Dear

The increase of temperature is usually indication of reaction, however some compounds have high temperature but usually they consisted in a liquid form. Evaporation of the solution is high risk especially when you deal with organic compounds. Try to add few particles of the same compound or other easily isolation substances during the cooling process. You can try with small amount the fast cooling, from high temp. To very low temp. But be careful it's sometimes deform the crystals structure. 

If the solubility of the compound in your system is normal, then try with much lower cooling rate when no evaporation occurs.

I would first lower the cooling rate. If you cool and heat at the same rate and nucleation occurs around -5C only during heating, you may be near the nose temperature while effectively letting the system equilibrate at that range of temperature a little longer. Also, try to lower the minimum T to -10C just to check if the system crystallizes during cooling. Of course, I am assuming that your system nucleates homogeneously. Besides that, can you control in any way the evaporation? I think that could be the issue too.

I think LGV Gonclaves has the proper interpretation and suggestion for you.

I have a question for you.  Have you played with any other variables such as cooling to below -5 deg C (or is this a limitation of your equipment?), or holding your sample at -5 deg C for longer periods of time?

As suggested by Goncalves, lowering the cooling rate is a good thing to try as the cooling rate applied can alter the MSZW of your system too - holding at your lowest temp for a longer period might go someway to verify this; however nucleation might occur still after only long time.

If you continue to have problems with this system, might an anti-solvent addition be a useful way to crystallize your target, i.e. are you looking for crystalline material or specimen candidates for single-crystal study?

I agree with Dimitrov, Gonclaves and Gliksman

Maybe the other compounds are so-called "stronger crystallizers" than naringenin?  Might also be worth checking the form, you could be isolating a kinetic phase...

Your system is behaving exactly like that of unstable glass systems such as those used to create glass ceramics. In these systems on the cooling stage supersaturation occurs but there are no nuclei on which crystals can grow. At the lower temperatures, nucleation occurs and very small nuclei form in large numbers but at these lower temperatures, the growth kinetics is hampered by very high viscosity, so the nuclei are unable to grow. As the temperature is increased, in the case of glassy systems, the viscosity decreases, reaction kinetics speed up and the nuclei grow into a very large number of small crystals. This controlled thermal treatment, especially the time spent at the lower temperatures, is very succesfully used to control the microstructure of glass ceramics. The process is very reporoducible.

Dear

The crystallization is gouverned by the nucleation and growth mechanism as explained by Pr. Albert F. Wright. Morover the density of activated nuclei depends also on the crystallization conditions (cold crystallziation or crystallization from the melt). See below the density of activated nuclei given as function of temperature for the crystallization from the glassy state (cold crystallization) and during cooling (crystallization from the melt) (for more details see 1 )

See the Attached file 

From calorimetric measurements one can observe this fact also in terms of the half crystallization time ( the time needed to transform the half of the sample, it depends on the nucleation density and the growth rate ) see for example Figure 6 (2).

I hope that helps

See the Attached file 

(1)         Janeschitz-Kriegl, H.; Ratajski, E. Colloid and Polymer Science. 2010, pp 1525–1537.

(2)         De Santis, F.; Pantani, R.; Titomanlio, G. Thermochim. Acta 2011, 522 (1-2), 128–134.

Thank you all for the answers.

Now I was able to generate nucleation, while performing cooling crystallization. I started at 60 C and then went to -5 C at 15 C/h. Nuclei appeared around 20 C this time. The crystals I obtained by evaporation were yellow colored. By cooling crystallization they now were almost white and they appear to grow as spherulites.

I tried today to use these obtained white needles/spherulites as seed crystals. Now there was decrease in the liquid concentration (with the yellow ones as seeds there was not), but they only seemed to grow at relatively high supersaturations (3.5 relative supersaturation) and in larger and larger spherulites. I was decreasing the temperature at 10 C/h.

Do you have an idea why I couldn't see the white needle crystals growing at lower supersaturation?

It is not strange since this behavior strongly depends on whether it is endothermic or exothermic when nucleation occurs

How about the purity of solution? If there is high concentration of impurity, the result you observed will appear