The Short Answer
Melting point is a physical property. That is the textbook answer, and it is correct, but the way you measure it and what happens during the measurement is where people get confused. When a substance melts, it goes from solid to liquid. The molecules do not break apart or form new bonds. Water stays H2O whether it is ice or liquid. Sugar stays C12H22O11 when it melts. The identity of the molecule does not change. That is the definition of a physical change, so the temperature at which it occurs qualifies as a physical property. Here is what actually happens in the lab when you are running a melting point on a capillary tube. You pack maybe 2 to 3 millimeters of powder into the bottom of a thin glass tube, clamp it into a melting point apparatus, and heat it at a rate you control. The standard rate is 1 to 2 degrees Celsius per minute. If you heat faster than that, your reading will be wrong. I learned that the hard way on a batch of crude acetanilide where the thermometer lagged behind the heating block and I recorded a range that was 8 degrees too high. Once I switched to a smaller heat ramp and let the block equilibrate between samples, the readings dropped to the expected 113 to 115 degree range immediately.
The reason heating rate matters so much comes down to thermal conductivity. Glass capillaries are thin, but the sample inside is not a point source. The outer crystals feel the heat before the inner ones do. A fast ramp creates a temperature gradient across the sample itself. What you are really measuring is not a single transition temperature. You are measuring the span between when the first droplet of liquid appears and when the last solid crystal disappears. That span is called the melting range, and for a pure compound it should be tight. One to two degrees is typical. A range wider than three degrees usually means the sample is impure or the heating rate was too aggressive. Impurities depress the melting point and broaden the range. This is colligative behavior, the same principle behind salt on icy roads. If you mix even a small amount of a second compound into your solid, the lattice destabilizes. The structure does not hold together as well, so less thermal energy is needed to break it apart. The effect is predictable enough that melting point depression is used routinely as a diagnostic tool. If you have an unknown white solid and you suspect it might be benzoic acid, you run the melting point, get 120 to 124 degrees, and then do a mixed melting point with a known benzoic acid sample. If the mixture melts at 110 to 118 degrees, your unknown is not benzoic acid. If it melts at 122 to 124 degrees, you are probably looking at the same thing. The mixed melting point technique has been around since the early twentieth century and it still works because the underlying thermodynamics do not change. There are edge cases where the line between physical and chemical gets blurry, and this is where people trip up. Some compounds decompose on melting rather than melting cleanly. Copper sulfate pentahydrate loses water molecules as it heats. That is a chemical change happening at roughly the same temperature as the phase transition. If you see darkening, gas evolution, or a melting range that just keeps widening instead of resolving, decomposition is occurring. In those situations the observed temperature is not a true melting point. It is a decomposition temperature masquerading as one.
I ran into this with a synthesized intermediate that was supposed to melt around 156 degrees. Every batch came back at 148 to 154 degrees with slight browning. I spent two days troubleshooting the purification before I realized the compound was undergoing slow thermal degradation. Switching to a sealed capillary tube eliminated the decomposition because the volatile byproducts could not escape. The reading jumped to 155 to 157 degrees, which matched the literature value. Sealed tubes are the standard workaround for thermally sensitive samples, and they are cheap enough that there is no reason not to keep a box on hand. Polymorphs add another layer of complication. The same molecule can crystallize in different lattice arrangements, each with a different melting point. Ritonavir, an HIV medication, is the classic pharmaceutical example. A new polymorph emerged unexpectedly during manufacturing and had a lower solubility than the original form. The entire product had to be withdrawn and reformulated. For routine melting point work this is less dramatic, but if you are characterizing a newly synthesized compound and the lit value does not match your range, polymorphism should be on your checklist. Running differential scanning calorimetry or checking the X-ray powder pattern will tell you which form you actually have. Another practical detail that beginners overlook is calibration. Melting point apparatuses drift. The thermometer inside the block or the thermocouple reading can shift over time. I calibrate mine annually using pure standards: urea at 133 degrees, benzoic acid at 122 degrees, and cinnamic acid at 133 degrees. If my readings are off by more than a degree across all three, I adjust the instrument or flag the data. Un calibrated equipment gives false confidence, and false confidence leads to wrong conclusions about purity or identity.
Get the Full Details
So to answer the original question directly: melting point is a physical property because the phase transition does not alter molecular identity. But the measurement itself requires attention to heating rate, sample preparation, capillary sealing, and instrument calibration if you want the number to mean anything. The concept is straightforward. Getting a reliable number is where the work sits.
