Basically, the pharmaceutical melting point isn't the same as that used by
chemists and others.
In the days before there was any kind of organic spectroscopy, the melting
point was a very important property of a compound. It isn't exactly a point,
for practical reasons and because impurities tend to spread it out in a way
that depends (in practice, forget the theory) on the rate of heating.
If you're writing a chemistry paper for the primary literature, you should
give a range: the lower temperature is where the product just starts to show
signs of melting, and the upper limit is the point where the last particle
In pharmaceutical practice, you record only the upper limit, for a quantity
of product in a so-called capillary tube that seems astronomical in other
contexts. There's nothing wrong with this provided the heating rate is
defined - it's 1°C/min if I remember rightly. The objective in this context
is to identify the compound, not in the absolute sense, but with respect to
other compounds it might be mistaken for. The only problem is that some
people of a regulatory bent want a melting point when there isn't one; if
there are any signs of decomposition refuse to quote a value.
To answer you question, you can use a flask of oil heated by a bunsen,
provided you keep within an undefined (pharmacists...!) range of the
prescibed heating rate. You can buy equipment that provides programmed
One problem with USP and EP general methods like this one is there that
there is no traceability - you can't easily find any background information
that explains why a method is defined the way it is, so you feel like an
idiot when you find a problem with it.
Finally, I wonder whether you confuse digital with automatic. There exist
instruments that detect the melting point automatically (generally
optically). They detect when the sample moves, which isn't the USP melting
point, and don't tell you if the compound is melting or decomposing; they
are worth the money in some production environments. That's very upsetting
for quality assurance people as well as analysts, cos the only way to prove
that nobody is cheating on the melting point is to make a - validated -
video of the sample & the display. But here be dangerous territory...
"information_person" <[Only registered users see links. ]> a écrit dans le message de news: [Only registered users see links. ]...
Thanks - i found your insight useful. However your discussion does bring up
a point i didn't formally consider, decomposition.
Wouldn't the literature not distinguish melting point from decomposition and
thus we are really measuring one of the crystal homologs of the compound
melting point and /or decomposition. If the decomposition occurs at a lower
temperature than melting point how is one supposed to determine the melting
Do you suppose that the manual analog melting points are acceptable as the
digital melting point devices?
And also - what approximate rate of temperature increase should be used for
accurate and reproducable results ? is 0.1 C per minute too slow, how about
1 C per minute or even 5 C per minute?
"Chris R. Lee" <[Only registered users see links. ]> wrote in message
news:[Only registered users see links. ].clara.net...
On Tue, 11 Nov 2003 00:22:35 -0500, "information_person"
<[Only registered users see links. ]> wrote:
Most pharmacopeia methods aren't applicate to materials that decompose
below the melting point, and may also require the sample be dried to
ensure a sharp melting point. Note that the sample is only added as
the expected melting point is approached to minise degradation.
You should get a copy of the pharmacopeia method that's applicable.
For USP 27 ( 2004 ) the section is <741> pages 2324-2325. Note that
USP reuires you to check using one of their standards ( they have 6
reference compounds with specified melting points ).
There are several differing procedures, depending on the compound and
melting point instrument. Basically, you heat the bath or block up to
30C below the expected melting point, remove the thermometer, and
attach the capillary with the sample, put it back in the bath, and
heat at ~3°C/min until 3°C below the beginning of the melting range,
then reduce heating to achieve about 1-2°C/min, and continue heating
until the melting is complete.
For compounds with no method specified you usually heat to 10°C below
the expected melting point, but rising at 1.0+-0.5°C/min and add the
sample when it's about 5°C below the beginning of the melting range.
Youd continue heating until melting is complete.
The first appearance of melting or slumping in the tube is defined as
the beginning of melting, the temperature when the sample is liquid
throughout is called the melting point, and the difference is called
the melting range.