- Email: [email protected]
Apparatus for the determination of melting points, molecular weights, freezing points and purity, and for the study of melting
Talmta, 1970.Vol. 17,pp. 999lo 1003. Pcr~amon Prear.Rintedin NorthernIreland
SHORT COMMUNICATIONS
Apparatus for the determination of melting points, molecular weights, freezing points and purity, and for the study of melting (Received10 June 1969. Revised 4 May 1970. Accepted 12 May 1970) AN APPARATUS for the rapid precise determination of molecular weights has been developed. It is based on the measurement of melting point lowering, because there are difficulties in obtaining accurate values for freezing points. *sl Notice is taken of the conditions which at&t precision and system. An equation is deduced for accurate calculation arise from processes occumnginthemel’ of the melting point, from an estimate Yo the time at which the melting point occurred, and the co-ordinates of two points in the melting region of the time-temperature record. EXPERIMENTAL The vessel was a lo-ml Pyrex glass bulb with two ground-glass necks, the one for filling and em tying having a ground-glass stopper, and the other containing a sealed-in thermistor armnged to Kave its tip immersed 5 mm into the liquid. The vessel contained 7 g of liquid and five glass balls 6 mm in diameter. The thermistor (FS 23), Standard Telephones and Cables Ltd.) had a resistance of 2000 haat 20”, and tem ture coetiicient 132 n/K, and was made one of the ratio arms of a Wheatstone bridge, the out-o p” -balance current being amplified and recorded on the chart of a recording potentiometer adjusted to give full-scale deflection for the concentration range being used. The flask was aled with a known weight of solvent and immersed in a freezing-bath for 1 min. It was then placed in an insulated container mounted on the arm of a reciprocating machine and swung uniformly, 90 times a minute, through an arc of 120”. The temperatur+time trace of the pure solvent formed a base line from which subsequent meltingpoint lowerings could be measured. Some examples are shown in Fig. 1. Conditions to be observed in design
6. 7. 8. 9.
Crystallization of the solvent should be rapid. Residual supercooling in the solid/liquid system or in the temperature measuring device is undesirable. Rapid heat exchange between solid and liquid is necessary so that equilibrium between the two phases can be established quickly. Melting should proceed through a series of equilibrium states, each state, for partially frozen solutions, being in accordance with the freezing point law. -sensing device must detect small changes accurately. The temperature The tem measureZera ._ _.should .-be that of the solution in equilibrium with frozen solid. At themeltingpotnt theconcentration of the solution in equilibrium with solid must be constant. The melting int should be relatively indifferent to the rate of melting. A clearly def!Zed melting point is desirable. The apparatus should be simple to use, and the time of measurement short.
DISCUSSION Crystallization of the organic liquids used, of water and of molten naphthalene and diphenylamme was rapid and clearly de&xl, largely because of the presence of the glass balls. The melting point was well defined on the recorder trace and by calculation could be measured accurately. The time required to remelt the partially frozen liquid was short, usually about 5 min. The thermistor tip was small and hence the temperature response rapid. Efficiency of mixing was the main factor involved in satisfying the other conditions, and was achieved by means of the glass balls, the umformity of the mechanical motion and the jerk at the end of each stroke. The uniformity of motion was most important. The efRciency was demonstrated by the smoothness of the trace for pure solvent and by the shape of that for a solution. The shape of the curve follows from the freezing point depression law, which may be ezpressed as
1000
Temperature FIG.
1 .-Examples
of time-temperature
f--
curves;
(a) pure solvent;
(b) solution.
where k = the cryoscopic constant, n = moles of solute, wO= weight of solvent at beginning of melting, T, = melting point of solvent, s = speed of melting, tI = time of melting to temperature TX. For any given trace for melting, the only variables are T and t and when plotted against each other they should give a concave curve. The shape of the trace in Fig. 1 (b) illustrates this. That all such traces were curved confirms the efficiency of mixing. The rate of heat exchange between the two phases was not only a function of mixing, but also of crystal size and the thermal properties of the two phases. Thus the heat exchange was least for pure solvents where the crystals were large, and greatest for the solvent cyclohexane, the latent heat of which was so low that melting was too rapid for a trace like Fig. 1 to be developed. Detemination of the melt@ point
The melting point of solvents (TL) is that portion of the trace which gives a constant temperature over an interval of time. The melting point for solutions can be taken as that temperature where the amount of frozen solid becomes vanishingly small (ZY,). At this point, T&-
where w is the weighed amount of solvent.
TV =
ATE =t
Short communications
1001
The expression used to calculate the melting point is ATIATOa ATs = AT,,(a + 1) - AT,
(3)
TO are two temperature differences on the trace, correwhereAT,=Tc-TlandAT,=TLsponding to times of melting I1 and Q a = (tl - t,,)/(r, - ra, and where te is the time at Te (see Fig. 1). From equations (1) and (2) we obtain ATew wo+stl=x’ andas w,,AT, = kn = ATew, then ATew ATew F + stx = m. 0
(5)
Thus,
If the rate of melting is the same throughout the melting process, and it is probable that this is so over a small temperature range and relatively short melting times, then the expression above is true whatever the temperature points chosen. Hence,
AT,AT,a ATe = AT,,(a + 1) - AT, ’ Estimation of precision The precision of cryosopic results is usually expressed in terms of temperature as this is the unit of measurement. It was thought that concentmtion would be a more meaningful term. The cryoscopic constant k estimatedfrom the slope of a plot of melting point depression VS.molal concentration m, may be used to estimate the concentration m’ corresponding to any given value of ATe, m’ = AT.Jk. The difIerences (m - m’) can than be used as a measure of precision. Results are shown in Table I. APPLICATIONS Molecular weights The precision with which the molecular weights of organic compounds can be measured by the apparatus can be seen from Table 1. The mean value of the standard deviations is &to402 mole/kg and the experimental value for a substance of molecular weight of 1000 in 1-molal solution would be found to lie between 1004 and 9%. If, however, the solution were 0.1 molal, the value would lie between 1040 and 960. It follows that within the bounds of ideality. the more concentrated the solutions the smaller the error. The samples in Table I included solutions up to l-2 molal. For a molal solution of a substance of molecular weight 10000, the limits would be 9960 and 10040, 8
1002
Short communications TABLEI.-Rxsu~~s Solvent
OF m
ANDTHEIRPRRCISION
Solutes
Water Benxene Dioxan
Nitrobenxene
Sucrose, glucose, urea, catechol, resorcinol, hydroquinone, phenol, aniline, benxylamine Naphthalene, cr-naphthol, catechol, benxil, fats and oils Naphthalene, benxoic acid, 1,5-, 1,6-, 1,3-, 1,4-, 1,7-, 2,3-, 2,7-dihydroxynaphthalenes, methyl salicylate, nitrobenxene, pyrogallol, catechol, fats and oils Naphthalene, coumarin
No. of
Std. devn.,
results
mole/kg
72
*0~0035
65
&to*0016
105
~O+k321
20
f@OO15
but it often becomes more dilhcult to make such a concentrated solution as the molecular weight rises, It would seem unwise to claim that it is possible to extrapolate the readings to high molecular weights. i.e., in excess of 10000, as is sometimes done. Purity
Takhrg the mean standard deviation as 0002 mole/kg and assuming that twice this value is needed for detection, then the minimum amount of im urity that could be detected in organic materials soluble in any of the four solvents suitable, woul B be OXlOOmole % if a l-molal solution were used. Other uses
It is possible to examine self-associated species in benzene and nitrobenxene* and weak complexes in an inert solvent.’ The number of particles present under certain experhnental conditions can be determined and this might be useful in the study of reaction mechanisms. The nature of the recorder traces enables a study to be made of the last stages of the melting of liquids and dilute solutions. Melting points can be determined and although it is not claimed that T. is the physical melting point, it is based on theory and reproducible to within rtO*OO5”. Department of Chemistry The City University St. John Street, London E.C.I.
R.F. GREENWU~D
apparatus for the determination of the melting point of a partially frozen liquid is described. Thermal equilibrium is rapidly attained, largely because of the presence of glass balls in the experimental flask, and the method of agitation. The use of a thermistor and a recording potentiometer allows charting of the temperature changes with time. The melting point can be calculated from points on the trace. The apparatus has been used to study melting point depressions for solutions of hydrocarbons and phenols in benzene, nitrobenxene, water and dioxan. Summary-An
Zusammenfassnng-Em Geriit xur Bestimmung des Schmelxpunktes einer teilweise gefrorenen Fltlssigkeit wird beschrieben. Das thermische Gleichgewicht wird mit Hilfe eines Vorrats von Glaskugeln im Versuchskolben und eines besonderen Rtihrverfahren rasch erreicht. Ein Thermistor und ein Schreiber ermijglichen es, die xeitliche Anderung der Temperatur zu registrieren. De~Schmelxpunkt kann aus Punkten auf der Reaistriersour berechnet werden. Mit dem Gerltt sind Schmelxpur&ennied&ungen einiger Liisungen von Kohlenwasserstoffen und Phenolen in Benxol, Nitrobenxol, Wasser und Dioxan untersucht worden. R&ua&On d&it un a. pp”il pour la determlpfltion du point de fusion d’un liouide narttel ement congele. L’&nhbre themuque cst rapidement at&n, &entiellement 51&se de la pr6sence de b&s de verre dans la fiole d’ex&rhnentation et de la methode d’agitation. L’emploi d’un therm&& et d’un potentiom&re enregls&permet
1003
Short communications l’emegistrement des variations de temperature en fonction du temps. Le point de fusion peut Btm calculC de points sur 1s trace. On a utihs6 l’appareil pour 6tudier les depressions du point de fusion pour des solutions d’hydrocarbures et de phenols en be&me, nitrobenxene, eau et dioxane. REFERENCES 1. C. A. 2. N. E. 2354. 3. R. V. 4. R. F.
Kraus and R. A. Vingee J. Am. Chem. Sot. 1934,!6,511. Smidt, Y. A. Buslaew, U. G. Murator and L. D. Kulikovskaya,
Zh. Fiz. Khim., 1968,42,
Huggett, Srudentproject, The City University, London, 1%9. Greenwood, Ph.D. Thesis, University of London, 1967.
Talattta, 1970, Vol. 17, PP. 1003 to 1006. Pcntatttott Press. Printed in Northern Ireland
A silver electrode in the potentiometric titration of thiols*t (Received 26 Jury 1969. Revised 20 January 1970. Accepted 11 March 1970) Tus M~PEROMETRIC method of Bencsch et al? using silver as the titrant has been used extensively for measuring thiols. Kolthoff and co-workers* have made a critical study of the method and discussed the merits of several titrants. We had difBculty in locating the end-point in amperometric titrations, so we tried potentiometry. We have found that simply immersing a piece of silver wire in a solution of a thiol for a short time produces an electrode responsive to thiol concentration. Cecil and McPhee” had earlier used a silver-silver sulphide electrode prepared by an involved procedure and therestrictive conditions they reported may have discouraged extensive use of their method. Since the electrode follows the thiol level, any thiol reagent should be satisfactory as titrant, and the reagent forming the most stable compound should give the sharpest end-point. Mercury(H) salts, monosubstituted organo-mercuric salts and silver nitrate have been used successfully. The operating procedure was designed to accommodate the slow electrode response, and a simple treatment prior to each titration eliminated the reduction in electrode response that occurred with continued use. The response of the electrode would be expected to arise from RSH+RS-+H+;
IRS-I=&+$
A@) + RS- e E=E*-@o59log~
A@S(s)
+ e
1
= E”’ + O-059 log [RSH] (at constant pH) The electrode potentials for three different levels of cysteine at pH 7.0, measured with two electrodes (the Rrst a silver wire immersed in the solution until a steady potential was attained, the second deliberately coated with a heavy silver sulphide layer) gave slopes of @054 and 0.057 V, co&rming the proposed mechanism, though the E” value-s were different. EXPERIMENTAL Preparation of electrode Coil 0.15 m of l -27-mm diameter silver wire for several turns round a 4-mm tube. The first time the electrode is used, immerse it briefly in ammonia solution (1 + 3), rinse with demineralized water, immerse it in dilute nitric acid (1 + 3) for 2 min and rinse with demineralized water. * This work was supported in part by a USPHS Grant No. IT1 DE175 and in part by the United States Atomic Energ yCommission, Contract No. W-7401-Eng-49 and has been assigned Report No UR-49-960. t Thiols are generally referred to as “sulfhydryls” in American and biochemical terminology.
SHORT COMMUNICATIONS
Apparatus for the determination of melting points, molecular weights, freezing points and purity, and for the study of melting (Received10 June 1969. Revised 4 May 1970. Accepted 12 May 1970) AN APPARATUS for the rapid precise determination of molecular weights has been developed. It is based on the measurement of melting point lowering, because there are difficulties in obtaining accurate values for freezing points. *sl Notice is taken of the conditions which at&t precision and system. An equation is deduced for accurate calculation arise from processes occumnginthemel’ of the melting point, from an estimate Yo the time at which the melting point occurred, and the co-ordinates of two points in the melting region of the time-temperature record. EXPERIMENTAL The vessel was a lo-ml Pyrex glass bulb with two ground-glass necks, the one for filling and em tying having a ground-glass stopper, and the other containing a sealed-in thermistor armnged to Kave its tip immersed 5 mm into the liquid. The vessel contained 7 g of liquid and five glass balls 6 mm in diameter. The thermistor (FS 23), Standard Telephones and Cables Ltd.) had a resistance of 2000 haat 20”, and tem ture coetiicient 132 n/K, and was made one of the ratio arms of a Wheatstone bridge, the out-o p” -balance current being amplified and recorded on the chart of a recording potentiometer adjusted to give full-scale deflection for the concentration range being used. The flask was aled with a known weight of solvent and immersed in a freezing-bath for 1 min. It was then placed in an insulated container mounted on the arm of a reciprocating machine and swung uniformly, 90 times a minute, through an arc of 120”. The temperatur+time trace of the pure solvent formed a base line from which subsequent meltingpoint lowerings could be measured. Some examples are shown in Fig. 1. Conditions to be observed in design
6. 7. 8. 9.
Crystallization of the solvent should be rapid. Residual supercooling in the solid/liquid system or in the temperature measuring device is undesirable. Rapid heat exchange between solid and liquid is necessary so that equilibrium between the two phases can be established quickly. Melting should proceed through a series of equilibrium states, each state, for partially frozen solutions, being in accordance with the freezing point law. -sensing device must detect small changes accurately. The temperature The tem measureZera ._ _.should .-be that of the solution in equilibrium with frozen solid. At themeltingpotnt theconcentration of the solution in equilibrium with solid must be constant. The melting int should be relatively indifferent to the rate of melting. A clearly def!Zed melting point is desirable. The apparatus should be simple to use, and the time of measurement short.
DISCUSSION Crystallization of the organic liquids used, of water and of molten naphthalene and diphenylamme was rapid and clearly de&xl, largely because of the presence of the glass balls. The melting point was well defined on the recorder trace and by calculation could be measured accurately. The time required to remelt the partially frozen liquid was short, usually about 5 min. The thermistor tip was small and hence the temperature response rapid. Efficiency of mixing was the main factor involved in satisfying the other conditions, and was achieved by means of the glass balls, the umformity of the mechanical motion and the jerk at the end of each stroke. The uniformity of motion was most important. The efRciency was demonstrated by the smoothness of the trace for pure solvent and by the shape of that for a solution. The shape of the curve follows from the freezing point depression law, which may be ezpressed as
1000
Temperature FIG.
1 .-Examples
of time-temperature
f--
curves;
(a) pure solvent;
(b) solution.
where k = the cryoscopic constant, n = moles of solute, wO= weight of solvent at beginning of melting, T, = melting point of solvent, s = speed of melting, tI = time of melting to temperature TX. For any given trace for melting, the only variables are T and t and when plotted against each other they should give a concave curve. The shape of the trace in Fig. 1 (b) illustrates this. That all such traces were curved confirms the efficiency of mixing. The rate of heat exchange between the two phases was not only a function of mixing, but also of crystal size and the thermal properties of the two phases. Thus the heat exchange was least for pure solvents where the crystals were large, and greatest for the solvent cyclohexane, the latent heat of which was so low that melting was too rapid for a trace like Fig. 1 to be developed. Detemination of the melt@ point
The melting point of solvents (TL) is that portion of the trace which gives a constant temperature over an interval of time. The melting point for solutions can be taken as that temperature where the amount of frozen solid becomes vanishingly small (ZY,). At this point, T&-
where w is the weighed amount of solvent.
TV =
ATE =t
Short communications
1001
The expression used to calculate the melting point is ATIATOa ATs = AT,,(a + 1) - AT,
(3)
TO are two temperature differences on the trace, correwhereAT,=Tc-TlandAT,=TLsponding to times of melting I1 and Q a = (tl - t,,)/(r, - ra, and where te is the time at Te (see Fig. 1). From equations (1) and (2) we obtain ATew wo+stl=x’ andas w,,AT, = kn = ATew, then ATew ATew F + stx = m. 0
(5)
Thus,
If the rate of melting is the same throughout the melting process, and it is probable that this is so over a small temperature range and relatively short melting times, then the expression above is true whatever the temperature points chosen. Hence,
AT,AT,a ATe = AT,,(a + 1) - AT, ’ Estimation of precision The precision of cryosopic results is usually expressed in terms of temperature as this is the unit of measurement. It was thought that concentmtion would be a more meaningful term. The cryoscopic constant k estimatedfrom the slope of a plot of melting point depression VS.molal concentration m, may be used to estimate the concentration m’ corresponding to any given value of ATe, m’ = AT.Jk. The difIerences (m - m’) can than be used as a measure of precision. Results are shown in Table I. APPLICATIONS Molecular weights The precision with which the molecular weights of organic compounds can be measured by the apparatus can be seen from Table 1. The mean value of the standard deviations is &to402 mole/kg and the experimental value for a substance of molecular weight of 1000 in 1-molal solution would be found to lie between 1004 and 9%. If, however, the solution were 0.1 molal, the value would lie between 1040 and 960. It follows that within the bounds of ideality. the more concentrated the solutions the smaller the error. The samples in Table I included solutions up to l-2 molal. For a molal solution of a substance of molecular weight 10000, the limits would be 9960 and 10040, 8
1002
Short communications TABLEI.-Rxsu~~s Solvent
OF m
ANDTHEIRPRRCISION
Solutes
Water Benxene Dioxan
Nitrobenxene
Sucrose, glucose, urea, catechol, resorcinol, hydroquinone, phenol, aniline, benxylamine Naphthalene, cr-naphthol, catechol, benxil, fats and oils Naphthalene, benxoic acid, 1,5-, 1,6-, 1,3-, 1,4-, 1,7-, 2,3-, 2,7-dihydroxynaphthalenes, methyl salicylate, nitrobenxene, pyrogallol, catechol, fats and oils Naphthalene, coumarin
No. of
Std. devn.,
results
mole/kg
72
*0~0035
65
&to*0016
105
~O+k321
20
f@OO15
but it often becomes more dilhcult to make such a concentrated solution as the molecular weight rises, It would seem unwise to claim that it is possible to extrapolate the readings to high molecular weights. i.e., in excess of 10000, as is sometimes done. Purity
Takhrg the mean standard deviation as 0002 mole/kg and assuming that twice this value is needed for detection, then the minimum amount of im urity that could be detected in organic materials soluble in any of the four solvents suitable, woul B be OXlOOmole % if a l-molal solution were used. Other uses
It is possible to examine self-associated species in benzene and nitrobenxene* and weak complexes in an inert solvent.’ The number of particles present under certain experhnental conditions can be determined and this might be useful in the study of reaction mechanisms. The nature of the recorder traces enables a study to be made of the last stages of the melting of liquids and dilute solutions. Melting points can be determined and although it is not claimed that T. is the physical melting point, it is based on theory and reproducible to within rtO*OO5”. Department of Chemistry The City University St. John Street, London E.C.I.
R.F. GREENWU~D
apparatus for the determination of the melting point of a partially frozen liquid is described. Thermal equilibrium is rapidly attained, largely because of the presence of glass balls in the experimental flask, and the method of agitation. The use of a thermistor and a recording potentiometer allows charting of the temperature changes with time. The melting point can be calculated from points on the trace. The apparatus has been used to study melting point depressions for solutions of hydrocarbons and phenols in benzene, nitrobenxene, water and dioxan. Summary-An
Zusammenfassnng-Em Geriit xur Bestimmung des Schmelxpunktes einer teilweise gefrorenen Fltlssigkeit wird beschrieben. Das thermische Gleichgewicht wird mit Hilfe eines Vorrats von Glaskugeln im Versuchskolben und eines besonderen Rtihrverfahren rasch erreicht. Ein Thermistor und ein Schreiber ermijglichen es, die xeitliche Anderung der Temperatur zu registrieren. De~Schmelxpunkt kann aus Punkten auf der Reaistriersour berechnet werden. Mit dem Gerltt sind Schmelxpur&ennied&ungen einiger Liisungen von Kohlenwasserstoffen und Phenolen in Benxol, Nitrobenxol, Wasser und Dioxan untersucht worden. R&ua&On d&it un a. pp”il pour la determlpfltion du point de fusion d’un liouide narttel ement congele. L’&nhbre themuque cst rapidement at&n, &entiellement 51&se de la pr6sence de b&s de verre dans la fiole d’ex&rhnentation et de la methode d’agitation. L’emploi d’un therm&& et d’un potentiom&re enregls&permet
1003
Short communications l’emegistrement des variations de temperature en fonction du temps. Le point de fusion peut Btm calculC de points sur 1s trace. On a utihs6 l’appareil pour 6tudier les depressions du point de fusion pour des solutions d’hydrocarbures et de phenols en be&me, nitrobenxene, eau et dioxane. REFERENCES 1. C. A. 2. N. E. 2354. 3. R. V. 4. R. F.
Kraus and R. A. Vingee J. Am. Chem. Sot. 1934,!6,511. Smidt, Y. A. Buslaew, U. G. Murator and L. D. Kulikovskaya,
Zh. Fiz. Khim., 1968,42,
Huggett, Srudentproject, The City University, London, 1%9. Greenwood, Ph.D. Thesis, University of London, 1967.
Talattta, 1970, Vol. 17, PP. 1003 to 1006. Pcntatttott Press. Printed in Northern Ireland
A silver electrode in the potentiometric titration of thiols*t (Received 26 Jury 1969. Revised 20 January 1970. Accepted 11 March 1970) Tus M~PEROMETRIC method of Bencsch et al? using silver as the titrant has been used extensively for measuring thiols. Kolthoff and co-workers* have made a critical study of the method and discussed the merits of several titrants. We had difBculty in locating the end-point in amperometric titrations, so we tried potentiometry. We have found that simply immersing a piece of silver wire in a solution of a thiol for a short time produces an electrode responsive to thiol concentration. Cecil and McPhee” had earlier used a silver-silver sulphide electrode prepared by an involved procedure and therestrictive conditions they reported may have discouraged extensive use of their method. Since the electrode follows the thiol level, any thiol reagent should be satisfactory as titrant, and the reagent forming the most stable compound should give the sharpest end-point. Mercury(H) salts, monosubstituted organo-mercuric salts and silver nitrate have been used successfully. The operating procedure was designed to accommodate the slow electrode response, and a simple treatment prior to each titration eliminated the reduction in electrode response that occurred with continued use. The response of the electrode would be expected to arise from RSH+RS-+H+;
IRS-I=&+$
A@) + RS- e E=E*-@o59log~
A@S(s)
+ e
1
= E”’ + O-059 log [RSH] (at constant pH) The electrode potentials for three different levels of cysteine at pH 7.0, measured with two electrodes (the Rrst a silver wire immersed in the solution until a steady potential was attained, the second deliberately coated with a heavy silver sulphide layer) gave slopes of @054 and 0.057 V, co&rming the proposed mechanism, though the E” value-s were different. EXPERIMENTAL Preparation of electrode Coil 0.15 m of l -27-mm diameter silver wire for several turns round a 4-mm tube. The first time the electrode is used, immerse it briefly in ammonia solution (1 + 3), rinse with demineralized water, immerse it in dilute nitric acid (1 + 3) for 2 min and rinse with demineralized water. * This work was supported in part by a USPHS Grant No. IT1 DE175 and in part by the United States Atomic Energ yCommission, Contract No. W-7401-Eng-49 and has been assigned Report No UR-49-960. t Thiols are generally referred to as “sulfhydryls” in American and biochemical terminology.