Dielectric material and method for forming same

Abstract

A dielectric material and a method of forming the same. A dielectric material is provided comprising a film comprising a layer of a two-dimensional (2D) material wherein the film has a dielectric constant ([kappa]) < = 3.0. A method of forming the dielectric material is also provided.

Description

Technical Field

[0001] This invention relates to a dielectric material and a method for forming the same. Background Technology

[0002] Conventionally, increasing the porosity of dielectric materials reduces the dielectric constant. However, although this compromise is detrimental to device fabrication and structure, particularly mechanical strength, it remains used due to a lack of alternative materials.

[0003] Existing dielectric materials cannot reduce their thickness without sacrificing dielectric constant properties.

[0004] Therefore, there is a need for an improved dielectric material and its formation method. Summary of the Invention

[0005] The present invention seeks to solve these problems and / or provide an improved dielectric material, particularly a dielectric material comprising two-dimensional (2D) materials.

[0006] According to a first aspect, the present invention provides a dielectric material comprising a film, the film comprising a layer of two-dimensional (2D) material, wherein the dielectric constant (κ) of the film is ≤ 3.0.

[0007] According to one particular aspect, the 2D material may comprise a monolayer of amorphous carbon (MAC).

[0008] The thickness of the membrane can be ≤ 20 nm. For example, the thickness of the membrane can be 0.5-3 nm.

[0009] Specifically, the membrane may comprise at least two layers of a 2D MAC. For example, the membrane may comprise two to five layers of a 2D MAC.

[0010] The hardness of the membrane can be > 10 GPa. The dielectric strength of the membrane can be > 8 MV cm. -1 .

[0011] According to one particular aspect, the membrane may be non-porous.

[0012] According to a particular aspect, the film can be formed on at least a portion of a non-catalytic substrate. Specifically, the non-catalytic substrate can include silicon-based substrates, carbon-based substrates, metal-based substrates, oxides, transition metal dichalcogenides, mXene, or any combination thereof. For example, the non-catalytic substrate can include cobalt, gold, copper, nickel, tungsten, molybdenum, ruthenium, niobium, silicon dioxide, silicon nitride, titanium nitride, tantalum nitride, cobalt oxide, tungsten carbide, germanium, gallium arsenide (GaAs), silver, stainless steel, Fernico alloys (iron-nickel-cobalt alloys or Fernico alloys), manganese, aluminum, or any combination thereof.

[0013] According to a second aspect, a method for forming a dielectric material is provided, the method comprising depositing carbon radicals on a non-catalytic substrate.

[0014] According to one particular aspect, the method may also include repeatedly depositing carbon radicals to form up to five layers of 2D MAC.

[0015] According to another particular aspect, the deposition of carbon radicals can include the formation of carbon radicals via photodissociation of a carbon source. The deposition of carbon radicals can also include the formation of carbon radicals via absorption at a UV wavelength of the carbon source. Specifically, the UV wavelength can be 200-400 nm. The carbon source can include acetylene, methane, ethylene, ethanol, propane, adventitious carbon (accidental or foreign carbon), or any combination thereof.

[0016] According to one particular aspect, the substrate may include a non-catalytic substrate. Specifically, the non-catalytic substrate may include silicon-based substrates, carbon-based substrates, metal-based substrates, oxides, transition metal dichalcogenides, mXene, or any combination thereof. For example, the non-catalytic substrate may include cobalt, gold, copper, nickel, tungsten, molybdenum, ruthenium, niobium, silicon dioxide, silicon nitride, titanium nitride, tantalum nitride, cobalt oxide, tungsten carbide, germanium, gallium arsenide (GaAs), silver, stainless steel, Fenico alloys, manganese, aluminum, or any combination thereof.

[0017] The deposition can be performed in a plasma environment. According to a particular aspect, the method may also include generating a plasma environment prior to deposition. For example, the plasma environment may include a remotely inductively coupled plasma. Brief description of the attached diagram

[0019] To enable a full understanding and ease of practice of the invention, exemplary embodiments are now described by way of non-limiting example, with reference to the accompanying illustrative drawings. In the drawings:

[0020] Figure 1This demonstrates the direct growth of monolayer amorphous carbon; Figure 1 (a)-(c) show cross-sectional transmission electron microscopy (TEM) images of MAC at 3L, 2L and 1L (layer) thicknesses on SiO2 / Si, respectively, where EELS overlays show the distribution of C and O2 for accurate thickness determination. Figure 1 (d) shows a 4-inch Si / SiO2 wafer (wafer) that is mostly covered by a 1.5nm MAC layer (ML-AC) and is clearly distinguished from the substrate by optical contrast (i.e., ML-AC represents the first part and SiO2 represents the second part, which is located on the side of the wafer). Figure 1 (e) shows atomic force microscopy (AFM) thickness measurements of 1–4L MAC films directly grown on SiO2 as a function of growth time. Each layer is numbered, with 1L denoted by 1a and 1b, 2L by 2a and 2b, 3L by 3a to 3d, and 4L by 4a to 4d (the letter next to each layer number indicates integrity; maximum integrity for single and double layers is reached at the letter “b”, and maximum integrity for multilayer layers is reached at the letter “d”). Figure 1 (f)-(j) show top-view TEM images with atomic resolution, sequentially illustrating layer-by-layer growth: Figure 1 (f), 1(h), and 1(j) correspond to the complete 1L, 2L, and 3L layers of MAC, respectively; Figure 1 (g) and 1(i) correspond to incomplete second and third layers, respectively, where the growth of the second layer transitions from 1L to 2L and the growth of the third layer transitions from 2L to 3L over time; Figure 1 (k)-(m) show scanning transmission electron microscopy (STEM) images of 1L, 2L and 3L MAC after transfer from SiO2, respectively;

[0021] Figure 2 (a)-(l) show the thickness AFM analysis of MAC as a function of time, where the synthesis time and number of layers are indicated;

[0022] Figure 3 A summary of the spectral analysis of the MAC samples is presented; Figure 3 (a) shows typical Raman spectra at different growth durations (the numbers and letters to the right of each line indicate the layer and its integrity); Figure 3 (b) shows an optical image of a set of metal lines covering the MAC, with the metal lines shown in the inset; Figure 3 (c) Shows individual Raman spectra collected from SiO2, Cu, and Co surfaces (illustrated in the inset). Figure 3 (b) Schematic diagram); Figure 3(d) shows the C1 s lines as a function of growth time (the numbers and letters to the right of each line indicate the layer and its integrity), and the sample with... Figure 1 match; Figure 3 (e) shows the similarity of the C 1s lines of MACs grown on metallic and insulating substrates; Figure 3 (f) shows the NEXAFS spectra of monolayer (solid line) and four-layer samples (dashed line), as well as the dependence of the 1s-π* and 1s-σ* transitions on the incident angle (for monolayer samples, experimental geometry is shown in the inset).

[0023] Figure 4 This shows the uniform growth of MAC; Figure 4 (a) shows the corresponding Figure 1 EELS carbon k-edge spectra of 1L, 2L and 3L samples (a)-(c); Figure 4 (b) and 4(c) show the uniformity of the Raman signal across the entire 4-inch wafer, where points on the wafer are marked p1, p2 and p3, and the spectrum is collected from these points; Figure 4 (d) shows an AFM image of the near-atomic-level smooth surface of the MAC film grown on Si / SiO2; Figure 4 (e)-(g) show the original and pure state of the substrates grown after MAC deposition in the cases of Co, Cu and Si, respectively;

[0024] Figure 5 The hardness of the MAC is shown; Figure 5 (a) shows the hardness measured by AFM indentation of a 2.1 nm thick film of MAC on SiO2 and compared with a reference sample of SiO2 only; Figure 5 (b)-(c) show the AFM morphology of the equivalent indentation of SiO2 and MAC on SiO2;

[0025] Figure 6 A schematic diagram of a metal interconnect stack is shown, in which a low-k dielectric surrounds conductive elements, and the complex structure of multiple layers surrounding each line is shown in the illustration, with a simplified structure illustrated on the right as presented in the enlarged area.

[0026] Figure 7 This illustrates conformal coating (or conformal coating) via MAC. Figure 7 (a)-(c) show cross-sectional TEM and corresponding EELS plots, illustrating the conformal coating of trenches in silica;

[0027] Figure 8(a)-(c) show cross-sectional TEM and corresponding EELS plots, illustrating the conformal coating of cobalt wires utilizing the MAC layer; Figure 8 (a) shows an overview of a large-format scan with multiple lines covering the ML-AC; Figure 8 (b) shows an enlarged view of one of these lines; Figure 8 (c) shows an EELS overlay diagram, in which the individual layers are marked;

[0028] Figure 9 (a)-(b) show AFM images and thickness distribution curves over time of an 80 nm thick cobalt line on Si / SiO2 substrates without a MAC layer (left) and with a 1.5 nm MAC layer (right) after 72 hours of exposure to environmental conditions.

[0029] Figure 10 (a)-(b) show a set of 100 nm wide copper wires after exposure to 7% APS solution for 30 seconds (with and without ML-AC layer protection);

[0030] Figure 11 The dielectric and metallic diffusion barrier properties of directly grown MACs are shown. Figure 11 (a) shows low-frequency dielectric spectral data, where the dielectric permittivity κ ~ 1.34 is independent of the MAC thickness; Figure 11 (b) shows the capacitance as a function of frequency at different thicknesses, and the inset shows the impedance trajectory for MAC measurements, where the MAC capacitance does not change in the frequency range of 100 Hz to 100 kHz. Figure 11 (c) shows the dependence of breakdown voltage on dielectric thickness, indicating a dielectric strength of 28–31 MV cm. -1 ; Figure 11 (d) shows a copper ion diffusion experiment, demonstrating that when ln(TTF) ~ E and ln(TTF) ~ When modeled, the failure time is at least two orders of magnitude higher than the existing 10-year baseline (the triangle symbol (SiO2) illustrates the comparison of substrate behavior); Figure 11 (e) shows a comparison of linear model fits for the TTF of recently reported existing tunneling barrier alternatives (TaN, represented by a triangle pointing to the left; and hBN, represented by a triangle pointing to the right);

[0031] Figure 12 An optical characterization of the dielectric constant is shown; Figure 12 (a) shows optical ellipsometer measurements of the imaginary and real parts of the dielectric constant supporting the results obtained in the low-frequency range; Figure 12 (b)-(c) show dielectric spectroscopic fittings of ψ (Psi) and Δ (Delta) spectra acquired by the Cody-Lorentz-Urbach model at different incident angles (angles increasing from 40 degrees to 70 degrees in 5-degree steps from the top curve to the bottom curve, and other layers considered in the model being a silicon substrate and a reflective surface of silicon dioxide (with a thickness of 87.4 nm) as determined by cross-sectional TEM measurements); Figure 12 (d)-(e) show the UV-vis spectroscopic data of MAC at various thicknesses; Figure 12 (e) The data are presented in the form of Tauc plots, and the corresponding direct band gaps are extracted;

[0032] Figure 13 The results of breakdown and leakage current measurements in the MIM device are shown; Figure 13 (a) shows an optical image that illustrates a size of 50 x 50 µm. 2 d, 500x500 nm 2 h (changing the area of ​​the capacitor plates) of the MIM capacitor (Au metal plates); Figure 13 (b) and (f) respectively show the... Figure 13 The IV curves measured by the apparatus shown in (a) and 13(d) are as follows: circles (1L) represent single layers, crosses (2L) represent double layers, stars (3L) represent triple layers, and squares (4L) represent quadruple layers. The horizontal and vertical dashed lines in (b) represent 1.5 x 10 -2 A cm -2 (Low power limiting current density and 0.7V transistor operating voltage, and the acceptable transistor operating range is located in the lower right quadrant); Figure 13 (c) and (g) respectively show the results from Figure 13 Weibull plots of breakdown voltage extracted from (b) and 13(f). Figure 13 (h)-(j) show the IV curves for scaling the capacitor plate size for 1, 2, and 3 layers, respectively;

[0033] Figure 14 The results of breakdown and leakage current measurements using CP-AFM are shown; Figure 14 (a) shows a schematic diagram of the CP-AFM experiment; Figure 14 (b) shows the IV curves of MAC for layers 1, 2 and 3 on Au substrate (each with more than 100 data points, and the solid line indicates the average curve); Figure 14 (c) shows a Weibull plot of the data presented in 14(b). Detailed Implementation

[0034] As explained above, an improved dielectric material is needed.

[0035] In general, the present invention provides a dielectric material comprising a film comprising a layer of two-dimensional (2D) material, wherein the dielectric constant (κ) of the film is ≤ 3.0. It is desirable to use ultra-low κ materials (κ < 2.5) with a thickness of < 3 nm for next-generation integrated circuits (ICs). However, it is generally difficult to reduce the thickness of dielectric materials without increasing κ. The dielectric material of the present invention advantageously possesses low κ and low thickness, while also being non-porous, exhibiting excellent mechanical properties, and being corrosion resistant.

[0036] According to a first aspect, the present invention provides a dielectric material comprising a film comprising a layer of two-dimensional (2D) material, wherein the dielectric constant (κ) of the film is ≤ 3.0.

[0037] For the purposes of this invention, the use of the singular form includes the plural form unless specifically stated otherwise. It should be noted that, as used in the specification and appended claims, the singular forms “a,” “an,” and “the” include plural references unless the context clearly indicates otherwise. Furthermore, the use of the terms “comprising,” “including,” and “having,” as well as other forms such as “comprising,” “including,” and “having,” is not considered limiting.

[0038] For the purposes of this invention, the 2D material mentioned refers to a material having a single atomic layer thickness. This 2D material may have an in-plane amorphous structure.

[0039] The film may have a κ ≤ 3.0. For example, the film may have a κ ≤ 2.5. In particular, the film may have κ ≤ 2.0, κ ≤ 1.9, κ ≤ 1.8, κ ≤ 1.7, κ ≤ 1.6, κ ≤ 1.5, κ ≤ 1.4, κ ≤ 1.3. Even more particularly, the film may have a κ ≤ 1.3. The film may comprise one or more layers of 2D material. According to one particular aspect, the dielectric material may comprise the film and any suitable non-2D material having a κ ≤ 3.0. According to another particular aspect, the dielectric material may comprise the film and any suitable non-2D material having a κ ≤ 3.0. The non-2D material may comprise a wafer or any suitable non-catalytic substrate.

[0040] The 2D material can be any suitable amorphous 2D material. The 2D material can have a κ ≤ 3.0. According to a particular aspect, the 2D material can include monolayer amorphous carbon (MAC). For the purposes of this invention, MAC is defined as an analogue of monolayer crystalline carbon (graphene), wherein it is predominantly sp2 carbon with randomly oriented in-plane bonds, where π bonds are broken and the relative contribution of σ bonds to material properties is increased, and the low-polarity carbon bonds in the disordered structure minimize the total polarizability. Since MAC contains only carbon, diffusion problems are advantageously absent, and it is compatible with many different substrates. MAC can have any suitable interlayer spacing. For example, MAC can have an interlayer spacing of 0.6–0.9 nm. In particular, the interlayer spacing can be 0.65–0.85 nm, 0.65–0.8 nm, or 0.7–0.75 nm. The interlayer spacing can be approximately twice the spacing of graphene layers. In each layer of 2D MAC, the number of carbon atoms in MAC can be similar to the number of carbon atoms in each monolayer of its analogue, monolayer crystalline carbon (graphene). Therefore, the density of MAC can be about half that of graphene. Specifically, the density of MAC can be 0.6–1.5 g cm⁻¹. -3 Therefore, using MAC or any other suitable amorphous 2D material (such as monolayer amorphous boron nitride) is beneficial for reducing material density while allowing for low κ.

[0041] The films included in the dielectric material can have any suitable thickness. For example, the film thickness can be ≤ 20 nm. In particular, the film thickness can be ≤ 10 nm, ≤ 9 nm, ≤ 8 nm, ≤ 7 nm, ≤ 6 nm, ≤ 5 nm, ≤ 4 nm, ≤ 3 nm, ≤ 2 nm, or ≤ 1 nm. Even more particularly, the film thickness can be ≤ 3 nm. The film thickness can be 0.5-3 nm, 0.65-2.6 nm, 0.8-2.4 nm, 1.3-2.1 nm, or 1.5-2 nm. The film can include any suitable number of 2D MAC layers. For example, the film can include one or more 2D MAC layers. In particular, the film can include one, two, three, four, or five 2D MAC layers. Based on the scaling technology nodes and dimensions for optimal performance, various specific thickness requirements exist, and the dielectric material of the embodiments of the present invention advantageously allows for the growth of different thicknesses by adjusting the number of single-atom-thick MAC layers.

[0042] Each of one or more layers in a 2D MAC can have κ ≤ 3.0. Therefore, the membrane can have κ ≤ 3.0, as described above.

[0043] The films included in the dielectric material can have suitable hardness. For example, the hardness of the film can be >10 GPa. Specifically, the hardness of the film can be >20 GPa, >30 GPa, >40 GPa, >50 GPa, >60 GPa, >70 GPa, >80 GPa, >90 GPa, >100 GPa. Even more specifically, the hardness of the film can be about 100 GPa. This mechanical stability enables the fabrication of standard integrated circuit devices without dielectric material collapse.

[0044] As explained above, in integrated circuits, reducing the thickness of the dielectric layer typically leads to higher leakage current (degraded performance). The dielectric material of the embodiments of the present invention advantageously enables device scaling while maintaining higher dielectric strength to prevent increased leakage current. In particular, the film can have high dielectric strength, thereby enabling it to meet requirements down to a single atomic layer thickness. For example, the dielectric strength of the film can be > 8 MV cm⁻¹. -1 The dielectric strength of the film can be > 10 MV cm⁻¹ -1 >15MV cm -1 > 20 MV cm -1 > 25 MV cm -1 > 30 MV cm -1 Even more specifically, the dielectric strength of the film can be > 30 MV cm⁻¹. -1 .

[0045] The films included in the dielectric material can be non-porous. For the purposes of this invention, "non-porous" means that there are no pores on or within the film. The absence of pores may, but is not limited to, the absence of pores or voids with a size of about 1 nm or larger. Non-porous films advantageously allow for improved resistance to degradation caused by moisture absorption or ion diffusion into the film. Furthermore, because the film resists such degradation, additional barrier layers (to prevent metal ion diffusion) and liner material layers (for adhering the dielectric material to the device structure) are unnecessary, as these layers further limit the scalability of integrated circuits. Simplified device architectures with metals in direct contact with the dielectric material can be achieved, allowing for better interconnect performance, which can utilize larger volumes for metal lines to improve line conductivity or utilize more aggressive scalability.

[0046] According to a particular aspect, the film can be formed on at least a portion of a non-catalytic substrate. For the purposes of this invention, the term "non-catalytic substrate" refers to any suitable substrate that does not participate in the growth chemistry. In particular, the substrate can be considered non-catalytic and not participate in the growth chemistry under the temperature conditions used to form the film, even if it may be catalytically active at a temperature higher than the temperature used to form the film. The non-catalytic substrate may not have κ ≤ 3.0. Examples of suitable non-catalytic substrates include, but are not limited to, silicon-based substrates, carbon-based substrates, metal-based substrates, oxides, transition metal dichalcogenides, mXene, or any combination thereof. In particular, the non-catalytic substrate may include cobalt, gold, copper, nickel, tungsten, molybdenum, ruthenium, niobium, silicon dioxide, silicon nitride, titanium nitride, tantalum nitride, cobalt oxide, tungsten carbide, germanium, gallium arsenide (GaAs), silver, stainless steel, Fenikol alloys, manganese, aluminum, or any combination thereof. Even more particularly, the non-catalytic substrate may include silicon dioxide, silicon nitride, titanium nitride, copper, cobalt, and tungsten.

[0047] According to a second aspect of the invention, a method for forming the dielectric material of the first aspect is provided, the method comprising depositing carbon radicals on a non-catalytic substrate. For the purposes of this invention, the carbon radicals referred to are carbon species having one unpaired electron and being readily reactive with other atoms or molecules. Deposition can be performed by any suitable method. For example, deposition can be performed by chemical vapor deposition (CVD). In particular, deposition can be performed by laser plasma-enhanced chemical vapor deposition (LPE-CVD), laser CVD (LCVD), UV lamp-assisted CVD, or UV lamp plasma-assisted CVD.

[0048] Carbon radical deposition can be sustained for any suitable amount of time to enable the formation of one or more layers of 2D material on a substrate, thereby forming a dielectric material. For example, carbon radical deposition can be sustained for >1 minute. Specifically, carbon radical deposition can be sustained for >1.5 minutes, >2 minutes, >2.5 minutes, >3 minutes, >3.5 minutes, >4 minutes, >4.5 minutes, >5 minutes, >5.5 minutes, >6 minutes, >6.5 minutes, >7 minutes, >7.5 minutes, >8 minutes, >8.5 minutes, >9 minutes, >9.5 minutes, >10 minutes, >11 minutes, >12 minutes, >13 minutes, >14 minutes, >15 minutes, >16 minutes, >17 minutes, >18 minutes, >19 minutes, or >20 minutes. According to a particular aspect, the method may include repeating the deposition to form up to five layers of 2D MAC.

[0049] Deposited carbon radicals can be continued at any suitable step size increment for any suitable time until the desired number of layers is achieved. For example, deposited carbon radicals can be generated from t iMinutes (initial time) increase to t f The time (final time) is measured in minutes, in increments of 1 minute, 2 minutes, or any combination thereof. Specifically, the step size increment can vary in alternating intervals (such as 1 minute, 2 minutes, 1 minute, 2 minutes) and is repeated until time t is reached. f This will mean smaller step sizes, but at the same time, increased uniformity across each layer.

[0050] The deposition of carbon radicals can include the formation of carbon radicals via photodissociation of a carbon source. Photodissociation of the carbon source can be carried out via any suitable light source having any suitable wavelength. For example, the light source can be an excimer laser, a halogen lamp, a diode lamp, a diode laser, or a gas laser. For example, the wavelength of the light source can be 0.01-2500 nm. Specifically, the wavelength of the light source can be 0.01-0.1 nm, 0.1-1 nm, 0.1-2000 nm, 1-1500 nm, 10-1000 nm, 100-500 nm, 150-450 nm, 200-400 nm, or 250-350 nm. In particular, the deposition of carbon radicals can include the formation of carbon radicals via absorption at a UV wavelength of the carbon source. This UV wavelength can be any suitable UV wavelength that causes the formation of carbon radicals when the carbon source is exposed to this UV wavelength. For example, the UV wavelength can be 200-400 nm. The UV wavelength can be generated by any suitable device. For example, the UV wavelength can be generated by an excimer laser or a UV lamp. The excimer laser source can include XeCl or KrF. The carbon source can be any suitable carbon source capable of forming carbon radicals under appropriate conditions. For example, the carbon source can include acetylene, methane, ethylene, ethanol, propane, anisotropic carbon, or any combination thereof. In particular, the carbon source can include acetylene. By using acetylene as the carbon source, the concentration of C2 radicals is advantageously increased, and the self-limiting decomposition behavior is disrupted, which allows for the growth of multiple layers of MAC without the need for a catalyst.

[0051] Carbon radical deposition can occur at any suitable pressure based on the carbon source chosen. For example, carbon radical deposition can occur at 10... -3 Up to 10 -2 It was carried out under pressure.

[0052] According to one particular aspect, the substrate may include a non-catalytic substrate. The non-catalytic substrate can be as described above. The substrate can be considered non-catalytic and not involved in growth chemistry at temperatures of about 20-500°C, even if it may be catalytically active at higher temperatures. Specifically, the substrate can be considered non-catalytic at temperatures of about 50-450°C, 100-400°C, 150-350°C, and 200-300°C. Even more specifically, the substrate can be considered non-catalytic at temperatures of about 250-350°C.

[0053] The substrate can be directly exposed to the UV wavelengths described above. Alternatively, direct exposure of the substrate to UV wavelengths can be avoided to prevent potential surface damage.

[0054] Carbon radical deposition can be performed in a plasma environment. The plasma environment advantageously provides additional energy to the carbon radicals. Therefore, a higher proportion of carbon radicals will be within the optimal energy range for photolysis by a specific wavelength UV source, thereby increasing the concentration of active carbon radicals and thus facilitating deposition on the substrate. According to a particular aspect, the method may also include generating a plasma environment prior to deposition. For example, generating a plasma environment may include remotely inductively coupled plasma.

[0055] Therefore, the method described is an improved approach that allows for direct and conformal growth of films. This is particularly important for applications in integrated circuits, where the film must cover all peripheries of trenches and vias, and thus needs to be able to grow continuously and connect on all sides of any high aspect ratio device structure (such as, but not limited to, pre-patterned columnar structures and trenches). The method allows for the elimination of lining materials typically used to facilitate conformal growth, which is not ideal in many deposition methods. Furthermore, film growth occurs throughout the entire volume, thus eliminating the need for direct surface irradiation using excimer lasers, a crucial characteristic for any further industrial applications.

[0056] The invention has now been generally described and will be more readily understood by reference to the embodiments provided below, which are provided by way of example and are not intended to be limiting.

[0057] Example

[0058] Materials and methods

[0059] Non-catalytic direct growth is achieved by using 10 -8 Up to 10 -7The process was performed in a chemical vapor deposition (CVD) chamber operating at a base pressure level of millibars (mbar). A UV excimer XeCl laser (λ = 308 nm) was used to initiate the photodecomposition of the carbon source (acetylene, C2H2), and Ar plasma was introduced into the device using a remote inductively coupled plasma (PIE Scientific) source. Samples were mounted on stainless steel supports (two operating modes: direct and indirect laser exposure). For Cu foil, silicon, and Si / SiO2 substrates, the sample surface was directly exposed to the excimer laser. For fabrication of the same substrates or devices in indirect exposure mode, direct laser radiation was avoided to prevent potential surface damage. Since the carbon source also dissociates when the laser is near the sample but not in contact with it, the same conditions could be used, but with gradual heating up to 300°C to support growth.

[0060] The growth of monolayer amorphous carbon was achieved by excimer laser photodecomposition of the carbon source. Acetylene was used to increase the concentration of C2 radicals, and remote inductively coupled plasma was added, which helped to improve the concentration and activity of carbon radicals and promote deposition on a substrate (e.g., silica).

[0061] Similar results for MAC synthesis were also obtained by using a UV lamp source instead of an excimer laser (XeCl or KrF source).

[0062] Characterization

[0063] Utilizing any location on a uniformly covered 2-inch wafer ( Figure 1 (d) Cross-sectional transmission electron microscopy (TEM) image acquired Figure 1 (a)-(c)) and through electron energy loss spectroscopy (EELS, Figure 1 Elemental mapping of illustrations (a)-(c), atomic force microscopy (AFM) Figure 1 (e) and Figure 2 Different MAC thicknesses were examined using ellipsometry and TEM measurements. More than 10 samples were measured in cross-sectional TEM and more than 20 samples were measured by AFM and ellipsometry, and the thicknesses corresponded to layered MAC. Figure 1 (a)-(c) show three representative samples with thicknesses corresponding to 3, 2, and 1 layers of MAC (3L, 2L, and 1L, respectively). The EELS spectra shown in the insets of each figure enhance the accuracy of the thickness measurements, with red corresponding to the sp2 shoulder of the carbon peak and blue corresponding to oxygen. Discrete variations in thickness were observed: from 2.1 nm in the 3L sample (… Figure 1 (a)), to 1.45 nm of the 2L sample ( Figure 1 (b)), and decreased to 0.8 nm in a single layer ( Figure 1 (c)). The measured thickness interval of 0.65 nm is consistent with the interlayer spacing, and the single-layer thickness is consistent with previous literature. Figure 1 (e) shows the AFM thickness measurements of 1–4 L MAC films grown directly on SiO2. The order of the data points and their labels correspond to the thickness measured by AFM—from bottom to top, with growth times increasing from 1 minute to 12 minutes (in 1-minute increments).

[0064] Atomic-resolution TEM was performed on the suspended sample transferred onto a Si3N4 TEM lattice to reveal the amorphous structure. The layer-by-layer growth of ML-AC is shown in top-view images of ML-AC grown at different synthesis times. Figure 1 (f)-(j)). Figure 1 (f), 1(h), and 1(j) correspond to the complete 1L, 2L, and 3L MAC layers, respectively, while Figure 1 (g) and 1(i) show the intermediate growth times during the growth process of the second layer (1-2L) and the third layer (2-3L), respectively. Three representative examples are shown: monolayers transferred from SiO2 substrates ( Figure 1 (as shown in (k)); the bilayer transferred from SiO2 ( Figure 1 (as shown in (l)), and the three layers transferred from SiO2 ( Figure 1 (as shown in m). Furthermore, atomic resolution TEM images reveal connected but twisted carbon rings composed of varying numbers of atoms. Localized moiré fringe is visible in the overlapping nanocrystalline regions within each layer of the multilayer sample. Despite their lack of periodicity, these structures advantageously lack any pores or defects.

[0065] A detailed analysis of the thickness variation over time of a sample with patterned trench assemblies etched in a MAC was performed using AFM (see [link to AFM]). Figure 2 During the first 2 minutes of growth, the thickness remained constant in the sub-nanometer range. Figure 2 (a) and 2(b)). Within the following 2 minutes, the thickness increased by 0.65 nm, but the thicknesses of the two samples remained similar. This may be because AFM could not distinguish the island-like structure of the sample grown for 3 minutes, while the sample grown for 4 minutes had a complete second layer. Furthermore, samples corresponding to growth times of 5, 6, 7, and 8 minutes ( Figure 2 Samples (e), 2(f), 2(g), and 2(h)) all had a thickness of approximately 2.1 nm, and the top layer became more complete with increasing growth duration. Finally, for the last group of samples grown for 9, 10, 11, and 12 minutes... Figure 2(i), 2(j), 2(k), and 2(l)), the measured average thickness is 2.7 nm, corresponding to 4 layers. These data were used to plot... Figure 1 (e).

[0066] The MAC grown directly on SiO2 was characterized at the maximum cm scale to evaluate its uniformity. A 4” diameter Si / SiO2 (90 nm thick SiO2) wafer was covered with a double layer of MAC. Figure 1 Optical photographs of the wafer are shown in (d), demonstrating visually uniform thickness and clear contrast between the MAC-covered area and the substrate. Raman spectroscopy (10,000 points) was used to confirm its amorphous nature and uniformity down to the micrometer level. Figure 3 (a) shows the averaged spectra on the wafer surface at various growth durations. It is also noteworthy that the spectra from Cu and Co surfaces are similar to those observed on SiO2, and therefore the surface does not affect the growth mechanism (e.g., Figure 3 (As seen in (b) and 3(c)). Figure 3 As seen in (b), the Raman spectrum of the G peak intensity demonstrates the continuity of the formed MAC film on a large scale. It can be seen that I... G The metal lines are uniformly distributed on the patterned surface of SiO2. Figure 3 (c) shows a single spectrum, further confirming the surprising technical effect of the embodiments of the invention in allowing surface-independent deposition (with only limited variation) on Cu and / or Co surfaces.

[0067] X-ray photoelectron spectroscopy (XPS) was used to investigate the elemental composition, chemistry, and electronic states of atoms in the MAC film and substrate used for growth. XPS confirmed that the sample was sp2 hybridized over a large area, consistent with local findings obtained by EELS. Figure 3 (d) shows the C 1s core level spectra captured at various growth times (times increased from 1 minute to 12 minutes, in 1-minute increments, from bottom to top). Figure 3 (e) shows the C 1s core level spectra captured from MAC grown on different substrates. These confirm the sp2 hybridization of carbon, with negligible or near-zero sp3 hybridization contributions. These data, combined with the core level spectra of the substrate materials ( Figure 4 (e)-(g)) also indicates that no bonds were formed with the substrate. From Figure 4 (e)-(g) show that for the substrate formed below the MAC, there is neither bonding nor the formation of carbides or any other compounds.

[0068] Near-edge X-ray absorption fine structure (NEXAFS) was used to study the σ and π bond system to demonstrate the layered nature of the MAC, and the results were presented in... Figure 3 As shown in (f). The observed behavior is significantly different from that of graphene and more similar to that of amorphous systems, where the 1s-π* transition is independent of the incident angle of the linearly polarized X-ray beam. Figure 3 As shown in (f), the 1L and 4L layers of MAC produce a consistent total electron yield as photon energy increases at different angles (intervals of 30 degrees).

[0069] According to EELS spectrum ( Figure 4 (a) For all samples, a strong sp2 shoulder peak (indicated by the vertical dashed line) was observed at 284.4 eV, with an intensity similar to that of the main carbon peak. This indicates that the MAC structure remains unchanged for all reported thicknesses of 1L, 2L, and 3L, i.e., it is predominantly sp2 carbon. It can be seen that, despite the different growth times (1b, 2b, and 3d, respectively), the intensity of the peak relative to energy loss consistently corresponds for each sample at 1L, 2L, and 3L.

[0070] Representative points marked by dots of different shapes Figure 4 (b) are labeled as p1, p2, and p3, and the corresponding Raman spectra are in Figure 4 As shown in (c), the three identified locations on the wafer have the same spectrum, whose I D / I G The ratio is close to the expected value of 0.85. This further demonstrates the surprising effect that the signal is uniform and consistent across the entire 4-inch wafer. Therefore, growth is not limited by wafer size, as these results also apply to 1-inch wafers or even larger wafers, such as 8-inch wafers.

[0071] AFM is used to measure thickness uniformity at the microscale, such as Figure 4 As shown in (d), the AFM image shows no obvious features, and the roughness value of ~200 pm is consistent with that of SiO2. No adsorption layer, island structures, or clusters were observed (unlike conventional amorphous thin films), and the roughness distribution is consistent across the entire wafer. The histograms for the height distribution (surface roughness) of MAC on SiO2 and on the SiO2 surface are difficult to distinguish, confirming the conformal growth of MAC on SiO2.

[0072] Nanoindentation AFM mode was also used to probe the hardness of the deposited MAC, which is a key mechanical performance standard for its application as a dielectric material in metal interconnects. Figure 5As shown, its hardness is at least an order of magnitude higher than that of silicon dioxide. Unlike other existing materials, the unique combination of high hardness and low κ makes MAC a suitable candidate material for back-end-of-line integration.

[0073] application

[0074] Potential applications for metal interconnects were evaluated, such as Figure 6 As exemplified in the illustration. In addition to the surrounding low-k dielectric material, the interconnect typically has a diffusion barrier layer and a liner material layer, as shown in the enlarged illustration. Such an example covers a key requirement of the semiconductor industry for low-k dielectric materials, namely, conformal deposition of multiple materials simultaneously on non-planar surfaces.

[0075] To demonstrate the ability to conformally cover non-planar surfaces, 100 nm wide trenches were fabricated in silicon dioxide using electron beam lithography and fluorine-based plasma etching. Furthermore, these trenches were covered with MAC and analyzed using cross-sectional TEM combined with EELS, such as... Figure 7 As shown in (a) and 7(b), representative data for a two-layer MAC (1.5 nm) are presented. Figure 7 (a) shows a large-format scan of multiple trenches simultaneously. Each trench provides a uniform conformal coating, in which each different material interacting with each other grows uniformly. Figure 7 (b) provides a magnified view of one of the trenches, and SiO2, gold (Au), and platinum (Pt) can be seen. Figure 7 The magnified EELS pattern in (c) allows for thickness determination due to material contrast, where the region marked as SiO2 represents the signal from oxygen (corresponding to the SiO2 layer), and the line marked with an arrow represents the sp2 carbon EELS shoulder. The observed thickness is uniform throughout the trench, on its top, bottom, and sidewall surfaces. The 2L ML-AC is 1.45 nm thick and uniformly distributed along the interface between SiO2 and Au. Material deposition uniformity is crucial for reliable electronic performance, and it is desirable to avoid regions with thickness inhomogeneities that will lead to variations in electrical performance. Therefore, this example demonstrates that uniform deposition of the material at a thickness of 1.45 nm advantageously improves the uniformity of electrical performance, resulting in enhanced reliability of electronic performance.

[0076] This further confirms that growth also occurs simultaneously on different materials. For example... Figure 8 As shown in (a), MAC photolithography defines the surface growth of cobalt lines with a width of 100 nm, a height of 60 nm, and a spacing of 0.5 μm on a Si / SiO2 substrate. The angle between the sidewalls of the cobalt lines and the silicon dioxide substrate is approximately 90 degrees, corresponding to a spacing of 0.6 nm.-1 Extremely high curvature levels. Perfect deposition in such highly complete regions is challenging for 2D materials due to the lattice-limited possible curvature, which is overcome by the methods and / or materials of this invention. (From cross-sectional TEM) Figure 8 (b) and EELS Figure 8 The data in (c) demonstrate that the cobalt wires utilize a conformal coating of a MAC layer that transitions uniformly to the SiO2 surface through the corner. The 2L ML-AC is 1.45 nm thick and is uniformly distributed along the interface between SiO2, Co, and Au.

[0077] The resulting film advantageously contains no cracks or other defects. To demonstrate this, particularly on the sidewalls, tests were conducted utilizing cobalt's high sensitivity to oxygen and water. Two sets of cobalt electrodes (with and without the MAC layer) were left in air for 72 hours. Figure 9 As shown in (a)-(b), the morphology of the original line and the protected line were analyzed using AFM, respectively. Figure 9 In the profile of (a), a significant expansion of approximately 1.5 times due to oxidation is clearly visible. In contrast, the thickness of the protected line remains unchanged. Therefore, it can be seen that the cobalt line expands due to oxidation without MAC protection. The height (h) is approximately 150 nm. When the cobalt line is protected with MAC, even if there are chemical reactions acting on the MAC-protected cobalt line, the sample does not degrade over time, as shown in... Figure 9 (b) shows the bottom view. The height (h) is approximately 75 nm. Based on these data, it is clear that MAC deposition forms a continuous film that is impermeable to water and oxygen molecules. Similar results were obtained on Cu lines by testing the chemical stability against commonly used copper etchants. Figure 10 (a) and (b)).

[0078] The potential of MAC as an ultra-low κ dielectric and diffusion barrier layer was further evaluated. To demonstrate its low κ value, two independent experiments were performed: impedance spectroscopy in the low-frequency range and ellipsometer measurements in the optical range. The dielectric constant was measured using an electronic device—a set of capacitors with different dielectric layer thicknesses. The frequency dependence of the sample impedance in the 100 Hz to 100 kHz range was captured, and the dielectric constant was further extracted by fitting the data to an LR / C circuit (see [link to relevant documentation]). Figure 11 (a) and (b)). For each thickness, at least 30 different samples had a κ value of 1.3 that remained constant over a frequency range of 1–100 kHz. Thus, it can be seen that the dielectric constant remained consistent across the entire thickness of the MAC layer, from 1L to 4L.

[0079] In summary, these two independent techniques confirmed that MAC exhibits ultra-low κ values ​​in thicknesses ranging from 0.6 to 3 nm. The ultra-low κ values ​​are achieved due to the amorphous structure and single-element carbon properties of MAC. A distinct thickness independence was uniquely observed for this material.

[0080] This is through reference Figure 3 The NEXAFS data shown in (f) support this. The width of the spectral lines corresponding to the 1s-π transitions, and the overall weak dependence on the incident angle, typically reflect the degree of disorder, the presence of wrinkles, and the buckling of the layers when compared to graphene. The tailing effect of the 1s-σ transitions and the weaker sensitivity to the incident angle also indicate the random orientation of the in-plane bond arrangement. π bonds typically involve delocalized electrons that move freely within the structure—in crystalline graphene, they form “π clouds” above and below the plane of carbon atoms, thus contributing significantly to electronic conduction. In MAC, this π network is disrupted, thus increasing the relative contribution of σ bonds to the dielectric constant. The dipoles associated with both types of bonds appear to be negligible, and these data indirectly explain the observed ultra-low dielectric constant.

[0081] In addition to parasitic capacitance crosstalk, the dielectric material used in the metal interconnect stack should also block any breakdown leakage current between the conductive lines and the active semiconductor components. To verify this, the tunneling IV curve was measured using the conductivity AFM of the MAC on the metal pads, and the dielectric strength was analyzed. Furthermore, stress-voltage scans were performed in the capacitors used to extract the dielectric constant. The measured dielectric strengths reached 28–31 MV cm⁻¹. -1 This is the highest reported for both 2D and 3D materials. The data is shown in... Figure 11 (c) shows (Cap = Capacitor = 28 MV cm) -1 And C-AFM = Conductive Atomic Force Microscopy = 31 MV cm -1 ).

[0082] The metal (Cu) interdiffusion barrier properties of the MAC layer were measured by statistical analysis of the current change over time under various applied bias voltages. Figure 11 (d)). The failure time (TTF) determined by the linear ln(TTF) ~ E model under the operating electric field (~0.5 MV / cm) is given as 10. 12 The failure time of s is at least two orders of magnitude longer than that of all recently reported and commonly used materials. Figure 11(e)). The negligible change observed in the MAC between 1L and 2L indicates that the metal diffusion failure mechanism is not limited by the MAC. The results demonstrate that the MAC has additional properties that make it well-suited for novel advanced architectures for interconnect stacks (an architecture in which the low-k dielectric also acts as a metal ion diffusion barrier layer).

[0083] The low-frequency κ value was obtained Figure 12 (a) shows the support from independent ellipsometer measurements. The Cody-Lorentz-Urbach model was used to process the Ψ-Δ spectra acquired at seven different incident angles. Figure 12 (b)-(c)). The increase in dielectric constant within the short UV wavelength range coincides with the increase in absorption observed in this range (see...). Figure 12 (d)-(e)). Towards lower frequencies, the real part of the dielectric constant gradually decreases to approximately 1.3. For MAC, it is normal for its optical dielectric constant to be slightly higher than that in the low-frequency range.

[0084] Figure 13 The results of the breakdown IV curves are summarized in the paper. The same apparatus used previously to measure the breakdown voltage was employed (see [link to paper]). Figure 13 (a), (d), and (e)). Breakdown voltage in a large device (50x50 µm) 2 , Figure 13 (b)-(c)) and small devices (500x500nm) 2 , Figure 13 (f)-(g)) show consistency, and also exhibit a high degree of consistency for different thicknesses and different capacitor plate sizes. Figure 13 (h)-(i)-(j), representing 1L, 2L, and 3L respectively. The results show that leakage current and breakdown voltage do not significantly depend on the plate area. These results were also obtained from conductive probe AFM IV curves collected from gold metal plates covered with MAC of varying thicknesses. Figure 14 (b)) support.

[0085] Applications in semiconductor integrated circuits

[0086] MAC is an ideal ultra-low κ dielectric material that meets the requirement that dielectric material κ < 2 (MAC has κ = 1.3) and is a perfect barrier layer at 0.6 nm (currently existing ultra-low κ materials are porous and require a barrier layer to prevent metal diffusion).

[0087] Therefore, the width of the metal wire core is maximized by improving the device structure.

[0088] By utilizing a MAC barrier layer, since MAC is an ideal dielectric material for "space width," the liner + barrier layer is no longer part of the interconnect width. Increased metal core width can be achieved by combining the MAC dielectric with existing ULK node materials.

[0089] MAC can also be grown thicker to completely replace the ULK dielectric material filling the "gap width." Since MAC is superior to existing dielectric gap width materials, the gap width can be reduced. This further maximizes the volume of the metal wire for improved conductivity. Alternatively, the metal volume can remain constant at optimal dimensions while allowing overall interconnect scaling to focus on reducing the gap width. With a significant increase in interconnect metal volume, the bottleneck problem of interconnect scaling can be solved.

[0090] Applications in spacers for Gate All-Around (GAA) transistors and FinFETs

[0091] The high parasitic capacitance makes the spacing between the gate and source / drain contacts a major limiting factor for such transistors. The IRDS roadmap indicates that while first-generation GAA transistors can be implemented with a 6 nm spacer width and κ = 3.3, no solution exists to reduce this to 4 nm and κ = 2.7. With MAC, a reliable GAA transistor architecture is possible at 2.1 nm and κ = 1.3. Furthermore, the spacer width can be further reduced to up to 0.6 nm (and the thickness reaches 4–6 nm across the entire spacer width range) if desired. Table 1 shows the possible dimensions for replacing the spacer material with MAC.

[0092]

[0093] Table 1: Dimensions of spacer material replaced by MAC

[0094] Other advantages of MACs include ease of etching and selective etching for high lithographic resolution, and carbon-based MACs are highly favored in the semiconductor industry due to this property. Because of the etching chemistry of amorphous carbon, it can also be used as an etch protectant to fabricate with extremely low line and sidewall roughness. The structure of a MAC does not degrade and will provide good performance after the fabrication process.

[0095] While exemplary embodiments have been described above, those skilled in the art will understand that many modifications can be made without departing from the invention.

Claims

1. A dielectric material comprising a film comprising a layer of two-dimensional (2D) material, wherein the dielectric constant (K) of the film is < 3.

0.

2. The dielectric material of claim 1, wherein the 2D material comprises a monolayer of amorphous carbon (MAC).

3. The dielectric material of claim 1 or 2, wherein the thickness of the film is < 20 nm.

4. The dielectric material of any preceding claim, wherein the thickness of the film is 0.5-3 nm.

5. The dielectric material of any of claims 2-4, wherein the film comprises at least two layers of 2D MAC.

6. The dielectric material of any of claims 2-5, wherein the film comprises two to five layers of 2D MAC.

7. The dielectric material of any preceding claim, wherein the hardness of the film is > 10 GPa.

8. The dielectric material of any preceding claim, wherein the film has a dielectric strength > 8 MV cm -1 .

9. The dielectric material of any preceding claim, wherein the film is non-porous.

10. The dielectric material of any preceding claim, wherein the film is formed on at least a portion of a non-catalytic substrate.

11. The dielectric material of claim 10, wherein the non-catalytic substrate comprises a silicon-based substrate, a carbon-based substrate, a metal-based substrate, an oxide, a transition metal dichalcogenide, a mxene, or any combination thereof.

12. The dielectric material of claim 10, wherein the non-catalytic substrate comprises cobalt, gold, copper, nickel, tungsten, molybdenum, ruthenium, niobium, silicon dioxide, silicon nitride, titanium nitride, tantalum nitride, cobalt oxide, tungsten carbide, germanium, gallium arsenide (GaAs), silver, stainless steel, Fe-Ni-Co alloy, manganese, aluminum, or any combination thereof.

13. A method of forming the dielectric material of any preceding claim, the method comprising depositing carbon radicals on a non-catalytic substrate.

14. The method of claim 13, the method comprising repeating the depositing carbon radicals to form up to five layers of 2D MAC.

15. The method of claim 13 or 14, wherein the depositing carbon radicals comprises forming carbon radicals via photodissociation of a carbon source.

16. The method of any of claims 13-15, wherein the depositing carbon radicals comprises forming carbon radicals via UV wavelength absorption by a carbon source.

17. The method of claim 16, wherein the UV wavelength is 200-400 nm.

18. The method of any of claims 15-17, wherein the carbon source comprises acetylene, methane, acetylene, ethylene, ethanol, propane, indefinite carbon, or any combination thereof.

19. The method of any of claims 13-18, wherein the non-catalytic substrate comprises a silicon-based substrate, a carbon-based substrate, a metal-based substrate, an oxide, a transition metal dichalcogenide, a mxene, or any combination thereof.

20. The method of any of claims 13-19, wherein the non-catalytic substrate comprises cobalt, gold, copper, nickel, tungsten, molybdenum, ruthenium, niobium, silicon dioxide, silicon nitride, titanium nitride, tantalum nitride, cobalt oxide, tungsten carbide, germanium, gallium arsenide (GaAs), silver, stainless steel, Fernico alloy, manganese, aluminum, or any combination thereof.

21. The method of any of claims 13-20, wherein the depositing is performed in a plasma environment.

22. The method of claim 21, further comprising generating the plasma environment prior to the depositing.

23. The method of claim 21 or 22, wherein the plasma environment comprises a remote inductively coupled plasma.

Images