ASTM D1015 Standard Test Method for Freezing Points of High-Purity Hydrocarbons
9. General Procedure for Determining a Freezing Curve
9.1 Assemble the apparatus, with no refrigerant and no sample yet in place, but with a stream of air, freed of carbon dioxide and water, flowing at a rate of 10 to 20 mL/min. Fill the jacket of the freezing tube with air freed of carbon dioxide and water.
9.2 As required, the operator must be prepared to induce crystallization in the sample as soon as possible after the temperature has passed below the freezing point of the sample (to prevent excessive undercooling). In some cases, crystallization may be induced by introducing into the sample at the appropriate time a small rod (ABC in Fig. 4) which has been kept at an appropriate lower temperature (near 0°C, -80°C, or -180°C) (J in Fig. 4). In other cases, crystallization can be induced by introducing into the sample at the appropriate time crystals of the sample on the coiled end of the small rod (ABC in Fig. 4). When inducing crystallization, the cold rod (with or without crystals) should be immersed in the sample in the freezing tube for about 2 s (if necessary, this is repeated every 2 or 3 min). These crystals are made by placing several millilitres of the sample in a small test tube, incased in a thin metal tube, as shown at E in Fig. 4, immersed in a refrigerant whose temperature is below the freezing point of the sample. A slurry or mush of liquid and crystals is produced. The rod (ABC in Fig. 4), with wet crystals adhering to the helical coil C, is raised above the liquid level in the tube E and held in position with a cork stopper until required for seeding.
9.3 Fill the Dewar flask surrounding the freezing tube with the appropriate refrigerant. Temporarily remove the thermometer and stopper and then introduce the sample (usually 50 mL of liquid in amount) through a pipet if the material is normally liquid, or by pouring the refrigerated liquid sample through the tapered male outlet of the reservoir trap (E in Fig. 1 of Test Method D1016) if the material is normally gaseous. When specified in Test Method D 1016, filter the sample directly into a freezing tube (O in Fig. 1) through silica gel to remove water. A detailed drawing of a funnel used for this purpose is shown in Fig. 5. Each time a freezing or melting curve is determined after the sample is melted, it is necessary to remove the sample from the freezing tube and refilter it through silica gel into a dry freezing tube to remove water. When the sample is volatile or normally gaseous at room temperature, cool the freezing tube before introduction of the sample in order to minimize loss by evaporation. Continue the flow of air (freed of carbon dioxide and water) into the freezing tube in order to keep out water vapor. Start the stirrer and allow the sample to cool down to within about 15°C of the freezing point, then begin evacuation of the jacket of the freezing tube.
9.4 Observe the time and the resistance of the thermometer at even intervals of 0.02 to 0.05 Ω (about 0.2 to 0.5°C) to determine the rate of cooling, which is continually changing as the pressure in the jacket of the freezing tube is reduced. Care must be taken to close the stopcock to the freezing tube when the desired cooling rate is obtained. In case the cooling rate is allowed to become too slow, the pressure and likewise the cooling rate can be increased by bleeding in air (freed of carbon dioxide and water) through stopcocks P' and P (Fig. 1). When a cooling rate is obtained that will give a change of 1°C in about 1 to 3 min in the range of about 5 to 10°C above the freezing point, close the stopcock controlling the jacket of the freezing tube. (The optimum rate of cooling will vary with the material being examined.)
9.5 When the temperature reaches a point about 5°C above the expected freezing point, record the time to 1 s (or 0.01 min) at which the resistance of the thermometer equals 0.1 or 0.05 Ω. At the appropriate time (see 9.2) induce crystallization. The beginning of crystallization will be accompanied by a halt in the cooling of the liquid. After recovery from undercooling is substantially complete, record the resistances at intervals of about 1 min. If a galvanometer is being used, also record the galvanometer scale at full sensitivity and with no current through the galvanometer. These observations, together with the sensitivity of the galvanometer system in terms of ohms per millimetre of scale reading, yield a sensitivity of nearly 0.0001°C. Approximately equal sensitivity is obtained when using a microvolt ammeter. Continue observations until the stirrer begins to labor, then stop the stirrer. After several minutes (when a steady rate is obtained) make alternate N and R readings through the commutator at fixed intervals of about 1 min. Determine the difference between the two at any given time from a plot of the values against time.
10. General Procedure for Determining a Melting Curve
10.1 For determining a melting curve, proceed exactly as described in Section 9 for a freezing curve, up to the point where the stirrer begins laboring. When the stirrer shows signs of laboring, make a comparison of N and R readings through the commutator, as in 9.5 except that the stirrer is still operating. When the laboring of the stirrer becomes quite pronounced, the freezing curve (with the stirrer still operating) is changed to a melting curve. The energy for melting is supplied in either of the two following ways: (1) the cooling bath is replaced by a warming bath and simultaneously the jacket is evacuated for an appropriate length of time (3 to 10 min). The stopcock on the freezing tube is closed; or (2) the cooling bath is left in position or replaced by a warming bath and the jacket evacuated as much as possible, leaving the stopcock to the freezing tube open to the vacuum system during the entire melting curve. In this case, the thermal conductivity across the jacket is so small that the energy introduced by the stirrer provides the energy for melting. Continue the observations of time and resistance along the equilibrium portion of the melting curve as along the equilibrium portion of the freezing curve. When melting is substantially complete, as evidenced by a marked change in the rate of change of resistance, make observations of time at even intervals of 0.05 Ω (0.5°C). The experiment is concluded when the temperature has gone about 5 to 10°C above the freezing point.
11. Evaluation of the Freezing Point from a Freezing Curve 11.1 To locate zero time (the time at which crystallization would have begun in the absence of undercooling), make a preliminary plot of the time-resistance observations covering the liquid cooling line and the equilibrium portion of the freezing curve. For this plot, as shown in Fig. 6, the time scale is taken so that 10 mm is equivalent to 1 min and the resistance scale (for a 25-Ω thermometer) is taken so that 10 mm is equivalent to 0.02 Ω (0.2°C). Zero time is determined by a visual extrapolation, on this plot, of the equilibrium portion of the freezing curve back to its intersection with the liquid cooling line.
11.2 In order to locate accurately the resistance corresponding to the freezing point, plot the time-resistance observations as shown in with the time scale as before but with the scale of temperature magnified 10 to 200 times. The equilibrium portion of the curve, GHI, is extended back to its intersection at F with the liquid line by the simple geometrical construction shown in Fig. 8, selecting for this purpose three points (near the ends and the middle) of the equilibrium portion of the curve (Note 3). The point F gives the resistance corresponding to the freezing point.
NOTE 3 - The location of the resistance corresponding to the freezing point can be made using algebraic expressions derived from the geometrical construction. These are as follows:
Rf = Rg + [(Rg - Ri)/(uvw - 1)
where:
u = [(Rh - Ri)/(Rg - Rh)],
v = [(Zh - Zg)/(Zi - Zh)],
and
w = [(Zi - Zf)/(Zg - Zf)]
Zf, Zg, Zh and Zi are the times corresponding to the points F, G, H, and I, respectively, and Rf, Rg, Rh and Ri are the resistances in ohms corresponding to the points F, G, H, and I, respectively.
It is nearly always possible to select the point H equidistant in time between G and I, so the v = 1.
11.3 The observed resistance at the point F, corrected by one half the difference between the N and R readings, and by a bridge zero correction, appropriate calibration corrections to the coils of the bridge, and by an ice point correction, if necessary, is converted to temperature in degrees Celsius. (See Fig. 7.)
12. Evaluation of the Freezing Point from a Melting Curve
12.1 Determine zero time from a preliminary plot (as for the freezing curve (Section 11)) of the time-resistance observations covering the equilibrium portion of the melting curve and the liquid warming line, as shown in Fig. 9. Zero time can usually be determined by visual extrapolation, on this plot, of the equilibrium portion of the melting curve to its intersection with the liquid warming line extended down in temperature to its intersection with the extension of the equilibrium portion of the curve.
12.2 The location of the freezing point at F is done exactly as in the case of the freezing curve, except that the geometrical extrapolation is made to the right as shown in Fig. 10. (See Fig. 8 and the reference in Footnote 10 for details.)
12.3 Make the conversion of resistance to temperature as described in Section 11.
