Ice accretion on solid surfaces, which is an omnipresent phenomenon in daily life, presents a variety of challenges for applications ranging from infrastructures (such as transmission lines and roads) to industrial manufacturing (such as aviation, wind turbines, boat hulls, and refrigeration cycles). For example, ice accreting on aerodynamic surfaces such as wings of aircraft or blades of wind turbines changes the shape of the surfaces, leading to power loss or even to mechanical failure.

In the past decades, strategies to prevent ice accretion are mainly based on de-icing techniques, characterized as removing ice from surfaces, via heating, chemistry, mechanics, etc., which is cumbersome and energy-consuming. Recently, driven by the progress in material science and micro/nanofabrication technologies, new strategies have been discovered to shed light on icing problems. All these new strategies could be classified into three categories: 1. Self-bouncing: facilitating the instant removal of water imparted by the high mobility of water before freezing; 2. Anti-icing: preventing ice nucleation of water deposited on surfaces; 3. Icephobicity: significantly reducing adhesion strength between ice and the contacted surface when icing is inevitable.

The self-bouncing strategy aims to shed water away from surfaces before water freezes. However, due to the omnipresent of water vapor under cold weather, which will condensate on the nanotexture of superhydrophobic surfaces and results in wetting and then icing within the nanotexture, the droplet is not capable to bounce away from surfaces, instead, droplet will stick on surfaces. Hence, the anti-icing strategy, which can prevent the ice nucleation of droplet on droplet-substrate interface, is widely studied to compensate for the problem when droplets stick and then freeze on surfaces. Ideally, anti-icing strategy is capable of preventing droplet from icing. However, realistically, pervasive dust, impurities, even minor changes in environmental conditions can trigger ice nucleation, which makes ice nucleation inevitable. Hence, icephobic strategy is produced to significantly reduce adhesion strength between ice and the contacted surface when icing is inevitable so that ice can be shed readily by self-weight, wind, or even small vibrations. However, self-removal of ice with no external energy input (such as wind, heat, or vibration) and instantly shed freezing droplets away, before they fully freeze and subsequently form an integral ice film, are of critical importance under harsh conditions, specifically for aerodynamic surfaces such as aircraft and wind turbines which demand ice-free performance during the whole working time.

Herein, inspired by the shooting mechanism found in nature, we created an elastic macro-pillar (compared to droplets) architecture which enables a water droplet to shoot itself away from the substrate. To comprehensively understand the influencing factors of the self-shooting phenomenon and subsequently to rationally design the self-shooting architectures, we studied the influence of four variables (pillar height, original droplet volume, contact radius, and material stiffness) on the self-shooting phenomenon of freezing droplets and found the design criterion of self-shooting architecture. Apart from these, we also found that the self-shooting phenomenon of freezing droplets is characterized by three steps: 1. Energy is generated when a water droplet freezes inward from two interfaces (air-water interface and water-substrate interface) and forms an integral spheroidal ice shell; 2. The generated energy is stored as elastic energy via downward, axial, elastomer deformation of the pillar-substrate system; The elastic energy is released after freezing droplet detaches from the substrate and transferred into kinetic energy of freezing droplet when the pillar-substrate system recovers to its original shape. This self-shooting mechanism of freezing droplets is found identical to the shooting mechanism found in fungi and plants. Our study demonstrates an artificial spontaneously shooting mechanism that has potential relevance to fields such as soft robotics, self-cleaning, and condensation.