See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271513918

True Melting Point Determination

Article · January 2013

DOI: 10.1333/s00897132500a

CITATIONS READS
0 7,757

2 authors, including:

John Young
Saint Mary's University
15 PUBLICATIONS   14 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Autotitrator titrant dispenser hydrodynamics View project

All content following this page was uploaded by John Young on 25 August 2015.

The user has requested enhancement of the downloaded file.

 Chem. Educator 2013, 18, 203–208 203

True Melting Point Determination

John C. O’C. Young

Department of Chemistry, Saint Mary’s University, Halifax, Nova Scotia, Canada, B3H 3C3,
john.young@smu.ca
Received May 17, 2013. Accepted June 21, 2013.

Abstract: While pure compounds melt isothermally at constant pressure, their melting points are commonly
reported as the range between the temperatures recorded by a heated capillary-type melting range measuring
instrument at the onset and at the completion of melting of a sample as determined by a pharmacopeiac test
protocol. Based on the use of a simple, low-cost version of this type of instrument, a procedure is presented and
exemplified for determining the true “thermodynamic” melting points of pure compounds from such melting
range data.

Introduction
The melting points of chemical compounds are routinely
determined and reported, and they can serve as a useful
indicator of substance identity and purity. However the
meaning of these reported data is not always clear.
Organic chemistry textbooks and laboratory manuals
commonly report that pure substances melt over narrow
temperature ranges and reagent bottles are often labeled with
“m. p.” ranges of several degrees, for which Roberts et al.
offer the following explanation [1]:
The melting point of a pure compound is characteristic of
the substance being studied. In practice, most melting
point apparatus is designed for ease of use and does not
permit the determination of the exact temperature at which
a compound melts, that is the melting point. Instead a
melting range (authors’ emphasis) is actually measured
and is commonly reported as a “melting point” for most
compounds. If a solid substance is pure, it should melt
over a very narrow range, which should be about 1 oC if it
is determined carefully.
However we learn from physical chemistry textbooks that,
aside from the slight pressure dependency defined by the
Clapeyron relationship, chemically and physically uniform
solid substances melt at a specific, constant temperature, not
over a “narrow range.” At a given pressure, this is the
temperature at which the liquid and solid phases equilibrate,
and they cannot achieve equilibrium if the system is being
heated at a finite rate from an external source, as is the
circumstance in conventional instruments that purport to
measure melting point. At the particulate level, the equilibrium
temperature corresponds to the unique vibrational energy level
which is just sufficient to shake loose the surface layer of solid
atoms or molecules from the attraction of the underlying solid
layer, allowing them to enter the contacting liquid layer.
Assuming a finite rate of heating the system, the temperature
of the previously melted liquid phase will be higher, and that
in the interior of the residual solid phase will be lower, than the
true “thermodynamic” melting temperature of the thin surface
solid layer.
Roberts et al. continue [1]:
The melting point apparatus available in most organic
chemistry laboratories (presumably meaning Fisher-Johns
and capillary type melting range instruments) is not
sufficiently sensitive to allow reporting of a single melting
point temperature. Instead the melting point should be
expressed as the temperature range at which the solid
started to melt and at which it was completely converted to
liquid... Unless an exceedingly expensive and elaborate
melting point apparatus is used, it is improper and
incorrect to report a single temperature for a melting point.
However, the matter is not one of sensitivity. While the
solid is melting at constant temperature, what is observed are
the temperatures recorded by the temperature sensor at the
beginning and end of the melting interval. Typically, in
modern instruments, the solid sample is placed in a capillary
tube which is brought into contact with a metal block oven
containing the temperature sensor and a heating element that
receives an electrical input regulated to maintain a constant
rate of temperature increase of the oven, the so-called ramp
rate. However sensitively temperatures are measured, with this
arrangement the temperature range observed during melting
must necessarily lie entirely above the true melting
temperature by an amount that depends on the design of the
particular instrument, the chosen ramp rate and to a lesser
extent the size of the sample to which the heat transfer area
should be close to proportional.
It is hardly surprising that our students are confused by this
fallacious concept of a melting range of a pure substance and
have been known to suggest that the “m.p.” range on the label
of one of our bottles of pure lauric acid of 44 – 46 oC might be
due to melting point depression by an impurity. However, the
cryoscopic constant of this compound is only 1.3 oC molal–1,
which means for example that a 15 % w/w concentration of a
contaminant of molar mass 100 would be required to cause a
melting point depression of 2 oC. The purity of this reagent
was stated to be 99.5%+ and pure lauric acid is reported to
melt at close to 43.8 oC [2]. Likewise, pure benzoic acid does
not melt over a temperature range: its melting point at
atmospheric pressure is reported to be 122.383 oC [3].
While not directly indicating the true melting point value,
the melting range data delivered by heated capillary-type
instruments do provide a rapid and meaningful indication of

© 2013 The Chemical Educator, S1430-4171(13)12500-6, Published 07/19/2013, 10.1333/s00897132500a, 18130203.pdf

204 Chem. Educator, Vol. 18, 2013
Figure 1. The “Melt Station”.
compound purity for routine quality control purposes
providing that measurements are made under specified
conditions of system design and operation. In North America,
these specifications are provided by the United States
Pharmacopoeia (USP) [4] and an informative and
comprehensive review of these specifications and of issues
relating to melting range and melting point measurement have
been reported by Stanford Research Systems [5]. This source
also presents a procedure similar to the one that we propose for
determining true melting points from melting range
measurements.
Used in conjunction with Vernier’s “LabPro” A/D interface
and “Logger Pro” data analysis and presentation program, that
company’s “Melt Station” melting range instrument [6]
illustrated in Figure 1 is easy to operate and is neither elaborate
nor expensive: the current cost of this instrument is
approximately one third of the last quotation we received for
the obsolete Fisher-Johns melting range apparatus before its
manufacture was terminated. We will show how the “Melt
Station” and other capillary type melting range instruments
may be used to determine the true melting points of pure
substances.
The Melting Process
At any given instant during the melting process, the heat
input dq to the sample over a time interval dθ isgiven by
dq = U A (t – tmp) dθ = UA (t – tmp) dt / R
in which R = dt / dθ is the ramp rate at time θ, t is the heating
oven temperature, tmp is the true melting point of the
compound, U is the coefficient of heat transfer between the
oven and the capillary tube and A is the contact area between
the oven surface and the section of the capillary tube
containing the sample.
The total heat input during melting of a sample of mass m
and latent heat of fusion Lf
 top  t-t mp  dt/R
t Experimental
q = m Lf = U A cp
in which top is the onset point temperature at which the
beginning of melting is observed and tcp is the clear point
temperature at which the last trace of solid melts, and from
© 2013 The Chemical Educator, S1430-4171(13)12500-6, Published 07/19/2013, 10.1333/s00897132500a, 18130203.pdf
John C. O’C. Young
which upon integration and rearrangement yields the
relationship
(tcp–tmp)2 – (top–tmp)2 = 2 (m Lf/(U A)) R = a constant * R
reasonably assuming that the ramp rate R does not change
significantly over the melting interval. In more sophisticated
instruments, the heat input to the oven is regulated to maintain
a constant ramp rate in face of progressively increasing heat
loss from the oven to its surroundings as its temperature is
raised. In the case of the “Melt Station”, a chosen constant heat
input rate is applied to the oven, which means that the ramp
rate will decrease somewhat over the melting interval.
However this effect is small and may be accommodated by
computing the mean ramp rate over the melting interval as R =
(tcp – top) / (θcp – θop) or determining it from the slope of a
linear regression fitted to the temperature versus time data
obtained within the θop to θcp time interval.
As indicated by the above relationship, the melting range
expands as the ramp rate is increased. Oppositely, the onset
point temperature and the clear point temperatures must
converge on each other and also on the true melting point
value in the limit as R approaches zero, corresponding to the
state of equilibrium associated with an infinitely slow rate of
heating. This relationship satisfies the requirement that tcp =
top = tmp at R = 0. It also satisfies relationships of the form tcp =
tmp + ccp R0.5 and top = tmp + cop R0.5 in which ccp and cop are
constants.
Employing these relationships
(tcp–tmp)2 – (top – tmp)2 = (ccp2 – cop2) R = 2 (m Lf / (U A)) R
From the clear point temperature versus ramp rate relationship
tcp = tmp + ccp R0.5, it follows that an estimate of the value of the
true melting point tmp may be obtained from the ordinate
intercept of a linear regression fitted to a plot of tcp on the
square root of R. A second estimate of this value may be
obtained by applying the same procedure to the onset
temperature data set.
The onset point and clear point temperatures called for in the
USP procedure [4] are obtained by inserting a value of R = 1
o
C per minute into these relationships between temperature and
the square root of the ramp rate. However, the onset point
temperature is somewhat more difficult to detect than the clear
point temperature and its regressional relationship with R
usually exhibits a larger root mean square error. Thus to
determine the USP range, our practice is to apply any small
correction necessary to shift the onset point temperature
regression relationship up or down as necessary to make its
intercept coincide with that of the clear point data plot. Simply
to determine the true melting point, the onset point temperature
data may be ignored. To determine the true melting point, the
clear point temperature must be observed at at least two
different ramp rates, or heat input rates in the case of the “Melt
Station.”
Obtaining melting range data from previously melted
samples is occasionally discouraged [5] on grounds of possible
sample decomposition or other physical or chemical change.
However, we have yet to encounter this type of problem based
True Melting Point Determination Chem. Educator, Vol. 18, 2013 205

Figure 2. RTD Response Data at Ramp Rates of 0.62 – 5.47 oC per minute.

Figure 3. Onset Point and Clear Point Temperature Data vs. Ramp Rate Function.

Figure 4. RTD Response Data at Ramp Rates of 0.68–4.60 oC per minute.

© 2013 The Chemical Educator, S1430-4171(13)12500-6, Published 07/19/2013, 10.1333/s00897132500a, 18130203.pdf

206 Chem. Educator, Vol. 18, 2013 John C. O’C. Young

Figure 5. Onset Point and Clear Point Temperature Data vs. Ramp Rate Function.

Figure 6. Onset Point and Clear Point Temperature Data vs. Ramp Rate Function.

Figure 7. Onset Point and Clear Point Temperature Data vs. Ramp Rate Function.

© 2013 The Chemical Educator, S1430-4171(13)12500-6, Published 07/19/2013, 10.1333/s00897132500a, 18130203.pdf

True Melting Point Determination
Figure 8. Onset Point and Clear Point Temperature Data vs. Ramp Rate Function.
on repeated testing of the same sample. If it did occur, sample
deterioration at temperatures up to slightly above the clear
point accompanying a single melting would perhaps render any
melting point data so obtained of questionable significance. In
any event, if sample deterioration did occur, it should be
readily detectable by test repetition and a fresh sample could
be used.
We frequently expose our samples to repeated melt-freeze
cycling at different ramp rates. We usually begin by melting a
sample of an unknown compound in a capillary tube at the
maximum heating rate provided by the “Melt Station”
instrument to obtain an approximate indication of the melting
rangeof the sample and to expel air from it. Upon subsequent
freezing, which we accelerate by temporarily placing crushed
ice in test tubes in contact with the oven, the sample forms a
coherent air-free mass with a well-defined upper surface which
facilitates visual detection of the onset point temperature
because detectable melting generally begins at the sample
surface adjacent to the wall of the capillary tube.
Samples often supercool as much as 30 oC below their
melting points and our heating procedure usually begins at 40
o
C below the approximately known value of the melting point.
From this point we heat the sample at the maximum rate
obtainable to a fixed target temperature approximately 20 oC
below the anticipated melting point, at which juncture we dial
in a lower heating rate to give us approximately the desired
ramp rate during melting. Observations of the onset point and
clear point temperatures are manually entered on a real time
plot of oven temperature versus time by mouse clicks. The
sample is then cooled back to the start of heating temperature,
heated at the maximum rate to the same fixed target
temperature at which a new value of the heating rate is dialed
in to obtain melting temperature and time data at a new ramp
rate as exemplified in Figure 2.
Calibration of a melting point measuring instrument may be
carried out by observing the melting behavior of a standard
melting reference compound whose melting point is accurately
known. For this purpose, we initially chose a phenacetin (N-(4-
ethoxyphenyl) acetamide) melting point standard, for which
clear point temperatures of 135.3 + 0.13 and 135.7 + 0.13 oC
were reported by the supplier [7] for ramp rates of 0.2 and 0.5
o
C per minute respectively, indicating a true melting point of
134.6 + 0.13 oC, based on the ordinate intercept of a linear
© 2013 The Chemical Educator, S1430-4171(13)12500-6, Published 07/19/2013, 10.1333/s00897132500a, 18130203.pdf
Chem. Educator, Vol. 18, 2013 207
regression to a plot of these temperature data versus the square
root of the ramp rate, and a USP melting range of 135.5–136.1
o
C at a ramp rate of 1 oC per minute. We repeated the
calibration procedure based on a benzoic acid reference
standard [8], a compound for which a melting point of 122.383
o
C has been reported [3]. Following these calibration
procedures, we determined the melting points and USP melting
ranges of samples of lauric, myristic and salicylic acids.
Results and Discussion
Onset point and clear point temperature data that we
obtained from a sample of the phenacetin melting point
reference standard are shown in Figure 2. Plots of these data
versus the square root of the ramp rate are shown in Figure 3.
These relationships indicated a true melting point of 134.4 oC
and a USP melting range of 135.2–136.4 oC which are in
reasonable agreement with the values that we derived above
from the data supplied with the phenacetin reference
compound, namely 134.6 + 0.13 oC and 135.5–136.1 oC
respectively.
Onset point and clear point temperature data that we
obtained from a sample of our benzoic acid reference
compound are shown in Figure 4. Plots of these data versus the
square root of the ramp rate are shown in Figure 5. These data
plots indicate a true melting point of 122.3 oC and a USP
melting range of 123.4–124.5 oC. The melting point of benzoic
acid is reported elsewhere to be 122.383 oC [3].
The temperature sensing performance of the system thus
appeared to be acceptable, and we next turned to apply it to
studying the melting behaviours of lauric (dodecanoic) acid,
myristic (tetradecanoic) acid and salicylic (2-hydroxybenzoic)
acid.
The onset point and clear point temperature data shown in
Figure 6 were obtained from a sample of lauric acid [9] which
was labeled as 99.5% pure with an “MP” of 44–46 oC, which is
in reasonable agreement with the USP temperature range of
44.8–45.9 oC indicated by our data. Our data also yield a true
melting point value of 43.5 oC which lies between reported
values of 43.2 oC [10] and 43.8 oC [2].
The onset point and clear point temperature data shown in
Figure 7 were obtained from a sample of myristic acid [11]
which was labeled as 99.5+ % pure with an “m.p.” of 55.1 oC.
Our data indicate a true melting point of 53.8 oC and a USP
208 Chem. Educator, Vol. 18, 2013 John C. O’C. Young

melting range of 55.3–56.3 oC. Melting point values of 54.2 oC 2. Haynes, W., ed., CRC Handbook of Chemistry and Physics, 91st edn.,
(2) and 54.4 oC (10) been reported for this compound. CRC Press, Boca Raton, FL, 2010, (entries for dodecanoic,
tetradecanoic and 2-hydroxy benzoic acids).
The onset point and clear point temperature data shown in
Figure 8 were obtained from a sample of salicylic acid [12] 3. Crovini, L., Marcarino, P., Milazzo, G. Apparatus for the Accurate
Determination of Melting Points and Freezing Points. Anal. Chem.,
which was labeled as having a “Melting Point” of 159 oC to 1981, 53, 681–686.
161 oC. Our data yield a true melting point of 158.5 oC and a
4. USP25-NF20 US Pharmacopoeia, Section <741>, p. 2033–2034.
USP melting range of 159.6–160.6 oC. A melting point of
159.0 oC has been reported for this compound (2, 10). 5. www.thinksrs.com/downloads/PDFs//ApplicationNotes/MPProcedur
es.pdf. Application Note #3: Pharmacopeia vs. Thermodynamic
Melting Point Determinations. Application Note #4: Determination
Conclusion of Melting Points According to Pharmacopeia (accessed May 17,
2013).
The procedure we have described for determining the 6. Vernier Software & Technology, Beaverton, OR 97005–2886
melting point and USP melting ranges of pure compounds
7. Phenacetin Melting Point Standard, Catalog # PHR-1094, Research
from plots of their onset point and clear point temperatures Technology Corporation, Laramie , WY 82070.
versus the corresponding value of the square root of the 8. Benzoic Acid Certified Primary Standard, Catalog # A68–30,
applied ramp rate appears to yield acceptable results. Thermo Fisher Scientific Company, Ottawa, ON K2E 7L6.
9. Lauric Acid, Catalog # L-1400, ACP Chemicals Inc., Montreal, PQ
References and Notes H1R 1A5.
10. en.wikipedia.org/wiki/Lauric_acid, Myristic_Acid and
1. Roberts, R., Gilbert, J., Rodewald, L., Wingrove, A. Modern
Salicylic_Acid (accessed May 17, 2013).
Experimental Organic Chemistry, 4th edn., Saunders, Philadelphia,
PA,1985, p. 81. 11. Myristic Acid, Catalog # 15,379–6, Inc., Sigma-Aldrich Chemical
Company, Inc., Milwaukee, WI 53209.
12. Salicylic Acid , Catalog # 10230, BDH Chemicals Ltd., Poole,
England.

© 2013 The Chemical Educator, S1430-4171(13)12500-6, Published 07/19/2013, 10.1333/s00897132500a, 18130203.pdf

View publication stats