According to both caffeine's pubchem page and chemspider page its boiling point lies at 173°C and its melting point at ~235°C. How can it melt at that temperature if it already sublimated at 173°C? Or is that the temperature of the triple point? If so, how come none of those pages gives any pressure specifications for that temperature?

I heated up some anhydrous caffeine at home, and at my home atmospheric pressure it appears to melt (not sublimate) and starts creating vapors immediately after (but not before) melting. This is based just on what I can see with the naked eye.

  • 3
    $\begingroup$ The reported temperatures are indeed dependent on pressure. Caffeine sublimes in a vacuum but melts under normal conditions. I am unsure as to why the pressures wouldn't be reported adjacent to the reported temperatures but then again, these websites don't really have to. Refer to peer-reviewed literature for that sort of detail. Also, the pubchem page you linked says "Fast sublimation is obtained at 160-165 deg C under 1 mmHg pressure at 5 mm distance." so it does offer some idea what the pressure is for sublimation. $\endgroup$ – LordStryker Dec 18 '13 at 15:50
  • $\begingroup$ that still does not explain at what temperature caffeine boils under normal conditions. and if they specify a melting point under normal conds, why specify the boiling point under some other condition? $\endgroup$ – TheChymera Dec 20 '13 at 16:07

See "VAPOR PRESSURE AND HEAT OF SUBLIMATION OF CRYSTAL POLYMORPHS" Journal of Thermal Analysis and Calorimetry, Vol. 57 (1999) 45-60 and "PHASE TRANSITIONS AND THERMODYNAMIC PROPERTIES OF ANHYDROUS CAFFEINE" Journal of Thermal Analysis, Vol. 16 (1979) 267-275 for detailed information.

Anhydrous caffeine occurs in two different polymorphic crystal forms, with a transition between the two forms occurring at 136°C. The high temperature form has a melting point of 236°C.

The vapor pressure of the high temperature solid form is described by the equation:

logP[Pa] = -5477/T[K] + 14.395

and for the liquid

logP[Pa] = -3918/T[K] + 11.14

So 173°C is incorrect; the boiling point is more like 365°C.

| improve this answer | |
  • $\begingroup$ Ok, thanks for the references (which i cannot currently read - maybe after i get my institution log in back); but why then do those pages give a value of ~170°C for the melting point? $\endgroup$ – TheChymera Jan 2 '15 at 21:21
  • 1
    $\begingroup$ Interestly, the 1876 Yearbook of Pharmacy (British Pharmaceutical Conference) at page 22 discusses this issue. As caffeine is heated, some sublimation is noticed before the melting point is reached. The links in your question both state "(sublimes)" after the 178 C value. This value only indicates that sublimation will be noticed above that point. The boiling is much higher. $\endgroup$ – DavePhD Jan 2 '15 at 22:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.