This is due to capillary condensation in so-called "ink-bottle" shaped pores, i.e. having a entrance diameter smaller than the size of the pore body. Condensation occurs progressively in such pores until the pores are completely filled by the condensable vapour at high relative pressure. But when the relative pressure decreases, the liquid needs to go out of the pores and therefore needs to pass the narrow necks. This occurs at a lower relative pressure than what was required for filling the pores, and when the critical relative pressure is reached, the pores empty abruptly. Hence the loop in the isotherms.
This is due to capillary condensation in so-called "ink-bottle" shaped pores, i.e. having a entrance diameter smaller than the size of the pore body. Condensation occurs progressively in such pores until the pores are completely filled by the condensable vapour at high relative pressure. But when the relative pressure decreases, the liquid needs to go out of the pores and therefore needs to pass the narrow necks. This occurs at a lower relative pressure than what was required for filling the pores, and when the critical relative pressure is reached, the pores empty abruptly. Hence the loop in the isotherms.
Hysteresis is the name of generic phenomena with a same class of characteristic: after a stimulus some information about or generated by the stimulus is maintained. In this sense, as an example, it is possible to mention thermal hystheresis, ferroelectric hysteresis and magnetic hysteresis, obviously each one of them with a proper physical-chemistry stimulus.
The hysteresis of Adsorption/desorption of N2 is correlated with condensation of N2 at mesoporo. During step of adsorption N2 is compressed in the porous, in this sense, under pressure the N2 gas turn a liquid. In the desorption, the pressure is decreasing gradually but N2 is now a liquid and undergo the effects of capillarity, (major action) turning a gas again at different P/Po ratio. Here, it is possible a more deep insight, but seems not necessary.
The area and shape of hysteresis depend on the area and pore geometry. This phenomenon was studied by several reseachers but is more commonly reported from BET Isotherm theory, Sthephen Brunaeur, Paul Emmet and Edward Teller, Brunaer/Emmet/Teller, BET. A priori, from BET isotherm is possible to derive the surface area of solids, but a serie of parameters, as mentioned it is possible to determine the geometry of pores.
The answers above give a very good description of the physical basis. If you want more elaborate explanation, look in the literature for work by Matthias Thommes.
A hysteresis loop is formed by adsorption-desorption N2 curve in mesoporous materia.
in this case ,the hysteresis of Adsorption/desorption of N2 is correlated with condensation of N2 at mesoporos. During step of adsorption N2 is compressed in the porous, in this sense,at a certain p/p0 (i.e 0.45 to 0.50 ) pressure the N2 gas turn a liquid and it behave works as a capillary condensation in with in the pore like "ink- bottle" .when the N2 gas turn to liquid its entrance diameter smaller than the size of the pore body. Condensation occurs progressively in such pores until the pores are completely filled by the condensable vapour at high p/p0 relative pressure.
But during the desorption when the relative pressure decreases, the liquid needs to go out of the pores and therefore needs to pass the narrow necks. This occurs at a lower relative pressure than that is required for filling the pores, finally when the critical relative pressure is reached, the pores empty abruptly. this is called the loop in the isotherms.This is due to capillary condensation in so-called "ink-bottle" shaped pores, i.e. having a entrance diameter smaller than the size of the pore body. Condensation occurs progressively in such pores until the pores are completely filled by the condensable vapour at high relative pressure. But when the relative pressure decreases, the liquid needs to go out of the pores and therefore needs to pass the narrow necks. This occurs at a lower relative pressure than what was required for filling the pores, and when the critical relative pressure is reached, the pores empty abruptly. This is due to capillary condensation in so-called "ink-bottle" shaped pores, i.e. having a entrance diameter smaller than the size of the pore body. Condensation occurs progressively in such pores until the pores are completely filled by the condensable vapour at high relative pressure. But when the relative pressure decreases, the liquid needs to go out of the pores and therefore needs to pass the narrow necks. This occurs at a lower relative pressure than what was required for filling the pores, and when the critical relative pressure is reached, the pores empty abruptly. Hence, this is formed loop in the isotherms.
Basically the hystesis loop formed due to the capillary condensation of N2 inside the pore system at high relative pressure. Whatever the type of pores of mesoporous materials. Which may be Cyliderical, ink-bottle, slit-type, cone-shapped type of pores.
Where the N2 gas converted into liquid by the effect of high relative pressure P/Po.
The N2 uptake required to remove the physically adsorbed N2 gas inside the pores or at the surface during desorption process is much higher that taken during the adsorption process.
All answers are very descriptive and helpfull, but as Heba Gobara says, capillary condensation occurs whatever the mesopore shape is. The hysteresis loop is the consequence of the two different mechanisms that happen during adsorption-desorption of vapors in mesopores: while adsorption mechanism involves the creation of a mesopores-surface coverage layer by layer, that finlally yields to a delayed condensatation, desorption occurs from the evaporation of liquid in mesopores from the meniscus (as Kelvin equation describes). Ink-bottle pores are not necesary for hysteresis, but they can include pore-blocking or even cavitation phenomena, which reduce the P/P0 value where adsorption and desorption branches mathches (lower than the usuall 0.6 value).
You may also read: Pure & Appl. Chem., Vol. 61, NO. 11, pp. 1845-1852, 1989 iupac.org/publications/pac/61/11/1845/pdf/; https://hal.archives-ouvertes.fr/docs/.../la051030o.pdf and http://scholarworks.umass.edu/cgi/viewcontent.cgi?article=1015&context=open_access_dissertations