Modern materials science relies on a deep understanding of defects — interruptions to regular atomic arrangements in crystalline solids. Although ‘defects’ brings to mind imperfections and blemishes, they often make a material more useful than it otherwise would be. For example, metal impurities such as chromium and iron atoms in corundum (a crystalline form of aluminium oxide) are responsible for the colours of rubies and sapphires. Moreover, the addition of impurities to silicon has enabled the current era of computing and robotics. Writing in Nature, Du et al.1 report a method for producing a variety of technologically useful two-dimensional materials that contain deliberately introduced impurities, solving a fabrication problem for next-generation devices.
Transition-metal chalcogenides (TMCs) are emerging materials that hold great promise for their incorporation into a wide range of applications, from batteries and flexible electronics to biosensors and water-purification systems. They are composed of a transition metal such as molybdenum or tungsten and a chalcogen (an element in group 16 of the periodic table) such as sulfur, selenium or tellurium. The properties of TMC monolayers change greatly if the metallic element is altered. In particular, these structures can change from being normal metals to semiconductors, or even superconductors.
In the past few years, many researchers2–4 have focused on making ultrathin electronics that have superior properties to those of existing silicon devices, by combining different TMC monolayers into a single object known as a heterostructure, using a technique called chemical-vapour deposition. Other researchers5 have produced functional devices using a single TMC in which different regions of the material have different properties, such as being metallic or semiconducting. However, although these techniques are good for fabricating prototype devices, they are not practical enough for real-world applications.
The long-standing problem in incorporating TMC monolayers into a functional device has been the lack of a metallic-phase TMC monolayer that is stable in ambient conditions for more than a month6. Du and colleagues overcame this challenge, and made metallic-phase TMC monolayers that they show can exist in such conditions for about a year. The authors achieved this feat by introducing a technology based on a process known as doping.
Doping has shaped the digital revolution — the shift from analog to digital electronics that began in the second half of the twentieth century. The process involves changing the electrical conductivities of semiconductors such as silicon by adding impurities. Eighty years ago7, dopant atoms of boron and phosphorus were added to pure silicon to produce materials called p-type and n-type silicon, respectively; these form p–n junctions, the basis of computing. This doping technology continues to be useful today, and is found in our everyday electronics. Du and co-workers’ doping technology for 2D materials is also expected to have a long-term impact on the field.
The authors produced TMC monolayers in three steps (Fig. 1). First, they prepared a crystal that contained two different transition metals (one of which provided impurity atoms for TMC doping), an element in group 13 or 14 of the periodic table, and carbon. Second, they heated the crystal at high temperatures (873–1,373 kelvin) for 4 hours in an environment that contained two gases. One of these was a chalcogen-containing gas that supplied chalcogen atoms for the TMC; the other gas was phosphorus, which provided further impurity atoms for TMC doping. Third, the authors used a process called liquid exfoliation to convert the resulting TMC crystal into TMC monolayers in the form of liquid inks.
Du et al. used this three-step dual-doping technology to make, for example, metallic-phase TMC monolayers of tungsten disulfide that were doped with both yttrium and phosphorus atoms. They also produced undoped TMC monolayers by preparing layered crystals that contained one type of transition metal, rather than two, and removing the source of phosphorus gas. In total, the authors made six doped and seven undoped TMC monolayers, demonstrating the remarkable versatility of their approach for producing 2D materials.
One major advantage of Du and colleagues’ method is that the final 2D materials are in the form of liquid inks. There is clearly a shift in this field towards making high-quality monolayer inks for commercialization8,9, rather than films produced by techniques such as epitaxial growth or chemical-vapour deposition. Such films require a process known as delamination to separate them from their growth substrates, which deteriorates the material’s quality and necessitates further processing10,11. By contrast, monolayer inks can be readily deposited on arbitrary substrates using techniques such as inkjet printing or spin coating, and so are easily integrated into 3D systems12,13.
From a scientific standpoint, 2D materials need to be stable and usable in our immediate surroundings. Du and colleagues’ findings are promising for the field because they show that the presence of a low quantity (less than 1%) of impurity atoms can stabilize TMC monolayers. This result suggests that materials researchers should start to explore the use of chemical elements to stabilize 2D materials that would otherwise degrade in ambient conditions within hours, rather than using encapsulation layers, which complicate the monolayer systems.
The next steps will be for theorists to predict suitable ‘impurity stabilizers’ for TMC monolayers, and for experimentalists to investigate the use of elements that are abundant on Earth. In the meantime, it should still be possible to build advanced machines for precise and reliable dual doping of TMCs, because only a low quantity of relatively rare yttrium and phosphorus is needed to stabilize TMC monolayers. Du and colleagues’ work demonstrates that, whatever new materials are discovered, it is crucial that we understand, manipulate and use their atomic-level defects. Every atom matters.
Nature 577, 477-478 (2020)