Incorporation of arsenic dopants into semiconductor lattices

Abstract

1. A method (100) for incorporating arsenic dopant atoms (5) at defined locations in a semiconductor lattice (1), comprising: (i) forming (102) a passivation layer (9) on a surface (7) of the semiconductor lattice (1); (ii) selectively removing (104) the passivation layer (9) at one or more incorporation locations (15) to expose the surface (7) of the semiconductor lattice (1); and (iii) exposing (106) the incorporation locations (15) to a dopant precursor gas (17) to adsorb dopant precursor substructures (17) comprising arsenic atoms (5) onto the surface (7) of the semiconductor lattice (1) at at least some of the incorporation locations (15), wherein the arsenic dopant atoms (5) are incorporated into the lattice (1) at the incorporation locations (15).

Description

Detailed Description of the Invention

[0001] [Background technology] The present invention relates to the incorporation of arsenic dopant atoms within the surface of a semiconductor lattice, and more particularly, to the precise placement of arsenic atoms in a scalable manner.

[0002] Due to the significantly enhanced computational power offered by quantum computers, the goal of developing quantum computers is attracting increasing attention.

[0003] In a quantum computer, quantum bits (also known as qubits) can encode information. Qubits are similar to classical bits in that they can take on values ​​of 0 or 1, but differ from classical bits in that they can take on both values ​​simultaneously (in a state known as superposition). The computational power of a quantum computer lies in this superposition.

[0004] A qubit is implemented in a two-state quantum mechanical system, with the states corresponding to 0 and 1. For example, a qubit may be implemented by capturing a photon, capturing an electron, controlling electron spin, or controlling nuclear spin.

[0005] One promising candidate for implementing qubits is to use the electron spin of dopant atoms placed in a semiconductor host. In order to manipulate and control the qubits, the dopant atoms must be placed in precise locations. To date, development has focused on the placement of phosphorus atoms in a (001) silicon surface. The phosphorus / silicon system has been chosen because phosphorus has one extra valence electron compared to silicon, creating a spin 1 / 2 hydrogen defect state that can be used for quantum computing. Phosphorus atoms also have a similar size to silicon atoms, and thus are easily incorporated. Silicon is a well-understood system that allows for the relative ease of incorporation of phosphorus qubit structures with conventional silicon-based electronics.

[0006] WO2019 / 210370 discloses a method of scanning tunneling microscope (STM) hydrogen desorption lithography for incorporating phosphorus atoms in silicon. The method includes the steps of forming one or more lithography sites on a surface portion using an STM tip, doping the surface portion with phosphine gas at a temperature below 100K so that a portion of the molecule binds to the surface portion, and incorporating one or more phosphorus atoms in each of the lithography sites by transferring an amount of energy to the dopant atoms. The energy can be transferred by thermal annealing or by transferring energy through the STM tip. The number of dopant atoms incorporated in the lithography sites is deterministic and related to the size of the lithography sites.

[0007] Current methods for incorporating phosphorus dopant atoms do not allow for precise placement of the atoms in controlled locations, and furthermore, the dopant atoms are typically incorporated in only about 70% of the locations, making this process unsuitable for large-scale commercial production.

[0008] Additionally, dopant atoms can become incorporated in unwanted locations due to imperfections in the formation of lithographic sites.

[0009] A need exists for a scalable method of precisely placing single dopant atoms within the surface of a semiconductor lattice to form qubits for a quantum computer.

[0010] According to one aspect of the invention, there is provided a method for incorporating arsenic dopant atoms at defined locations in a semiconductor lattice, comprising the steps of: (i) forming a passivation layer on a surface of a semiconductor grating; (ii) selectively removing the passivation layer at one or more incorporation locations to expose the surface of the semiconductor lattice; (iii) exposing the incorporation sites to a dopant precursor gas to cause dopant precursor moieties comprising arsenic atoms to adsorb onto the surface of the semiconductor lattice at at least some of the incorporation sites; Arsenic dopant atoms are incorporated into the lattice at the incorporation sites; A method is provided.

[0011] The step of exposing the surface to a dopant precursor gas may be carried out at room temperature.

[0012] The semiconductor lattice may be a group IV semiconductor lattice. The lattice may have a (001) surface. For example, the semiconductor lattice may be a silicon lattice, a germanium lattice, or a diamond cubic structure elemental semiconductor.

[0013] The method may comprise subjecting the surface layer to a thermal anneal to incorporate arsenic dopant atoms into the semiconductor lattice.

[0014] For a germanium lattice, upon adsorption of the dopant precursor moiety, the dopant precursor molecule may dissociate and the arsenic atom may be incorporated into the semiconductor lattice in a manner that does not require energy transfer.

[0015] The method may further comprise overgrowing a surface of the semiconductor lattice after the arsenic dopant atoms are incorporated.

[0016] The area of ​​each incorporation location may be up to four semiconductor atoms, or up to six semiconductor atoms. The area of ​​each incorporation location may be four semiconductor atoms and the arsenic dopant atom may be incorporated in 100% of the incorporation location. The area of ​​each incorporation location may be a semiconductor dimer and the arsenic dopant atom may be incorporated in 80% of the incorporation location. The area of ​​each incorporation location may be a single semiconductor atom and the arsenic dopant atom may be incorporated in about 10% of the incorporation location.

[0017] The method includes, before incorporating the arsenic atom, (a) Identifying incorporation sites where no arsine adsorption occurs and where atoms of the passivation layer are re-adsorbed; (b) selectively removing the passivation layer to expose the identified incorporation locations; (c) exposing the identified incorporation sites to a dopant precursor gas to adsorb dopant precursor moieties onto the surface of the semiconductor lattice at the identified incorporation sites; The device may further include:

[0018] The method is: Repeating steps (a) to (c) in an iterative manner at least once; (d) incorporating arsenic atoms into the semiconductor lattice; The device may further include:

[0019] The method may further comprise, between steps (i) and (ii), heating the passivation layer to induce diffusion of defects in the passivation layer so that the defects pair up and cluster, and exposing the passivation layer to an additional precursor gas comprising additional precursor molecules that contain the same atoms as the passivation layer to fill holes in the passivation layer.

[0020] The additional precursor gas may comprise the same atoms as the passivation layer plus phosphorus atoms. The additional precursor gas may comprise phosphine PH3.

[0021] The passivation layer may be heated to at least 200° C. to induce diffusion of holes in the passivation layer.

[0022] The method may further comprise forming a second passivation layer on the surface of the semiconductor lattice; selectively removing the second passivation layer at one or more second incorporation locations to expose the surface of the semiconductor lattice; and exposing the second incorporation locations to a second dopant precursor gas to adsorb second dopant precursor substructures comprising atoms of a second dopant, different from arsenic, onto the surface of the semiconductor lattice at at least some of the second incorporation locations, where the atoms of the second dopant are incorporated into the lattice at the second incorporation locations.

[0023] The step of selectively removing the passivation layer at one or more incorporation locations to expose a surface of the semiconductor lattice may be performed by one of scanning tunneling microscopy and electron beam microscopy. The step of selectively removing the passivation layer at one or more incorporation locations to expose a surface of the semiconductor lattice may be performed by an array of scanning tunneling microscopy probes.

[0024] The method may further comprise annealing the surface of the semiconductor lattice to form an atomically flat surface prior to forming the passivation layer on the surface of the semiconductor lattice.

[0025] The passivation layer may comprise atomic hydrogen.

[0026] The dopant precursor gas may comprise arsine AsH3.

[0027] The energy barrier for diffusion and / or desorption of atoms of the passivation layer may be lower than the energy barrier for desorption of the precursor substructures from the surface of the semiconductor lattice.

[0028] Various aspects of the invention can provide placement of dopant arsenic atoms at approximately 100% of lithographic locations, where only one dopant arsenic atom is incorporated. Dopant arsenic atoms can be placed with greater precision than phosphorus atoms and are incorporated more quickly than other dopants. This suggests that embodiments of the invention provide, for the first time, a process that is scalable to a commercial scale for fabricating many qubits.

[0029] The use of arsenic in embodiments allows the formation of two qubits using 1 / 2 electron spin and a qudit (with more than two states) using 3 / 2 nuclear spin.

[0030] The embodiment including a stage of exposing the passivation layer to phosphine (or other suitable precursor gas) ensures that defects in the passivation layer are repaired, so that dopant arsenic atoms cannot be placed in undesired locations, as could be used with other dopants, since phosphine will desorb from smaller defect sites in the passivation layer rather than being incorporated into the lattice.

[0031] The inventors have found that embodiments using germanium can be implemented to incorporate dopant atoms without the need for annealing or other energy transfer, which reduces the likelihood of detrimental desorption of donors, a key limiting factor in scaling up to large numbers of qubits for phosphorus in silicon, and also reduces the risk of incorporated dopant atoms migrating away from their desired locations.

[0032] The inventors have also surprisingly discovered that in embodiments where the lithographic location is no more than a single silicon atom, performing an iterative process can allow for precise placement of a single dopant atom at the single atomic location.

[0033] Embodiments of the present invention will now be described, by way of example only, with reference to the accompanying drawings in which: [Brief description of the drawings]

[0034] [Figure 1] 1 shows a flow diagram of a method for embedding dopant atoms into a surface of a semiconductor lattice. [Figure 2A-2C] 2A-2D are schematic illustrations of the surface of a semiconductor grating at corresponding steps of the method shown in the flow diagram of FIG. 1; [Fig. 2D-2F] 2A-2D are schematic illustrations of the surface of a semiconductor grating at corresponding steps of the method shown in the flow diagram of FIG. 1; [Figure 2G] 1 shows the dissociation of an arsine molecule on a free silicon surface. [Figure 3A-3B] 2A-2D show scanning tunneling microscope images of the surface of a grating at certain steps of the method illustrated in the flow diagram of FIG. 1. [Figure 3C] 2A-2D show scanning tunneling microscope images of the surface of a grating at certain steps of the method illustrated in the flow diagram of FIG. 1. [Figure 4] 1 shows a flow diagram of an alternative method for embedding dopant atoms into a surface of a semiconductor lattice. [Diagram 5] 1 shows a flow diagram of a method for forming an atomic passivation layer. [Figure 6A-6C] 6A-6C are schematic diagrams illustrating the surface of a passivation layer at corresponding steps of the method of FIG. 5. [Figure 7] 1 shows an example of an STM image of a Ge(001) surface formed at room temperature and doped with arsine. [Figure 8] An example of an STM image of the exchange of germanium adatoms between two As-Ge-H heterodimer features in the surface shown in FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0035] Figure 1 shows a first method 100 of providing a single dopant atom 5 into a lattice 1 of semiconductor atoms 3. Figures 2A-F show schematic views of a surface 7 of the lattice 1 at different steps, as discussed below. The upper rows of Figures 2A-F show a portion of the lattice 1 in top view, and the lower rows show a portion of the lattice in side view.

[0036] 1, each dopant atom 5 replaces a semiconductor atom 3 in the lattice 1. As discussed below, each dopant atom 5 is provided at a well-defined location and is sufficiently spaced from the other dopant atoms 5 so that the spin state of one dopant atom 5 does not affect the spin state of another dopant atom 5. Each dopant atom can be manipulated by conductive contacts (not shown) without the electric fields of the conductive contacts affecting each other.

[0037] In one example, as discussed below, arsenic dopant atoms 5 are incorporated into the (100) surface of a crystalline silicon lattice 1 .

[0038] Prior to doping, the surface 7 of the silicon grating 1 is cleaned. Various methods of cleaning the surface of the semiconductor grating 1 will be known to those skilled in the art. This may include cycles of sputtering and annealing, with thermal annealing repeated with cooling in between, and cleaning with acetone and / or isopropyl alcohol. Figure 2A shows the prepared grating 1.

[0039] Cleaning was performed under vacuum conditions (<5×10 -10 In addition, steps 102, 104, 106, 108, and 110 discussed below are performed under vacuum conditions and at room temperature (e.g., between 15° C. and 30° C.).

[0040] 2A-2F show the semiconductor atoms 3, dopant atoms 5, and hydrogen atoms 11 as being the same size, it will be appreciated that this is not the case and the figures are merely schematic. However, in terms of the bonds formed, the ratio of semiconductor atoms 3 in the surface layer to hydrogen atoms 11 in the passivation layer is 1:1 because each silicon atom 3 on the surface 7 has a single dangling bond (not shown) that is terminated with a single hydrogen atom 11.

[0041] The first step 102 of the method 100 (see FIG. 2B) is to form a passivation layer 9 on the surface 7 of the grating 1. The passivation layer 9 is formed of hydrogen atoms 11 and has a thickness of a single atomic layer. Various methods for forming such a hydrogen passivation layer are known to those skilled in the art.

[0042] In a second step 104 of the method (FIG. 2C), a scanning tunneling microscope (STM) tip 13 is used to selectively remove hydrogen atoms 11 from the passivation layer 9 to expose the surface 7 of the grating 1. The removed atoms 11 define incorporation sites 15 where dopant atoms 5 are to be incorporated.

[0043] The precise position of the STM tip 13 can be controlled via the STM control routine. By controlling the location where the incorporation site 15 is formed, the location of the dopant atom 5 can be controlled.

[0044] The size of the incorporation sites 15 can vary. For example, an incorporation site 15 may have an area of ​​four silicon atoms 3, preferably arranged in a square (i.e., two adjacent silicon dimers) to provide four dangling bonds. Alternatively, each incorporation site may be the size of a pair of silicon atoms 3 (one silicon dimer or two silicon dangling bonds) or a single silicon atom 3 (single dangling bond).

[0045] In Figure 2C, the examiner shows an incorporation site having an area of ​​two silicon atoms 3. Figure 3A illustrates a scanning tunneling microscope (STM) image showing four incorporation sites 15 having an area of ​​two silicon atoms 3 formed in the passivation layer 9.

[0046] In a third step 106, the incorporation site 15 is exposed to arsine gas AsH317. -9 The exposure to arsine gas is for 5 minutes at 1000 psi (mbar), although it will be appreciated that this is by way of example only and that any suitable range of pressures and times may be used.

[0047] At the incorporation site 15, various AsH x Portions 17' (where x=0-3) are adsorbed. Passivation layer 9 inhibits and prevents portions 17' from adsorbing away from incorporation sites 915. FIG. 3B shows AsH adsorbed onto silicon surface 7 at incorporation sites 15 shown in FIG. 3A. x An STM image of portion 17' is shown.

[0048] The moiety 17 ′ that is adsorbed at the entrapment site 15 will depend on the size of the entrapment site 15 .

[0049] The AH3 molecule 17 reacts with different AsH x AsH dissociates from moiety 17' x The moiety 17' bonds to a silicon atom 3 on the surface 7. Depending on the AsHx moiety 17', the arsenic atom 5 may either bond to a single silicon atom 3 at the incorporation site 15 (referred to as a terminal configuration) or may bond across two silicon atoms 3 (referred to as a bridge configuration). In an incorporation site 15 where the AsHx moiety 17' is bonded to a single silicon atom 3, a dangling bond at the other silicon atom 3 at the incorporation site 15 bonds to a hydrogen 11 from a dissociated arsine molecule 17. The table below shows the percentage of dangling bond terminations at incorporation sites 15 of different sizes. [Table 1] When x=1, the AsH moieties 17' are attached in a terminal configuration. When x=2, there is a mixture of moieties 17' attached in terminal and bridge configurations.

[0050] Typically, two arsenic atoms 6 are visible only at incorporation sites 15 having an area of ​​six or more silicon atoms 5 .

[0051] In a fourth step 108 of the method 100, the surface 7 is annealed to remove the arsenic dopant atoms 5 from the surface 7 and incorporate them into the lattice structure 1. The annealing is performed by heating the sample to 350° C. for approximately 1 minute.

[0052] It will be appreciated that the annealing temperature may be chosen over a range of temperatures. The minimum temperature required is dictated by the activation temperature for diffusion and desorption of hydrogen atoms 11 in the passivation layer 9. The incorporation of arsenic atoms 5 requires that sufficient free silicon sites (dangling bonds) are available for complete dissociation in the case of arsenic moieties 17', as discussed below. These free sites are provided by diffusion of one or two dangling bond windows to adsorption sites via annealing and / or by thermal desorption of H atoms. The minimum temperature required for hydrogen desorption is 350°C, the minimum temperature required for diffusion of a single dangling bond is 275°C, and the minimum temperature required for diffusion of two dangling bonds is 375°C. The maximum temperature is determined by the activation temperature for the diffusion of recombined AsH x The activation temperature for desorption of arsenic is determined by the activation temperature, which is approximately 450°C. Thus, annealing can be between 350°C (to achieve hydrogen desorption) and 450°C (to prevent arsenic desorption).

[0053] It will also be appreciated that the annealing time and the heating ramp rates may be varied.

[0054] An arsine molecule 17 can completely dissociate at room temperature when it is adsorbed onto a free silicon surface 7 without a passivation layer 9. Figure 2G shows the path of an arsine molecule 17 adsorbing onto a silicon surface 7 and dissociating therefrom. It includes the following steps: First, an arsine molecule 17 is coordinately adsorbed on the silicon surface with an adsorption energy of −0.52 eV (determined by density functional theory (DFT) calculations). In the first reaction step, a coordinately adsorbed arsine molecule 17 dissociates to an AsH2+H species by means of an inter-row hydrogen shift. This reaction transfers one hydrogen 11 atom to an adjacent dimer row, giving rise to a hemihydrate dimer. The adsorption energy for this structure calculated by DFT is -1.85 eV. The activation barrier for this reaction step calculated by DFT is only 0.23 eV, corresponding to a reaction that is completed on the nanosecond time scale at room temperature. The inter-row hydrogen shift is kinetically more favorable than the alternative (and perhaps more intuitive) supra- and inter-dimer hydrogen shifts, which have DFT-calculated activation energies of 0.53 and 0.29 eV, respectively. The second reaction step is also a hydrogen shift reaction. The hydrogen shift reaction occurs preferentially in the dimer up direction to give a dimer terminal AsH+2H structure. This reaction transfers one hydrogen 11 atom on the same dimer as the AsH moiety 17'. The adsorption energy for this structure calculated by DFT is -2.20 eV. The activation energy for this reaction step calculated by DFT is 0.63 eV, corresponding to a millisecond reaction time scale at room temperature. The alternative reaction in which the hydrogen atom shifts in the interdimer direction has a slightly larger activation barrier of 0.72 eV. Thus, the alternative reaction is not favored kinetically. The dimer terminal AsH+2H configuration, involving a singly charged AsH fragment, is a highly transient intermediate that corresponds to a very shallow minimum on the potential energy surface. In the third reaction step, this species stabilizes almost instantly to a terminal bridging AsH+2H species, in which the AsH fragment bridges between the two dimer ends. The adsorption energy calculated by DFT for this structure is -2.62 eV. The activation energy for this reaction calculated by DFT is less than 0.05 eV. The final reaction step transfers a hydrogen atom from the AsH fragment to the silicon dangling bond at the opposite dimer end. This reaction yields an inter-row terminal bridge As+3H. The adsorption energy for this structure calculated by DFT is -3.09 eV. The activation energy calculated by DFT for this reaction step is 0.89 eV, which is the rate-limiting activation energy along the full dissociation path. This barrier is still low enough that the full dissociation of the adsorbed AsH3 molecule to the inter-row terminal bridge As+3H is completed on a timescale of tens of seconds.

[0055] For lithography locations 17 with a dangling bond size of four or less, the dissociation mechanism is stopped midway along this reaction path, resulting in the formation of AsH x The moieties are bonded to the lattice 1 because, due to the passivation layer 9, the dangling bonds are not available for the hydrogen atoms 11 to move within or between the dimers. As mentioned above, during annealing, vacancies are created to complete the dissociation by diffusion or desorption. Also, as mentioned above, the barrier to the release of H vacancies is the AsH x Therefore, a high proportion of arsenic atoms 5 are incorporated.

[0056] The table below illustrates the incorporation rate (ie, the percentage of arsenic moieties 17' bound to the surface 7 that result in incorporated arsenic atoms 11) for incorporation sites 15 of different sizes. [Table 2]

[0057] As can be seen from Fig. 2E (bottom row), the dopant atom 5 moves into the lattice 7 and the silicon atom 3 replaced by the dopant atom is displaced towards the surface. The arsenic atom 11 is incorporated only in the top layer of the silicon 1. This means that the silicon atom 5 achieves three bonds to the silicon atoms 3 in the top and second layers of the lattice 1.

[0058] In a final step 110, additional silicon is used to overgrow the silicon surface 7 to cover the dopant atoms 5. As the overgrowth grows, it is believed that hydrogen atoms 11 in the passivation layer 9 have segregated to the surface, meaning that the passivation layer remains on the surface of the grating 1. FIG. 3C shows the silicon surface 7 after the dopant atoms 5 adsorbed in FIG. 3B have been incorporated and overgrown. The dopant atoms 5 may be overgrown by a layer of any desired depth. Typically, the dopant atoms are overgrown by a layer between 2 and 20 nm.

[0059] The result of the method 100 is a lattice 1 with arsenic dopant atoms 5 provided at known locations throughout the silicon lattice 1. The location of each dopant atom 5 is known to be within a limited number of atoms in the lattice 1, determined by the size of the lithography site. The valence spin of the electron can be used as a qubit for data storage or data processing. Alternatively, the spin of the nuclei can be used as a qudit for data storage or data processing. The nuclear spin can retain its state for several days. Additionally, the nuclear spin or the electron spin may be used for the operation of other electronic devices.

[0060] Those skilled in the art will appreciate that a variety of methods can be used to control, manipulate, and read out the qubits and qudits. For example, highly doped regions (not shown) may be provided within the lattice 1 above or below the plane into which the dopant atoms 5 are incorporated to act as electrical contacts and / or gates. Alternatively, other conductive materials such as gold can be used as gates and contacts.

[0061] It will be appreciated that the method 100 described above can be tailored to provide any number of dopant atoms 5 in a surface. For example, a silicon wafer (not shown) can be provided with an array of dopant atoms 5 and cut into individual chips or devices. Wafers with thousands of qubits or qudits can be fabricated by this method.

[0062] Processes for automatic alignment and actuation of the STM tips 13 will be apparent to those skilled in the art. In one example, an array of cantilevered STM tips 13 or AFM tips can be used for faster processing. See, for example, P. Vettiger et al., "The "Millipede"-More than thousand tips for future AFM storage," IBM Journal of Research and Development, Vol. 44, No. 3, pp. 323-340, May 2000, doi:10.1147 / rd.443.0323, or R. Huff et al., "Atomic White-Out: Enabling Atomic Circuitry through Mechanically Induced Bonding of Single Hydrogen Atoms to a Silicon Surface," ACS Nano, Vol. 11, No. 9, pp. 8636-8642, July 2017.

[0063] In the method 100 presented above, the position of the dopant atom 5 is known to be within several silicon atoms 3 in the lattice 1 depending on the size of the incorporation site 15 .

[0064] Generally, it is known that the larger the incorporation sites, the greater the percentage of sites that are successfully doped, but the less precise the positioning of the dopant atoms 5. For example, For each incorporation site 15 whose area is four silicon atoms 11, it is known that, except for a negligible number of incorporation sites 15s, a single dopant atom 5 is always incorporated and its location is within the area of ​​four atoms. For example, for an incorporation site 15 with an area of ​​two silicon atoms 11, it is known that approximately 80% of the sites incorporate a single dopant atom 5 and the positions are within an area of ​​two atoms. For example, for an incorporation site 15 with an area of ​​one silicon atom 11, it is known that about 10% of the sites incorporate a single dopant atom 5 and the positions are within a single atom.

[0065] Figure 4 illustrates an alternative method 200 that can be used to provide even greater precision in the positioning of dopant atoms 5. The method 200 of Figure 4 is the same as the method 100 of Figure 1 unless otherwise stated.

[0066] According to the method 200 of Figure 4, in a first step 202, a passivation layer 9 is formed, in a second step 204, incorporation sites 15 are formed in the passivation layer 9, and in a third step 206, the incorporation sites 15 are exposed to arsine gas AsH317. The formation of the passivation layer 9, the formation of the incorporation sites 15, and the exposure to arsine gas 17 are the same as the corresponding steps 102, 104, and 106 in the method 100 shown in Figure 1. However, in the method 200 shown in Figure 4, the area of ​​each incorporation site 15 is only the area of ​​a single silicon atom 3 (i.e., a single dangling bond) in the lattice 1.

[0067] As discussed above, for incorporation sites of single silicon atoms 11 or single silicon dimers, arsine moieties 17' are not adsorbed at all sites. The method 200 of Figure 4 includes, after exposing the incorporation sites 15 to arsine gas 17 in step 206, using an STM tip 13 to identify incorporation sites 15 that do not have adsorbed moieties 17'. The STM tip 13 may be the same tip used to remove hydrogen atoms 11 to form the incorporation sites 15, or it may be a different STM tip 13.

[0068] In the next step 214, a check is made to determine if the method termination criteria have been reached, as will be discussed in more detail below.

[0069] If the termination criterion has not been reached, the method 200 returns to the second step 204 .

[0070] Where no arsine moiety 17' has been absorbed, hydrogen 11 dissociated from the arsine molecule will fill the incorporation site 15 by terminating the dangling bonds.

[0071] When the method 200 moves from the verifying step 214 to the removing step 204 of the passivation layer 9 at the incorporation site 15, AsH x The passivation layer 9 is removed only at those incorporation locations 15 that are identified as not yet having adsorbed moieties 17'. Locations that are determined to have adsorbed moieties 17' are not exposed. The adsorbed moieties 17' inhibit further adsorption in subsequent stages and thus form part of the passivation layer 9. Due to the energy of the bond between the moieties 17' and the silicon surface 7, the adsorbed moieties 17' do not desorb from the surface and rebond.

[0072] The loop 216 of forming incorporation sites 15, exposing to arsine gas 17, identifying incorporation sites 15 that do not have arsine moieties 17', and checking the termination criteria is repeated iteratively until the termination criteria are met. Thermal annealing is then performed in step 208 to incorporation arsenic atoms 5 from the adsorbed moieties 17'. As discussed above, approximately 100% of the adsorbed moieties result in adsorbed arsenic atoms 5.

[0073] Finally, in step 210, the silicon layer is overgrown.

[0074] The thermal annealing and overgrowth are the same as the corresponding steps 108 and 110 discussed with respect to the method 100 shown in FIG.

[0075] Various termination criteria may be used in the validation step 214. These include, for example: · 100% adsorption of arsenic moieties 17' at the uptake sites 15 (i.e. no free sites identified). The proportion (or number) of uptake sites 15 having adsorbed arsenic moieties 17' is above a threshold value (the threshold value being less than 100%). A predefined number of loops has been completed. The predefined number of loops may be determined based on the statistical likelihood of reaching 100% adsorption of arsenic moieties 17' at the incorporation sites (or the percentage of incorporation sites 15 having adsorbed moieties 17' being above a threshold). A predefined number of loops have been completed with the number of locations that do not have absorbed arsenic moieties 17' remaining unchanged; or Default time. Other suitable termination criteria may be used.

[0076] By the method shown in Figure 4, high levels of incorporation of dopant atoms 5 can be achieved with precise placement of the dopant atoms 5 such that the position of the dopant atoms is known to be within a single atom in the lattice 1 when a single atom incorporation site is used, or within two atoms in the lattice 1 when a single silicon dimer incorporation site 15 is used. It will be appreciated that, as with the method of Figure 1, step 200 of Figure 4 may be automated.

[0077] It will be appreciated that the iterative loop 216 need not be used only in combination with incorporation sites 15 having areas the size of a single silicon atom 3 or dimer. If a high percentage of dopant incorporation is required (for any dopant, not just arsenic), an iterative process may be used to ensure a reduction in the number of incorporation sites 15 that do not have a dopant.

[0078] It will be appreciated that the higher the percentage of moiety 17' that adsorbs in each loop, the fewer loops that will be required to achieve a high level of overall incorporation. For example, when the incorporation sites 15 are the size of a single dimer, 80% of the sites have adsorbed moiety 17' in a single loop, currently requiring fewer loops than when the incorporation sites 15 are the size of a single atom. When the incorporation sites 15 are the size of a single atom, only 10% of the sites have adsorbed moiety 17' in a single loop.

[0079] It will be appreciated that in some cases the capture location may increase in size in subsequent loops so that the termination criteria can be achieved.

[0080] As mentioned above, the passivation layer 9 is a single atom thick. It will be appreciated that the passivation layer 9 may contain defects (holes) 19 where atoms of the passivation layer have been removed. These defects 19 have dangling silicon bonds and can form (unwanted) incorporation sites 15. Even at the size of a single atom, this results in a 10% probability of adsorption and incorporation of arsenic atoms 5 in undesired locations.

[0081] Figure 5 shows a method 300 for forming a passivation layer 9 that repairs / removes any defects 19. This method 300 is performed before the incorporation sites 15 are formed. Figures 6A-6C show schematic diagrams of the passivation layer 9 at different stages, as discussed below. Figures 6A-6C show the passivation layer 9 in a top view.

[0082] In a first step 302, as described above, an atomic layer of hydrogen atoms 11 is deposited. As shown in Figure 6A, there are many defects 19 (missing hydrogen atoms 11, also referred to as holes in the passivation layer 9). The defects 19 are the size of a single missing atom and are generally spaced apart from one another.

[0083] In a second step 304, the passivation layer 9 is heated to about 200° C. This induces diffusion of the defects 19 to form defect pairs 19′ as shown in FIG 6B.

[0084] In a third step 306, the passivation layer 9 is exposed to a precursor gas having molecules 21 containing phosphorus atoms 23 and hydrogen. For example, the precursor may be phosphine gas PH3 21, which dissociates into phosphine moieties (usually PH2) 21' and H atoms, which are adsorbed at the defect pairs 19' and inhibit the defects 19 from forming incorporation sites 15 when the passivation layer 9 is subsequently exposed to arsine gas 17.

[0085] The phosphine repaired passivation layer 9 is then used in the method 100, 200 shown in Figure 1 or Figure 4. The arsenic moieties 17' are unable to bond at the defect sites 19' because they are blocked by dissociated phosphine 21. Thus, the dopant arsenic atoms 5 are incorporated only at the locations 15 where they are intended to be present.

[0086] In the annealing steps 108 and 208 used to incorporate the arsenic atoms 5, the annealing temperature is sufficient to cause the dissociated phosphine to desorb and recombine rather than be incorporated into the lattice 1.

[0087] During step 304 of heating the passivation layer 9 as part of repairing the defects 19, the layer should be heated to a temperature sufficient to induce diffusion of the defects 19 to form defect pairs 19', but below a temperature which would induce any material changes to the lattice 1. Thus, the layer should be heated to a minimum of 275°C, and no higher than 350°C. As mentioned above, this ensures diffusion of hydrogen vacancies, but does not induce hydrogen desorption.

[0088] There may be a small number of single dangling bond sites remaining in the resist, however, as mentioned above, over 90% of these sites will be re-terminated, so the amount of defects in the resist can be considered negligible.

[0089] It will be appreciated that the process 300 shown in Figure 5 is optional and may not be used. Additionally, this process may be used upon incorporation of any dopant, not just arsenic.

[0090] In the above described methods 100 and 200 , arsenic dopants are incorporated into the silicon lattice 1 .

[0091] It will be appreciated that the arsenic dopant may be incorporated into other semiconductor lattices 1. For example, a germanium lattice may be used instead of a silicon lattice.

[0092] The inventors have surprisingly found that no annealing step is required, since the arsenic atoms 5 from the arsine moieties 17' are incorporated into the (001) surface of the germanium lattice 1 upon adsorption at room temperature. In other words, the arsine / germanium system progresses from the state shown in Figure 2D to the state shown in Figure 2E without the need for heating (or other energy transfer) as described in step 108. This occurs necessarily, except for a negligible number of incorporation sites 15.

[0093] In the following discussion, the temperature is 5×10 -10 STM experiments were carried out on a Sien Taomicron low-temperature STM system operating under ultra-high vacuum at less than mbar. Highly resistive germanium 001 oriented samples (antimony doped, 1-100 Ωcm) were degassed overnight at 200°C and then warmed to 760°C by DC heating for 1 hour. The samples were subsequently cleaned by repeated cycles of sputtering (1.2 kV, 10 mA, 30 min) and DC annealing (700°C, 30 min). Sample preparation was then completed by three 30 s annealings to 760°C and cooling from 600°C at 25°C / min.

[0094] An infrared pyrometer (IMPAC IGA50-LO plus) was used to measure the sample temperature and provided absolute temperature measurements with an accuracy of ±30°C.

[0095] With the STM tip macroscopically housed, arsine dosing was performed in-situ. -10 The sample was exposed to 99.999% pure arsine (ATMI Inc.) at mbar total chamber pressure. After arsine dosing, the STM tip was brought close again and imaging was started within 15 min of dosing.

[0096] The adsorption energy ΔE of the surface structure resulting from the adsorption and dissociation of arsine on Ge(001) was calculated as the formation energy of the adsorption as follows: ΔE=E (クラスター+AsH3) -E (クラスター) -E (AsH3) In the formula, E クラスター is the total energy of the bare germanium cluster, E (クラスター+AsH3) is the total energy of the cluster with adsorbed or dissociated arsine molecules, and E AsH3 is the total energy of the gas-phase arsine molecule.

[0097] In some of the adsorbed structures considered here, the germanium atom has been displaced away from the adsorption site to a position farther away than can be accommodated by a 3-dimer cluster. In this case, the adsorption energy is calculated as ΔE = (1 / 2) × [2E (クラスター+AsH3-Ge) +E (クラスター+2Ge) -3E クラスター -2E (AsH3) ] In the formula, E (クラスター+AsH3-Ge) is the total energy of the cluster with a dissociated arsine molecule and one missing Ge atom, and E (クラスター+2Ge) is the total energy of the cluster with Ge-Ge add-dimers bridging the two surface dimers vertically. These add-dimers are directly observed in the experiment and are therefore an accurate reference for accounting for displaced germanium atoms in our calculations.

[0098] Figure 7 shows an STM overview image of a germanium (001) surface doped with arsine at room temperature. Rows of germanium dimers stretch horizontally across the image, forming either a buckled c(4x2) configuration or a more symmetric looking 2x1 configuration. Features related to arsine dosing are labeled B, h1, m1, and t1, and examples of these are highlighted with boxes.

[0099] The B feature is the brightest feature observed and appears as a single protrusion at the top of the dimer row and at a position halfway between two adjacent dimers. The feature type labeled "B" is also the observed displacement position on the surface. This is another known characteristic of germanium add-dimers at room temperature. The presence of germanium add-dimers on the surface after AsH3 dosing suggests that arsenic atoms were incorporated into the surface and germanium atoms were expelled during the treatment.

[0100] The m1 feature is the most abundant feature after arsine dosing and appears as two adjacent dark dimers. The feature is not perfectly symmetric and appears to be slightly less dark at one corner. This feature can be mapped to the following structure: [ka]

[0101] This is a fully dissociated arsine molecule, resulting in one arsenic atom incorporated at the surface and three adsorbed hydrogen atoms. The hydrogen atoms 11 are bonded resulting in a hydrogen-terminated arsenic-germanium heterodimer (As-Ge-H) and a monohydride dimer (H-Ge-Ge-H). This feature appears darker than the surrounding bare germanium dimers due to the absence of a germanium dimer pi bond and the formation of a lone pair on the arsenic atom induced by the adsorption of the three hydrogen atoms. The slightly less dark area at one corner of the feature is due to this arsenic atom.

[0102] The t1 and h1 features are three dimers wide. The t1 feature shows a complex appearance and spans two dimer rows, while the h1 feature shows a dark single dimer flanked by two protuberances that are asymmetrically positioned relative to the dimer rows.

[0103] The h1 feature corresponds to one hydrogen-terminated arsenic-germanium heterodimer (As-Ge-H) and two adsorbed hydrogen atoms. However, in this case, two hydrogen atoms are adsorbed at the ends of two adjacent dimers to form two germanium hemihydride dimers (Ge-Ge-H) rather than a single dimer as in the case for m1. These two protrusions observed in the STM image are caused by the two germanium dangling bonds of the hemihydride dimer, as highlighted by the circles in the structural schematic below. [ka]

[0104] A transition is sometimes observed between this feature and several others that are three dimers wide. One or both of the asymmetric protuberances change from the top to the bottom of the dimer row. This is due to one or both hydrogen atoms moving from one side of the germanium dimer to the other. If one hydrogen atom moves, both are called h2, and if both hydrogen atoms move, this is called h3.

[0105] In a further transition, two prongs of a hemihydride dimer with hydrogen atoms on the same side of the germanium dimer are merged to form a single prong, which is induced by the capture of a single germanium add atom into the germanium dangling bonds of two hemihydride dimers in a terminal bridge bond configuration, as shown below. [ka]

[0106] This is called the g1 feature. In the final transition, the g1 feature can become the above-mentioned feature t1 (FIG. 1f) that spans two dimer rows.

[0107] It was observed that the size and strength of the protrusions caused by the add atoms of the g1 feature increased. Thus, phason defects (i.e., defects in which two adjacent dimers are statically twisted in the same direction) can be pinned by the g1 feature on adjacent dimer rows. Spontaneous and localized pinning of phasons is expected only at very low temperatures. Thus, there is believed to be a chemical origin for this pinning, specifically, the capture of a second germanium add atom to form a dimer trough bridging the germanium add dimers, as shown below. [ka]

[0108] This structure illustrates both the expansion of the protrusion at the location of the addatom in structure g1 and the dimer pinning that induces a phason-like defect.

[0109] Thus, the transitions h3→g1 and g1→t1 result from the dynamic trapping of germanium add atoms that are bound in the g1 or t1 features, but can occasionally overcome a diffusion barrier at room temperature to leave one feature and become trapped in another. Figure 8 shows two examples of such transitions, showing both the source and destination of the transitioning germanium add atoms.

[0110] The top panel of Figure 8 shows an STM image in which the t1 and g1 features are separated horizontally by the spacing of three dimers and vertically by a single germanium dimer row. The bottom panel of Figure 8 shows a subsequent image of the same area of ​​the surface, where the t1 feature becomes a g1 feature and vice versa, mediated by the exchange of germanium adatoms between the two. This exchange is observed to alternate several times in successive STM images, demonstrating the reversibility of these transitions.

[0111] Further insight into the dynamics of the surface is gained by counting the transitions of features between successive STM images. The averaging time between images was 9 min, and the imaging parameters were -1.5 V, 200 pA. Table 1 shows the number of 50 × 50 nm 2 3 shows the counts of transitions observed in the scanned images of 1. As expected for reversible transitions, there is a strong symmetry about the diagonal for the table. [Table 3]

[0112] The largest number of transitions, 56% (129 of 231) of the total observed transitions, involved hydrogen atom transfer to the opposite side of the dimer row to induce transitions between h1, h2, and h3 features. The next largest category was the add atom exchange transitions g1←→t1, which accounted for 35% of the total. Transitions between h1-h3 and g1 were relatively rare, accounting for 9% of the total transitions, demonstrating that the majority of add atom transitions are exchanges between the dimer t1 configuration and the add atom g1 configuration. No transitions were observed between the h1-h3 and t1 structures, which is understandable since the g1 structure is an intermediate between h1-h3 and t1.

[0113] It should also be noted that there were no g1↔g1 or t1↔t1 transitions. Experimentally, these would appear to be mirror-symmetric transitions for the dimer train, but given the structural assignments used, such transitions are highly unlikely to occur.

[0114] The experimental observations are supported by the formation energies calculated by density functional theory (DFT). Each of the experimentally observed structures was calculated and the formation energies for h1-h3 were found to be all equal to -2.8 eV, and for structure g1, it was found to be nearly equal to -2.9 eV. The closeness of these energies suggests the possibility of reversible transitions between these structures.

[0115] Structure m1 is slightly more stable at -3.1 eV. In addition, energies for a wide range of AsHx+(3-x)H adsorption configurations were calculated. The most stable of these are the AsH+2H and As+3H configurations, both of which have formation energies of -2.2 eV. Thus, a large energy gain (0.6 eV) is observed for arsenic incorporation into the top surface layer, and a very large overall energy gain (~3 eV) for surface incorporation compared to arsine in the gas phase. These large energy gains are the driving force supporting the surface incorporation of arsenic into Ge(001) at room temperature.

[0116] Further supporting this interpretation, it was confirmed that the density of ejected germanium was equal to the density of incorporated arsenic. Feature t1 was one of the most common features observed, accounting for 30±11% of all observed arsenic-germanium heterodimer features (m1, h1-h3, g1, and t1). Feature g1 occurred less frequently, accounting for 9±4% of the heterodimer features. Meanwhile, the number of B-site germanium add-dimers is equivalent to 15±5% of the number of heterodimer features. Thus, keeping in mind that each t1 and add-dimer feature encompasses two germanium atoms, this explains why 99±25% of the germanium atoms ejected onto the surface result from arsenic incorporation.

[0117] It will be appreciated that germanium may be used in place of silicon in any of the methods 100 and 200 discussed above. Additionally, germanium may be used in any other suitable incorporation process.

[0118] It will be appreciated that while the incorporation of arsenic dopants in germanium does not require an annealing step, annealing (e.g., to a temperature of 140° C.) may be used to reduce surface defects. In general, the annealing temperature may be lower than that typically required for the incorporation of arsenic into silicon and will not induce migration of doping atoms. Furthermore, the onset of desorption of hydrogen atoms from the surface of germanium occurs at a lower temperature than for silicon, likely below 250° C. Thus, any temperatures mentioned above for silicon are defined by the desorption of hydrogen atoms and may be correspondingly modified.

[0119] Any of the above methods 100 and 200 may include the optional step of annealing the semiconductor grating 1 before applying the passivation layer 9. This may planarize the surface 7 of the grating 1 before any processing. Techniques for such annealing steps are known in the art. For example, the semiconductor may be heated to an elevated temperature (greater than 500° C.) and the heating step may be repeated with sputtering and / or cooling.

[0120] Above, a single dopant is provided. However, further processing steps are discussed to incorporate the resulting structure into an electronic device. It will be appreciated that this may involve doping regions of the lattice structure with other dopants, such as phosphorus. Additional dopants may be provided above or below the plane of the arsenic dopant, and may be included by hydrogen desorption lithography, which involves the following steps: (1) applying a hydrogen passivation layer of one atom thickness; (2) exposing the incorporation site; (3) exposing to a precursor gas having molecules containing dopant atoms and hydrogen atoms; and (4) (Optional) Thermal annealing to incorporate dopant atoms.

[0121] For example, when incorporating phosphorus, to achieve the same precision in placement, the incorporation site should be exposed to the precursor gas at a temperature below 100 K or larger. However, for other dopants, the same precision may not be required. Alternative processes may be used to include other dopants.

[0122] In the above, single atoms have been incorporated to provide a qubit (or qudit). However, it will be appreciated that the qubit / qudit (or device for controlling or reading out the qubit / qudit) can also be formed by quantum dots or quantum wires, or two-dimensional electron gas (sheets), formed by small clusters of dopant atoms 5 that still exhibit quantum properties. Larger structures of dopant atoms 5 that exhibit bulk properties may also be provided. Such areas of dopant atoms 5 can be formed by making the incorporation sites 15 larger. It will be appreciated that for reasons similar to those discussed above, the incorporation sites 15 for forming quantum dots of arsenic dopant atoms will be smaller than the incorporation sites 15 for forming quantum dots of similar size of phosphorus atoms.

[0123] Above, an STM tip 13 is used to remove the passivation layer 9 at the incorporation locations 15 and, if necessary, to identify locations 15 that do not have absorbed arsine moieties 17'. This is by way of example only. Any suitable method for removing atoms 11 from the passivation layer 9 may be used. For example, an electron beam (e-beam) and an STM scanning microscope may be used.

[0124] In some of the above methods, a thermal annealing step is used to incorporate the adsorbed dopant atoms 5. However, this is by way of example only. Any suitable method of energy transfer may be used. For example, the energy may be transferred via an STM tip or an e-beam.

[0125] In the above example, the passivation layer 9 is hydrogen and the precursor gas molecules contain hydrogen in addition to phosphorus or arsenic. In other examples, different passivation layers may be used and the precursor gas molecules may contain the same or different atoms as the passivation layer.

[0126] Arsine and phosphine are given as examples of precursor gases for applying dopants and for repairing defects 19 in the passivation layer, but these are merely examples and other suitable precursor gases for applying phosphorus and arsenic may be used.

[0127] As discussed above, it will be appreciated that high levels of incorporation at room temperature can be achieved when the energy barrier for diffusion and / or desorption of the passivation layer 9 is lower than the desorption of the precursor moiety 17', thereby causing the dopant to be preferentially incorporated rather than dissociated.

[0128] In the above, DFT calculations were performed using the B3LYP hybrid functional with exact exchange terms, atom-centered Gaussian-type orbital basis functions, and methods of energy calculation and geometry optimization, as implemented in the Gaussian09 or Gaussian16 software. Using an approach called the cluster complex model, the activation energies and adsorption ensembles were calculated for various configurations of arsine on silicon (001) and germanium (001) surfaces.

[0129] To model silicon, use Si 15 H 16 or Si54 H 44 A cluster model was used to represent the Si(001) surface and corrections were made to minimize finite cluster size effects. The reaction steps in Figure 2G were calculated using a 3D2R cluster for the first step and a 2D1R cluster with the hemihydride dimer represented by a separate cluster for the remaining steps.

[0130] For modeling germanium, a compact Ge 21 H 20 A cluster model was used to represent the Ge(001) surface, describing three Ge-Ge dimers in the atomic layer closest to the surface, and eight, four, and three atoms in the second, third, and fourth atomic layers, respectively. The 20 hydrogen atoms of this cluster provide chemical terminations for all germanium atoms except those in the surface layer. These hydrogen atoms terminating the cluster were kept in their default positions during all geometry optimizations to simulate the strain that would be exerted by the surrounding surface and the extended surface bulk atoms. All other atoms were fully relaxed.

[0131] For calculations on silicon, the 6-311G (2df,2pd) atom-centered basis set was used.

[0132] In the calculations for germanium, two types of composite basis sets, referred to below as large and small, are used. Geometry optimization and vibrational frequency calculations are performed using a small basis set consisting of standard 6-311++G(d,p) basis sets for all atoms of the adsorbate and the top surface layer of the cluster (i.e., dimer atoms), 6-311G(d,p) basis sets for the second layer atoms, and core pseudopotential LANL2DZ basis sets for the third and fourth layer atoms and hydrogen atoms terminating the cluster. After geometry optimization, single-point energies are calculated on the optimized structures using a large basis set consisting of 6-311++G(2df,2pd) basis sets for the adsorbate atoms and the top surface layer, and 6-311G(2df,2pd) basis sets for all other atoms. Thus, using quantum chemical notation, the total energy E of a given structure is given by E SCF It is calculated as (B3LYP / large / / B3LYP / small) + ZPE(B3LYP / small).

Claims

1. 1. A method for incorporating arsenic dopant atoms at defined locations in a semiconductor lattice, comprising: (i) forming a passivation layer on a surface of the semiconductor grating; (ii) selectively removing the passivation layer at one or more incorporation locations to expose the surface of the semiconductor lattice; (iii) exposing the incorporation sites to a dopant precursor gas to adsorb dopant precursor moieties comprising arsenic atoms onto the surface of the semiconductor lattice at at least some of the incorporation sites; Equipped with arsenic dopant atoms are incorporated into the lattice at the incorporation sites; method.

2. The method of claim 1 , wherein the step of exposing the surface to a dopant precursor gas is carried out at room temperature.

3. The method of claim 1 or 2, wherein the semiconductor lattice is a Group IV semiconductor lattice.

4. The method of claim 3 , wherein the semiconductor lattice is a silicon lattice.

5. 10. A method according to any preceding claim, wherein the method comprises subjecting the surface layer to a thermal anneal to incorporate the arsenic dopant atoms into the semiconductor lattice.

6. The method of claim 3 , wherein the semiconductor lattice is a germanium lattice.

7. 7. The method of claim 6, wherein upon adsorption of the dopant precursor moiety, the dopant precursor molecule dissociates and the arsenic atom is incorporated into the semiconductor lattice in a manner that does not require energy transfer.

8. 2. The method of any preceding claim, further comprising overgrowing the surface of the semiconductor lattice after the arsenic dopant atoms are incorporated.

9. 2. The method of any preceding claim, wherein the area of ​​each incorporation site is at most four semiconductor atoms.

10. 10. The method of claim 9, wherein the area of ​​each incorporation site is four semiconductor atoms and arsenic dopant atoms are incorporated at 100% of the incorporation sites.

11. 10. The method of claim 9, wherein the area of ​​each incorporation site is a semiconductor dimer and arsenic dopant atoms are incorporated in 80% of the incorporation sites.

12. 10. The method of claim 9, wherein the area of ​​each incorporation site comprises a single semiconductor atom, and optionally, arsenic dopant atoms are incorporated in about 10% of the incorporation sites.

13. Prior to incorporating the arsenic atom, (a) identifying incorporation sites where no arsine adsorption occurs and where atoms of the passivation layer are re-adsorbed; (b) selectively removing the passivation layer to expose the identified incorporation locations; (c) exposing the identified incorporation locations to the dopant precursor gas to adsorb the dopant precursor moieties onto the surface of the semiconductor lattice at the identified incorporation locations; The method of any of claims 9 to 12, further comprising:

14. Repeating steps (a) through (c) in an iterative manner at least once; (d) incorporating the arsenic atoms into the semiconductor lattice; The method of claim 13 further comprising:

15. Between step (i) and step (ii), heating the passivation layer to induce diffusion of defects in the passivation layer such that the defects pair up and cluster; exposing the passivation layer to an additional precursor gas comprising additional precursor molecules containing the same atoms as the passivation layer to fill the holes in the passivation layer; 13. The method of any preceding claim, comprising:

16. The method of claim 15 , wherein the additional precursor gas comprises phosphorus atoms in addition to the same atoms as the passivation layer.

17. The additional precursor gas is phosphine PH ３ 20. The method of claim 16 comprising:

18. The method according to any of claims 15 to 17, wherein the passivation layer is heated to at least 200°C to induce diffusion of holes in the passivation layer.

19. forming a second passivation layer on a surface of the semiconductor lattice; selectively removing the second passivation layer at one or more second incorporation locations to expose the surface of the semiconductor lattice; exposing the second incorporation locations to a second dopant precursor gas to adsorb second dopant precursor substructures comprising atoms of a second dopant, different from arsenic, onto the surface of the semiconductor lattice at at least some of the second incorporation locations; Equipped with 2. A method according to any preceding claim, wherein atoms of the second dopant are incorporated into the lattice at the second incorporation sites.

20. Selectively removing the passivation layer at one or more incorporation locations to expose the surface of the semiconductor lattice comprises: Scanning tunneling microscopy, and Electron beam microscopy, 13. The method of any preceding claim, performed by one of

21. Selectively removing the passivation layer at one or more incorporation locations to expose the surface of the semiconductor lattice comprises: Arrays of scanning tunneling microscopy probes 20. The method of any preceding claim, performed by:

22. 13. A method according to any preceding claim, comprising the step of annealing the surface of the semiconductor lattice to form an atomically flat surface prior to forming the passivation layer on the surface of the semiconductor lattice.

23. 10. The method of any preceding claim, wherein the passivation layer comprises atomic hydrogen.

24. The dopant precursor gas is arsine AsH ３ 13. The method of any preceding claim, comprising:

25. 10. The method of any of the preceding claims, wherein the energy barrier for the diffusion and / or desorption of atoms of the passivation layer is lower than the energy barrier for desorption of the precursor substructures from the surface of the semiconductor lattice.

Images