Supplementary Vapo Pressue Data JChT,53,2012,114-118

8/12/2019 Supplementary Vapo Pressue Data JChT,53,2012,114-118

1/5

Supplementary vapor pressure data of the glycol ethers, 1-methoxy-2-propanol,

and 2-methoxyethanol at a pressure range of (15 to 177) kPa

Arturo Bejarano a, Laura J. Poveda b, Juan C. de la Fuente a,c,

a Departamento de Ingeniera Qumica y Ambiental, Universidad Tcnica Federico Santa Mara, Avda. Espaa 1680, Valparaso, Chileb Pontificia Universidad Catlica (PUC) de Chile, Avda. Vicua Mackenna 4860, Macul, Santiago, Chilec Centro Regional de Estudios en Alimentos Saludables CREAS, Blanco 1623, Valparaso, Chile

a r t i c l e i n f o

Article history:

Received 18 January 2012Received in revised form 12 April 2012Accepted 27 April 2012Available online 8 May 2012

Keywords:

Dynamic recirculation methodGlycol ethers1-Methoxy-2-propanol2-MethoxyethanolVapor pressure

a b s t r a c t

The vapor pressure of pure 1-methoxy-2-propanol and 2-methoxyethanol, commonly used as co-solventsin inks, paints, coatings, organic/water solutions among many other applications, were measured with adynamic recirculation apparatus at a pressure range of (15 to 177) kPa. The measurements were per-formed at temperature ranges of (342 to 412) K for 1-methoxy-2-propanol and (346 to 417) K for 2-methoxyethanol. The maximum likelihood method was used to estimate the parameters of the Antoineequation, the parameters of an extended Antoine equation and the Wagner equation were determined bynon linear least squares method. The three models showed root mean square deviations ( rmsd) of 0.39%,0.38%, and 0.29%, and 0.37%, 0.33%, and 0.32%, for 1-methoxy-2-propanol and 2-methoxyethanol, respec-tively. Additionally, the experimental data and correlation were compared with those available in theliterature.

2012 Elsevier Ltd. All rights reserved.

1. Introduction

Common glycol ethers are oxygenated hydrocarbons of majorindustrial and economic importance. Since they have both func-tional hydrophobic and hydrophilic groups they are suitable for alarge number of industrial and commercial applications includinghousehold products, paints, inks, coatings, cleaning solutions andbiochemical applications[1]. Nonetheless, several studies relatedto toxicity suggest that exposure to glycol ethers (1-methoxy-2-propanol and 2-methoxyethanol, among others) can cause adverseeffects on human health [25]. Vapor pressure of pure glycol ethersis a relevant property on which the vapor-liquid calculations havea strong dependence and therefore it is of great importance in thedesign of separation processes. Moreover, several derived physical-

chemical properties can be estimated from the vapor pressuredata. Besides, limited information regarding vapor pressure datawere reported in the literature for the glycol ethers 1-methoxy-2-propanol and 2-methoxyethanol.

The main objective of this work was to contribute with newexperimental information of the vapor pressures for the glycolethers, 1-methoxy-2-propanol and 2-methoxyethanol, in the rangeof (15 to 177) kPa by measuring the isobaric (vapor + liquid)

equilibrium. Additionally, the parameters of three commonly usedvapor pressure equations were estimated: the Antoine model, theWagner model[6], and an extended Antoine model[7].

2. Experimental

2.1. Materials

Pure nitrogen (N2) was supplied by AGA Chile with no less than99.999% of N2. HPLC-grade 2-methoxyethanol and pro-analysis1-methoxy-2-propanol were obtained from Aldrich (St. Louis,MO) with purity greater than 99.5%. These materials were usedwithout further purification.

2.2. Apparatus and procedures

The vapor pressure was measured using a commercial all-glassdynamic recirculation isobaric (vapor + liquid) equilibrium (VLE)apparatus (Labodest model 602D, i-Fischer Engineering GmbH,Waldbttelbrunn, Germany)[8]. Its operation procedure relies onthe principle of the recirculation of both liquid and vapor phasesat controlled pressure. The advantage of the recirculation methodis the rapid appearance of the equilibrium simultaneously withthe exact measurement of the boiling temperature. The experi-mental uncertainty was the uncertainty associated to the equip-ment which was estimated


2/5

in our equipment and those reported in literature. The comparison,the apparatus (figure 1) and methodology are described in moredetail in our previous work[9]. The experimental procedure usedwas as follows: the pure compound was charged in the apparatusby the filling tunnel (13), the N2 supply (7) to the system wasopened, once the liquid is at the desired level the magnetic stirrerbar was activated (3), the pressure throttle valve (9) is opened, andthe vacuum pump (10) was started to work. The desired value ofpressure was set on the controller panel, the immersion heater(2) was activated and finally, fine adjustments of pressure weremade by manual operation of valve (8). In order to verify that sys-tem reached the equilibrium, the temperature stability had to re-main constant (within 0.1 K) for a period of time of (15 to30) min[8].

3. Results and discussion

Experimental values of temperature and pressures measuredfor 1-methoxy-2-propanol at pressure range of (15 to 177) kPaand temperature range of (342 to 412) K and, for 2-methoxyetha-nol at the same pressure range and temperature range of (346 to417) K are listed intable 1. The regressed parameters of three va-por pressure equations, described below, along with their rootmean square deviation (rmsd) are reported intable 2.

The maximum likelihood method was used to estimate theparameters of the Antoine equation(1) in order to take into ac-count its non linear mathematical form and the fact that both tem-perature and pressure are subject to experimental variability[10].

lnfpcal=kPag A B

T=K C 1

where A, B and C are adjustable parameters. The maximumlikelihood objective function to be minimized has the form:

SXi

TTcalrT;i

2

ppcalrp;i

2" # 2

where rT,i and rp,i are estimated standard deviations in the mea-sured temperature and pressure for theith observation. These val-ues were assigned from the experimental set up as rT,i= 0.1 K andrp,i= 0.15 kPa.

The Wagner equation (3) [7] with four different functionalforms, was evaluated to represent the measured vapor pressures.

FIGURE 1. Experimental apparatus: (1), Cotrell pump; (2), immersion heater; (3), mixing chamber; (4), vapor Pt-100 temperature probe; (5) pressure controller; (6), vacuumpump; (7), N2supply; (8), vacuum throttle valve; (9), pressure throttle valve; (10), vacuum by-pass; (11), thermo regulated bath; (12), vapor condensers; (13), filling tunnel;(14) and (15), liquid and vapor samplers; (16), overpressure relief valve; (17), vacuum relief valve; (18), 3/2 way valve.

TABLE 1

Experimental vapor pressures (p) and percent deviations (102(ppcal)/p) from the

Wagner equation (3) for 1-methoxy-2-propanol and 2-methoxyethanol at temper-

aturesT.

1-Methoxy-2-propanol 2-Met hoxyet hanol

T/K p/kPa 102(p pcal)/p T/K p/kPa 102(p pcal)/p

342.3 15.0

0.25 346.4 15.0 0.06352.5 23.5 0.38 356.9 23.5 0.04360.1 32.0 0.33 364.7 32.0 0.29367.4 42.1 0.24 376.0 49.0 0.07371.6 49.1 0.21 380.6 57.5 0.22376.0 57.6 0.08 384.5 66.0 0.16380.0 66.1 0.09 388.1 74.5 0.19383.5 74.6 0.26 391.3 83.0 0.50386.9 83.3 0.04 394.4 91.5 0.28390.1 92.6 0.27 397.2 100.0 0.34392.8 99.9 0.68 400.1 108.5 0.49397.9 116.9 0.68 402.6 117.0 0.56400.0 125.5 0.04 404.9 125.5 0.45402.2 134.0 0.07 407.0 134.0 0.11404.3 142.5 0.08 409.1 142.5 0.09406.3 151.0 0.11 411.1 151.0 0.07408.2 159.5 0.13 412.9 159.5 0.27410.0 168.0 0.23 414.7 168.0 0.39

411.8 176.5 0.16 416.6 176.5 0.02

A. Bejarano et al. / J. Chem. Thermodynamics 53 (2012) 114118 115


3/5

The results for the regression analysis have shown that there werenot significant differences, in terms ofrmsd, among the four func-tional forms selected for both glycol ethers. However, the func-tional form depicted by equation (3) showed the bestrepresentation for the vapor pressures measured in this work.

lnpcal=pc fTc=T=Kg c1sc1:5s1:5 c2s

2 c4:5s4:5

3

wherepcalis the calculated vapor pressure, Tcandpcare the criticaltemperature and pressure,ciare the fitting model parameters, and,s= 1 T/Tcis a reversed reduced temperature variable.

The critical temperature and pressure in equation (3) were579.8 K and 4113 kPa for 1-methoxy-2-propanol, and 597.6 K and5285 kPa for 2-methoxyethanol[11].

The extended Antoine equation(4)[6]has the form:

lnfpcal=kPag A B

T=KC fT=Kg D lnfT=Kg E fT=Kg6 4

where A, B, C, D and Eare adjustable parameters. Non linear leastsquares method was used to estimate the parameters of the equa-

tions(3) and (4)by minimizing the objective function

S

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1

N

XNi

ppcalp

2vuut 5

The Wagner equation (3) was selected to compare the model re-sults with information found in the literature. Table 3 lists the datasets selected from the literature for the vapor pressures of bothcompounds studied in this work. Table 1 also lists the percent devi-ations from equation(3)for the vapor pressures measured for 1-

methoxy-2-propanol and 2-methoxyethanol.The relation between natural logarithm of the vapor pressure

and the inverse of temperature for the experimental data and thecalculated pressure from Wagner equation(3)is shown infigures2 and 3for 1-methoxy-2-propanol and 2-methoxyethanol, respec-tively.Figure 4 compares for 1-methoxy-2-propanol the percentdeviations from equation(3)with the data sets included in table3in the temperature range of (340 to 420) K. Experimental vaporpressure data from Psutka and Wichterle [12]were representedusing the equation(3)with a standard deviation of 0.22 kPa, andpercent deviations in the range of (3.6% to 1.2%), the minimumdeviation was shown at temperatures around 390 K. All percentdeviations were above the values measured in this work and theyincrease as temperature decreases. This result showed the highest

difference between this work and the selected data sets oftable 3for 1-methoxy-2-propanol. Chiavone-Filho et al. [13]reported va-por pressure measurements in a narrower temperature range, nev-ertheless, they were in very good agreement with this work results,equation(3)represented the values of this data set with standard

TABLE 2

Estimated parameters (A,B,C,D,E, and ci) and root mean square deviation (rmsd) from the Antoine, Wagner and extended Antoine equations for 1-methoxy-2-propanol and 2-

methoxyethanol.

Equation Parameter Deviation

A B C D E 102rmsd

1-Methoxy-2-propanol

Antoine(1) 14,557 3145,549 76,530 0.39Wagner(3) c1 c1.5 c2.5 c5 0.29

8.115 1.867 3.508 8.774Ext. antoine(4) 178.690 5.643 0.056 1.639 5.457 0.38

2-Methoxyethanol

Antoine(1) 14,696 3256,010 74,619 0.37Wagner(3) c1 c1.5 c2.5 c5 0.32

8.952 3.446 4.258 6.655Ext. antoine(4) 175.903 5.558 0.054 3.648 5.013 0.33

TABLE 3

Vapor pressure datasets for 1-methoxy-2-propanol and 2-methoxyethanol selected

from the literature.

Type of information Temperaturerange/K

Reference

1-Methoxy-2-propanol

Experimental:Tp318 to 392 Psutka and Wichterle (2004)[12]331 to 389 Chiavone-Filho

et al.[13]

332 t o 373 Ant osiket al. (2004)[14]342 t o 412 This work348 t o 378 Chylinskiet al. [15]

2-Methoxyethanol

Experimental:Tp315 to 363 Hauschild and Knapp[16]336 t o 397 Lladosaet al.[17]346 t o 417 This work350 to 397 Marrufoet al.[18]

Correlation329 to 397 NIST Chemistry WebBook[19],

parameters obtainedfrom Picket al.[21]

329 t o 397 Chandaket al. [20]343 to 363 Chiavone-Filhoet al. [13],

parameters obtained

from Boublket al. [22]

FIGURE 2. Natural logarithm of vapor pressure (lnp)vs. the inverse of temperature

(T1

) for the experimental data () and the calculated pressure ( ), from Wagnerequation(3)for 1-methoxy-2-propanol.

116 A. Bejarano et al. / J. Chem. Thermodynamics 53 (2012) 114118


4/5

deviation of 0.12 kPa, and the percent deviations were in the rangeof (0.08% to 0.48%), again the minimum deviation was aroundtemperatures of 390 K. For this data set a slight convex tendencyof the percent deviations was observed. The data set of Antosiket al.[14]was represented by equation (3)with standard deviationof 0.09 kPa and maximum percent deviations of 0.98%. It can beseen from figure 4that the data set[14]indicates a minor decreas-ing trend of the percent deviations as temperature increases, theminimum percent deviation took place at temperatures near to363 K and was 0.01%. A similar tendency was displayed by the per-cent deviations from the vapor pressures reported by Chylinskiet al.[15]. For this data set, the standard deviation from equation(3)was 0.07 kPa and the percent deviations were in the range of(1.0% to 0.03%), which was in very good agreement with this workresults as well as for data set[14]. The observed discrepancies be-tween the data sets [1315] and this work, and the values reportedby Psutka and Wichterle[12]were unexpected because the mea-

surements made by these authors were relatively recent compared

to those of Chiavone-Filhoet al.[13](1993), and were made withthe same DvorakBoublik type of equipment. In addition, data sets[14]and[15]were from the same year of data set[12],they werebased on the same principle of measurement (dynamic ebulliome-try), and they did showed good agreement with this work andamong themselves and the data set[13].

Figure 5 shows for 2-methoxyethanol the comparison of the va-por pressure, in the temperature range of (340 to 420) K, amongthe data selected from the literature included intable 3and theequation(3). According tofigure 3, for the references; Hauschildand Knapp[16], NIST[19], Chandaket al.[20], and Chiavone-Filhoet al.[13], there was a region where equation (3) represented thesedata sets with comparatively large positive percent deviations,especially at low temperatures, from(4.5% to 0.6%). Chiavone-Filhoet al.[13]estimated the A andB parameters of the Antoine equa-tion, from measurements made at temperatures (343 and 363) K,with a fixed parameter Ctaken fromBoublk et al.[22]. Apparently,the work of Boublk et al. [22]is a reviewed version of the collec-tion made by Pick et al. [21], which is, as well, the reference forthe parameters of the Antoine equation reported by NIST [19].Hence, this information is very similar to each other. As it can beseen fromfigure 3, the data sets[13,16,19,20]also showed signif-icant differences with the measurements of Lladosaet al.[17], andMarrufo et al.[18]mainly at low temperatures. Another region canbe described infigure 3, between temperatures of (375 to 395) K,the percent deviations from equation (3) for the data sets [1720] and this work results becomes smaller. In this region, thecorrelations from NIST[19]and Chandak et al. [20]were in goodagreement with the last values measured by Lladosa et al. [17]and some of the values reported by Marrufo et al.[18]. The devia-tions from this work results, in the same region, were


5/5

4. Conclusions

New and complementary information for the vapor pressure ofpure 1-methoxy-2-propanol, and 2-methoxyethanol was pre-sented in this work for the temperature ranges of (342 to 412) Kand (346 to 417) K, respectively. The measurements were carriedout using a VLE apparatus with an uncertainty

Documents

Supplementary Vapo Pressue Data JChT,53,2012,114-118