Identification and Characterization of Nonflammable Azeotropic Mixtures (Chapter 3)

CHAPTER III Results This chapter contains data that was collected using the methods outlined in the Materials and Methods chapter. This data is discussed in Chapter IV. All uncertainties shown arise from regression analyses, and represent one standard deviation. Table 3 shows the composition, boiling points, and flash points of all the azeotropic mixtures discovered in this study. Table 3. Compositions, boiling points (Tb), and flash points (Tf) of new azeotropic or near azeotropic mixtures. Label Component 1 %(v/v) Component 2 %(v/v) Tb / °C AZ1 Tf / °C 85% HFE1 15% MOH 46.6 45.5 AZ2 85% AE3 15% MOH 48.7 48 AZ3 53% HFE3 47% BAC 84 ≤ 21 AZ4 50% AE3 50% HFC3 N/R None AZ5 83% HFE5 17% EL 112 ≤ 50 AZ6 80% Sup 20% CBTF 110 None AZ7 65% HFE5 35% CBTF 125 None Note: “None” under the flash point column means that the substance was brought to boiling and still did not undergo a flash. “N/R” means not recorded. In Table 3, the measured boiling points were obtained by recording the temperature reading on the thermometer of the distillation apparatus after the first drop of distillate was collected and the temperature stabilized. The measured flash points were determined using a closed-cup flash point tester and a modified ASTM D56 method. The volume percentages were obtained using a Raman chemometric technique. The HFE-7500 / ethyl lactate azeotrope was a two-phase liquid where the light fraction (LF) made up 14% of the total volume and consisted of 5% HFE-7500 and 95% ethyl lactate. The heavy fraction (HF) of this mixture made up 86% of the total volume and consisted of 96% HFE-7500 and 4% ethyl lactate. For details about the methods used to obtain this data, refer to chapter II. A deviation from the method was made for the two-phase azeotrope, and this is discussed later in this chapter. The spectral analyses for all the azeotropes are shown in the following figures (Figure 14-Figure 21). The spectra in these figures are vertically offset from each other to allow each spectrum to be seen clearly. Figure 14. The Raman chemometric analysis of the azeotrope between AE3000 and methanol. Figure 15. The Raman chemometric analysis of the azeotrope between HFE-7100 and methanol. Figure 16. The Raman chemometric analysis of the azeotrope between HFE-7300 and tert-butyl acetate. Figure 17. The Raman chemometric analysis of the azeotrope between AE3000 and Solkane 365. AZ5 consisted of two immiscible layers that were each a mixture of the two components. This led to an interesting way of calculating the total volume composition. The azeotrope was placed in a graduated cylinder, covered with parafilm, and left to settle for a few days. At this point, the fraction of the total volume that each layer made up was calculated, and a sample of each layer was removed from the graduated cylinder with a pipette and analyzed via Raman spectroscopy. The results of these analyses are shown in Figure 18 and Figure 19. Knowing the composition of each layer and the fraction of the total volume that each layer made up enabled the calculation of the overall volume composition. Figure 18. The Raman chemometric analysis of the top layer of the azeotrope between HFE-7500 and ethyl lactate. Figure 19. The Raman chemometric analysis of the bottom layer of the azeotrope between HFE-7500 and ethyl lactate. Figure 20. The Raman chemometric analysis of the azeotrope between Suprion and pchlorobenzotrifluoride. Figure 21. The Raman chemometric analysis of the azeotrope between HFE-7500 and pchlorobenzotrifluoride. The volume percent verification data for this chemometric analysis is shown in Table 4. Table 4. Volume percent composition verification data. Chemometric % Composition Composition type CBTF HFE5 CBTF HFE5 Prepared Volume % 50 50 50 50 Computed Mole % 61 39 Computed Mass % 41 59 Prepared Volume % 40 60 39 61 Computed Mole % 51 49 Computed Mass % 32 68 Note: The numbers on the left are the compositions that were mixed, and the numbers on the right are the chemometric analysis results. The solvency screening test results can be seen in Figure 22. Figure 22. Solvency screening test results. These results were used to decide which azeotropes to characterize further. The azeotropes that passed the flash point and solvency screening, and the pure components of those azeotropes were characterized in terms of the thermal expansion coefficient (β), density (ρ), surface tension (γ), and viscosity (η). The pure substances were characterized solely for the purpose of testing the accuracy of the methods, and literature values from the SDSs are provided for comparison. These SDSs can be found in the SERDP report.14 The physical properties for the azeotropes and pure substances can be seen in Table 5. Table 5. Physical properties of select substances compared to literature at 25 °C. Substance β/ 10-3 °C-1 Exp. ρ/ g/mL Lit. ρ/ g/mL Exp. γ/ mN/m Lit. γ/ mN/m Exp. η/ cP Lit. η/ cP HFE5 1.55(2) 1.62(2) 1.614 - 16.2 1.08(3) 1.24 sup 1.4(1.1) 1.57(6) 1.59 - 18 1.06(18) 1.12 CBTF 1.03(4) 1.333(3) 1.34 - 25 0.84(7) 0.79 AZ6 1.47(2) 1.521(2) None 18.8(1) None 0.97(4) None AZ7 1.58(5) 1.516(6) None 19.3(2) None 0.94(5) None Note: The numbers in parentheses are the calculated uncertainties for the experimentally determined values. The thermal expansion coefficients (β) were found using the dependence of density on temperature. The determined HSPs and wetting index for AZ6, AZ7, and their components are shown in Table 6. Table 6. Hansen Solubility Parameters and wetting indices of select substances at 25 °C. Substance D/ (J/mL1/2) P/ (J/mL1/2) H/ (J/mL1/2) W HFE5 13.3 2.0 1.0 92.6(2) sup 12.8 2.0 1.3 82.2(8) CBTF 17.3 4.0 2.9 63.6(1) AZ6 13.7 2.4 1.6 83.9(1) AZ7 14.7 2.7 1.7 83.6(1) Note: The numbers in parentheses are the calculated uncertainties. In Table 6, the HSPs for the three pure substances are literature values used to calculate the HSPs of the two azeotropes. The wetting index values were calculated from the experimental values of density, surface tension, and viscosity. However, the literature surface tension values were used for the pure components. The results of the azeotrope prediction model are shown in Table 7. Table 7. Comparison of predicted azeotrope compositions to experimental compositions. Label Components (1/2) Component 1 Exp (v/v) Component 2 Exp (v/v) Component 1 Pred (v/v) Component 2 Pred (v/v) Error (v/v) AZ1 HFE1/MOH 0.85 0.15 0.76 0.24 -0.09 AZ2 AE3/MOH 0.85 0.15 0.80 0.20 -0.05 AZ3 HFE3/tBAc 0.53 0.47 0.68 0.32 0.15 AZ4 AE3/HFC3 0.50 0.50 0.35 0.65 -0.15 AZ5 HFE5/EL 0.83 0.17 0.75 0.25 -0.08 AZ6 Sup/CBTF 0.80 0.20 0.77 0.23 -0.03 AZ7 HFE5/CBTF 0.65 0.35 0.75 0.25 0.10 Note: The error is calculated for component A. Negative error in Table 7 indicates that the predicted composition of component A is lower than the experimental composition. Positive error indicates the opposite. Since the compositions that are shown in Table 7 are the volume fractions where every number is less than one and greater than zero, error is more appropriate than percent error.