin, Journal of Molecular Liquids 181 (2013) 105–109. [26] D. Fu, L. Wei, S.T. Liu, Fluid Phase Equilibria 337 (15) (2013) 83–88. [27] S. Glasstone, K.J. Laidler, H. Eyring, The Theory of Rate Process, McGraw-Hill, New York, 1941. [28] L. Grunberg, A.H. Nissan, Nature 164 (1949) 799–800.