The Fe3O4-HNTs provide 1-dimensional matrix to shorten the diffusion path of electrons and electrolyte ions as well as to absorb the mechanical stress during cycling along with excess sites for charge storage, while N-GQDs offer abundantly accessible electroactive sites for rapid electrons and electrolyte ions transport as well as enhance electrical conductivity of Fe3O4-HNTs.
A linearity increase of capacitance with O/C ratio can be attributed to the increase of the population of surface oxides on CNTs, which imparts excess sites for redox reaction (pseudocapacitance) and for the formation of double-layer (double-layer capacitance).
In contrast, nonRR units showed no excess reward site bias for replayed spatial content (excess reward site bias for nonRR units at centrifugal replay 0.002 ± 0.019, p=1, chi = 0.001, Chi-square test; at centripetal replay: 0.011 ± 0.022; p=0.6, chi = 0.2).
However, we were unable to detect a difference in the excess reward site bias of RR units at replay events immediately following correct trials versus error trials (excess reward site bias after correct trials 0.022 ± 0.020; after error trials -0.018 ± 0.031; p=0.19, nonparametric permutation test).
Interestingly, the excess reward site bias of RR units was greater on the SWM task than on the linear track (excess reward site bias for RR units on the SWM task 0.027 ± 0.014; excess reward site bias for RR units on the linear track 0.017 ± 0.015; p=0.045, nonparametric permutation test).
The contrast of excess reward site bias of RR units and nonRR units was significant (p<0.016, nonparametric permutation test; see Materials and methods).
However, this preferential coordination was not observed on the linear track (excess reward site bias for RR units at centrifugal replay on the linear track: 0.040 ± 0.024; p=0.11, chi = 2.5; excess reward site bias for RR units at centripetal replay: -0.004 ± 0.031; p=0.9, chi = 0.01).
Similarly, the excess reward site bias of RR units at replay events was no greater immediately prior to correct trials than prior to error trials (excess reward site bias prior to correct trials 0.001 ± 0.021; prior to error trials 0.013 ± 0.041, p=0.6, nonparametric permutation test).
Consistent with the latency of the SPW-R-associated VTA potential, the excess reward site bias of RR units was maximal at a 75 ms VTA lag relative to hippocampal activity (data not shown).
We did not detect a selective engagement of RR units with reward locations of reverse replay over forward replay (excess reward site bias for RR units at reverse replay 0.035 ± 0.020; at forward replay 0.034±0.020; p=1, nonparametric permutation test).
