Two-dimensional (2D) van der Waals (vdW) magnetic materials with high coercivity and high $T_\text{C}$ are desired for spintronics and memory storage applications. Fe$_3$GeTe$_2$ (F3GT) is one such 2D vdW ferromagnet with a reasonably high $T_\text{C}$, but with a very low coercive field, $H_\text{c}$ ($\lesssim$100~Oe). Some of the common techniques of enhancing $H_\text{c}$ are by introducing pinning centers, defects, stress, doping, etc. They involve the risk of undesirable alteration of other important magnetic properties. Here we propose a very easy, robust, and highly effective method of phase engineering by altering the sample growth conditions to greatly enhance the intrinsic coercivity (7-10 times) of the sample, without compromising its fundamental magnetic properties ($T_\text{C}\simeq$210K). The phase-engineered sample (F3GT-2) comprises of parent F3GT phase with a small percentage of randomly embedded clusters of a coplanar FeTe (FT) phase. The FT phase serves as both mosaic pinning centers between grains of F3GT above its antiferromagnetic transition temperature ($T_\text{C1}\sim$70~K) and also as anti-phase domains below $T_\text{C1}$. As a result, the grain boundary disorder and metastable nature are greatly augmented, leading to highly enhanced coercivity, cluster spin glass, and meta-magnetic behavior. The enhanced coercivity ($\simeq$1~kOe) makes F3GT-2 much more useful for memory storage applications and is likely to elucidate a new route to tune useful magnetic properties. Moreover, this method is much more convenient than hetero-structure and other cumbersome techniques.
Comment: 12 pages, 11 figures