The Majorana representation of spin
1/2 could be useful in the theory of strongly correlated electrons because it respects isotropy in spin space and has no constraints. To judge its reliability, we apply it to the
O(N) quantum Heisenberg antiferromagnet for general
N. At $N\ensuremath{\rightarrow}\ensuremath{\infty}$ we find a resonant valence bond-like spin-liquid ground state in precise agreement with previous results from an SU(\ensuremath{\infty}) extrapolation. At
N=3, however, the representation fails and gives an incorrect gap in the
O(3) Heisenberg chain. We trace this failure at
N=3 to (i) a redundancy of states and (ii) the presence of a local gauge invariance both of which are intrinsic properties of the Majorana representation of spins at any
N but irrelevant at $N=\ensuremath{\infty}.$ Specifically at
N=3, there is an unphysical ``longitudinal'' mode. A correct treatment of these difficulties must precede applications of the Majorana representation of spin-
1/2 operators to systems of physical interest.
We consider the properties of finite isotropic antiferromagnetic Heisenberg chains with
S=1/2,1,3/2 spins when a weak magnetic field is applied on a few sites, using White's density-matrix renormalization-group method. For the
S=1 chain there exists only one length scale in the system which determines the behavior of the one- and two-point correlation functions both around the local perturbation and near the free boundary. For the critical half-odd-integer spin cases the exponent of the spin-spin correlation function was found to be $\ensuremath{\eta}=1$ and the exponent of the decay of the site magnetization around the perturbed site is ${x}_{m}=\ensuremath{\eta}/2.$ Close to a free boundary, however, the behavior is completely different for
S=1/2 and
S>1/2.We consider the isotropic
S=1 Heisenberg chain with a finite Haldane gap $\mathrm{\ensuremath{\Delta}}$ and use state-of-the-art numerical techniques to investigate its dynamical properties at finite temperature, focusing on the nuclear spin-lattice relaxation rate
1/T1 measured in nuclear magnetic resonance (NMR) experiments, for instance. In particular, we analyze the contributions from modes with momenta close to $q\ensuremath{\approx}0$ and $q\ensuremath{\approx}\ensuremath{\pi}$ as a function of temperature. At high-temperature we observe spin diffusion, while at low-temperature we argue that a simple activated behavior $1/{T}_{1}\ensuremath{\propto}exp(\ensuremath{-}\mathrm{\ensuremath{\Delta}}/T)$ can be observed only at temperatures much smaller than the gap $\mathrm{\ensuremath{\Delta}}$.