to be performed over extended periods and potentially forming the foundation for a nanoscale assembly tool. The optical trapping of nanoparticles faces the challenge however that the gradient force varies approximately with the nanoparticle volume. [6][7][8] This means, for example, that reducing the nanoparticle radius by a factor of ten reduces the gradient force by a factor of one thousand. As the name suggests, this force also varies with the gradient of the intensity. One can therefore boost the gradient force by concentrating light as tightly as possible at the trapping site. Conventional optical tweezers however use lenses to focus light and are thus subject to the diffraction limit. This sets a limit on the gradient force on a given nanoparticle that can be obtained with a given laser power with conventional optical tweezers. [9] Optical nanostructures enable optical fields to be concentrated into deeply-sub wavelength regions, thereby presenting a means to surpass the performance of traditional optical tweezers for the trapping of nanoparticles. Optical tweezers based on nanostructures are sometimes known as optical nanotweezers. Plasmonic apertures have proven highly effective, [10][11][12][13][14][15][16][17] with the enhanced fields inside the apertures producing strong gradient forces. [18] The presence of heating is an Plasmonic apertures permit optical fields to be concentrated into subwavelength regions. This enhances the optical gradient force, enabling the precise trapping of nanomaterials such as quantum dots, proteins, and DNA molecules at modest laser powers. Double nanoholes, coaxial apertures, bowtie apertures, and other structures have been studied as plasmonic nanotweezers, with the design process generally comprising intuition followed by electromagnetic simulations with parameter sweeps. Here, instead, a computational algorithm is used to design plasmonic apertures for nanoparticle trapping. The resultant apertures have highly irregular shapes that, in combination with ring couplers also optimized by algorithm, are predicted to generate trapping forces more than an order of magnitude greater than those from the double nanohole design used as the optimization starting point. The designs are realized by fabricating precision apertures with a helium/neon ion microscope and are studied them by cathodoluminescence and optical trapping. It is shown that, at every laser intensity, the algorithm-designed apertures can trap particles more tightly than the double nanohole.