Halvledare, material som leder ström halvbra, är grunden för dagens teknik. Utan dem skulle vi inte ha några datorer, smarta telefoner, mikrovågsugnar, självkörande bilar eller talande tvättmaskiner. Chip med komponenter så små att de inte går att se med ett vanligt mikroskop, utgör den både minsta och viktigaste beståndsdelen i hela den digitala industrin. När efterfrågan på olika digitala verktyg snabbt steg under pandemin ledde det till en global brist på halvledare, vilket blev en ögonöppnare för den moderna teknikens sårbarhet. --- I Halvledaråldern förklarar kemisten Henrik Pedersen pedagogiskt vad en halvledare faktiskt är, och hur mänskligheten har lyckats nå fram till denna svindlande precisa nanoteknik som förenar vårt samlade kunnande kring kemi, fysik och elektronik. --- Om författaren: Henrik Pedersen är professor i oorganisk kemi vid Linköpings universitet. Hans forskning bedrivs i gränslandet mellan kemi, fysik, materialvetenskap och elektronik. Halvledaråldern är hans första populärvetenskapliga bok.

Semiconductor devices are constructed from stacks of materials with different electrical properties, making deposition of thin layers central in producing semiconductor chips. The shrinking of electronics has resulted in complex device architectures which require deposition into holes and recessed features. A key parameter for such deposition is the step coverage (SC), which is the ratio of the thickness of material at the bottom and at the top. Here, we show that adding a co-flow of a heavy inert gas affords a higher SC for deposition by chemical vapor deposition (CVD). By adding a co-flow of Xe to a CVD process for boron carbide using a single source precursor with a lower molecular mass than the atomic mass of Xe, the SC increased from 0.71 to 0.97 in a 10:1 aspect ratio feature. The concept was further validated by a longer deposition depth in lateral high aspect ratio structures. We suggest that competitive co-diffusion is a general route to conformal CVD.

Nanoscale hybrid inorganic-organic multilayers are attractive for accessing emergent phenomena and properties through superposition of nanomolecularly-induced interface effects for diverse applications. Here, we demonstrate the effects of interfacial molecular nanolayers (MNLs) of organo-diphosphonates on the growth and stability of titania nanolayers during the synthesis of titania/MNL multilayers by sequential atomic layer deposition and single-cycle molecular layer deposition. Interfacial organo-diphosphonate MNLs result in similar to 20-40% slower growth of amorphous titania nanolayers and inhibit anatase nanocrystal formation from them when compared to amorphous titania grown without MNLs. Both these effects are more pronounced in multilayers with aliphatic backbone-MNLs and likely related to impurity incorporation and incomplete reduction of the titania precursor indicated by our spectroscopic analyses. In contrast, both MNLs result in two-fold higher titania nanolayer roughness, suggesting that roughening is primarily due to MNL bonding chemistry. Such MNL-induced effects on inorganic nanolayer growth rate, roughening, and stability are germane to realizing high-interface-fraction hybrid nanolaminate multilayers.

Aluminum nitride (AlN) is a semiconductor with a very wide band gap and a potential dielectric material. Deposition of thin AlN films is routinely done by several techniques, including atomic layer deposition (ALD). In this study, we deposited AlN using ALD with trimethylaluminum (TMA) as the Al precursor and ammonia (NH3) with and without plasma activation as the N precursor in the temperature range from 100 to 400 degrees C while monitoring the surface reactions using mass spectrometry. Our results, combined with recent quantum chemical modelling, suggest that the surface chemistry of the deposition process is chemisorption of TMA followed by reductive elimination of the methyl groups to render mono methyl aluminum species. The NH3 chemisorption is done by ligand exchange to form CH4 and an -NH2 terminated surface.

Chemical vapor deposition of indium nitride (InN) is severely limited by the low thermal stability of the material, and, thus, low-temperature deposition processes such as atomic layer deposition (ALD) are needed to deposit InN films. The two chemically and structurally closely related materials—aluminum nitride and gallium nitride (GaN)—can be deposited by both plasma and thermal ALD, with ammonia (NH₃) as a nitrogen precursor in thermal processes. InN, however, can only be deposited using plasma ALD, indicating that there might be a limitation to thermal ALD with NH₃ for InN. We use quantum-chemical density functional theory calculations to compare the adsorption process of NH₃ on GaN and InN to investigate if differences in the process could account for the lack of thermal ALD of InN. Our findings show a similar reactive adsorption mechanism on both materials, in which NH₃ could adsorb onto a vacant site left by a desorbing methyl group from the surfaces. The difference in energy barrier for this adsorption indicates that the process is many magnitudes slower on InN compared to GaN. Slow kinetics would hinder NH3 from reactively adsorbing onto InN in the timeframe of the ALD growth process and, thus, limit the availability of a thermal ALD process.

SiC multilayer coatings were deposited via thermal chemical vapor deposition (CVD) using silicon tetrachloride (SiCl₄) and various hydrocarbons under identical growth conditions, i.e. at 1100 °C and 10 kPa. The coatings consisted of layers whose preferred growth orientation alternated between random and highly 〈111〉-oriented. The randomly oriented layers were prepared with either methane (CH₄) or ethylene (C₂H₄) as carbon precursor, whereas the highly 〈111〉-oriented layers were grown utilizing toluene (C₇H₈) as carbon precursor. In this work, we demonstrated how to fabricate multilayer coatings with different growth orientations by merely switching between hydrocarbons. Moreover, the success in depositing multilayer coatings on both flat and structured graphite substrates has strengthened the assumption proposed in our previous study that the growth of highly 〈111〉-oriented SiC coatings using C₇H₈ was primarily driven by chemical surface reactions.

Boron carbide in its rhombohedral form (r-BxC), commonly denoted B4C or B13C2, is a well-known hard material, but it is also a potential semiconductor material. We deposited r-BxC by chemical vapor deposition between 1100 degrees C and 1500 degrees C from triethylboron in H-2 on 4H-SiC(0001) and 4H-SiC(0001). We show, using ToF-ERDA, that pure B4C was grown at 1300 degrees C, furthermore, using XRD that graphite forms above 1400 degrees C. The films deposited above 1300 degrees C on 4H-SiC(0001) were found to be epitaxial, with the epitaxial relationships B4C(0001)[1010]||4H-SiC(0001)[1010] obtained from pole figure measurements. In contrast, the films deposited on 4H-SiC(0001) were polycrystalline. We suggest that the difference in growth mode is explained by the difference in the ability of the different surfaces of 4H-SiC to act as carbon sources in the initial stages of the film growth.

Indium oxide (In2O3) is an important transparent conducting material widely used in optoelectronic applications. Herein, we study the deposition of In2O3 by thermal atomic layer deposition (ALD) using our recently reported indium(iii) triazenide precursor and H2O. A temperature interval with self-limiting growth was found between similar to 270 and 385 degrees C with a growth per cycle of similar to 1.0 angstrom. The deposited films were polycrystalline cubic In2O3 with In : O ratios of 1 : 1.2, and low levels of C and no detectable N impurities. The transmittance of the films was found to be >70% in visible light and the resistivity was found to be 0.2 m omega cm. The high growth rates, low impurities, high optical transmittance, and low resistivity of these films give promise to this process being used for ALD of In2O3 films for future microelectronic displays.

In recent decades, indium nitride (InN) has been attracting a great deal of attention for its potential applicability in the field of light-emitting diodes (LEDs) and high-frequency electronics. However, the contribution from adsorption- and reaction- related processes at the atomic scale level to the InN growth has not yet been unveiled, limiting the process optimization that is essential to achieve highly crystalline and pure thin films. In this report, we investigate the reaction pathways that are involved in the crystal growth of InN thin film in atomic layer deposition (ALD) techniques from trimethylindium (TMI) and ammonia (NH3) precursors. To accomplish this task, we use a solid-state approach to perform the ab-initio calculations within the Perdew–Burke–Ernzerhof functional (PBE) level of theory. The results clarify the activation role from the N-rich layer to decrease the barrier for the first TMI precursor dissociation from Δ‡H= +227 kJ/mol, in gas phase, to solely +16 kJ/mol, in the surface environment. In either case, the subsequent CH3 release is found to be thermo- and kinetically favored with methylindium (MI) formed at the hcp site and ethane (C2H6) as the byproduct. In the following step, the TMI physisorption at a nearby occupied hcp site promotes the sequential hydrogen removal from the N-rich layer at the minimum energy cost of Δ‡H < +105 kJ/mol with methane (CH4) release. An alternative mechanism involving the production of CH4 is also feasible upon dissociation in gas phase. Furthermore, the high concentration of CH3 radicals, from precursor dissociation, might be the origin of the carbon impurities in this material under the experimental conditions of interest. Finally, the passivation methodology is not found to affect the evaluation of the surface-related processes, whereas the inclusion of spin-polarization is demonstrated to be essential to the proper understanding of the reaction mechanism.

Area selective deposition (ASD) of films only on desired areas of the substrate opens for less complex fabrication of nanoscaled electronics. We show that a newly developed CVD method, where plasma electrons are used as the reducing agent in deposition of metallic thin films, is inherently area selective from the electrical resistivity of the substrate surface. When depositing iron with the new CVD method, no film is deposited on high-resistivity SiO2 surfaces whereas several hundred nanometers thick iron films are deposited on areas with low resistivity, obtained by adding a thin layer of silver on the SiO2 surface. On the basis of such a scheme, we show how to use the electric resistivity of the substrate surface as an extension of the ASD toolbox for metal-on-metal deposition.