Topological spin textures have drawn
intense attention
due to interesting
fundamental physics and possible application in non-volatile information
carriers as well as logic gate devices. Here, in a specially designed
race track that consists of three narrow nanotracks connected to a
wide nanotrack, we investigate the role of geometry in domain wall
(DW) pair to skyrmion conversion using micromagnetic simulation. In
particular, tunable DW to skyrmion or fractional skyrmion conversion
is achieved for a selected material parameter with a separation length
of 10 or 30 nm (or combination of both) between the narrow nanochannels.
By suitably varying the spacing between the narrow nanotracks symmetrically
and asymmetrically, we control the dynamics of skyrmions and fractional
skyrmions along with the trajectory. Interestingly, if the separation
length between the top and middle (or middle and bottom) nanochannel
is 30 nm, a fractional skyrmion is formed. The DW pair to skyrmion
conversion time depends on the separation between the narrow nanochannels,
e.g., for 10 nm separation, the conversion time of DW pair to skyrmion
from the top nanochannel is ∼0.3 ns, and the same for the 30
nm separation is ∼2 ns. Analysis of the topological number
of spin texture suggests the creation of two skyrmions in the case
of 10 nm separation between the narrow nanochannels, whereas for 30
nm separation, a skyrmion and a fractional skyrmion are formed. Furthermore,
the analysis of total energy and other energy terms shows a non-monotonic
variation during the conversion of DW to skyrmion at the junction.
Finally, the increase or decrease in the total energy value depends
on the formation of skyrmions or fractional skyrmions. Thus, we infer
that the enforced geometrical constraints and the interplay of various
energies play a crucial role in controlling the topology and skyrmion
formation. Based on these findings, we believe that a skyrmion racetrack
made up of three nanochannels will help achieve efficient controllable
skyrmion dynamics, which may have application potential in magnetic
memory operations.