Patent Application
20250214331
The present disclosure relates to the field of material science and advanced material engineering, and, more specifically, the present disclosure relates to a method of transferring a graphene layer onto a substrate or support structure, such as a sample support for use in (e.g. transmission) electron microscopy.
Graphene possesses useful properties in terms of structure, thermal and electrical conductivity and particularly thickness, e.g. enabling layers of only a single atom thick. Its large surface-to-volume ratio, unique optical properties, high carrier mobility and density and other attributes may be particularly useful as well. These properties can be used in various applications, e.g. as a sample support in the field of electron microscopy.
The field of transmission electron microscopy (TEM) has significantly evolved in recent years. For example, aberration correctors have improved the attainable resolution, and stability and sensitivity have been drastically improved by novel stages and detector designs. However, significant room for improvement remains in the development of sample holders and supports. The sample support is, ideally, as thin and clean as possible, so the interference with the sample signal is limited and known. A sample support grid conventionally comprises a flat surface of a continuous layer, a layer with holes or a mesh. Depending on the sample type, the sample is deposited on the support by a drop-casting procedure, or thin sections of the sample are deposited on the grid via different sample preparation methods. Depending on the nature of the sample, the sample is analyzed on the edge of a hole or on top of a support layer, which allow electrons to pass through.
Conventionally, amorphous carbon-based supports are used, which may typically have a thickness of about 20 nm. However, such sample supports can cause a significant and undesirable background signal, which may be particularly problematic, e.g. for particular types of sample and/or particularly thin samples since they show poor contrast. A poor manufacturing reproducibility of these supports can furthermore easily lead to failed experiments, or at least a waste of time, resources and/or samples, which might be avoidable if only higher quality supports would be readily available.
Graphene films may be a good candidate for such high-quality TEM sample supports. A single layer of graphene, along its crystalline structure, can, at least presently, be considered as the thinnest material conceivable, such that the background signal from this layer, when used as a sample support, can be desirably low. Moreover, this background signal, when the support is sufficiently uniform and homogeneous, can be easily subtracted from the detected signal since graphene will yield a periodic pattern that can be easily distinguished.
Graphene has also an aspect of high thermal and electrical conductivity, such that it may be ideal for use as support in the analysis of beam-sensitive materials, e.g. reducing the most common damage processes in the TEM by allowing a good dispersion of currents and heat. This, in combination with its thickness, enables a good optimized characterization of beam sensitive materials, including soft-hard interfaces and organic compounds such as capping ligands in the TEM.
It is to be noted that a graphene support is not only useful in TEM imaging, but also in a wide range of other transmission techniques such as electron diffraction techniques, energy-dispersive X-ray spectroscopy (EDX) and electron energy-loss spectroscopy (EELS), in which the signal-to-background ratio can also be improved substantially by reducing the background signal received from the support. Graphene can also be useful for many more applications outside the field of (charged particle) microscopy, such as sensors and microelectromechanical systems. For example, its large surface-to-volume ratio, optical properties, (thermal and electrical) conductivity and high carrier mobility and density can be greatly beneficial for sensor functions.
Unfortunately, graphene TEM grids that are presently commercially available are relatively expensive consumables. More importantly, the graphene of such grids typically fails to achieve a perfect coverage, i.e. does not form a single intact layer over the entire grid area, and areas without poly-methyl methacrylate (PMMA) contamination, which is used in the transfer process, are considerably rare, e.g. the cleanness of the sample support surface is open for improvement. The graphene layer may even show wrinkles, while ideally it should be flat. A clear need exists in the art for a method of manufacturing a sample support having a reproducibly clean, flat, continuous, uniform and homogeneous graphene layer, which has proven to be considerably challenging.
To obtain graphene-based TEM grids, graphene is typically grown on a metallic surface, e.g. copper, by chemical vapor deposition (CVD), from which it is transferred to a desired target using either polymer-based or polymer-free transfer methods. However, depending on the method to transfer the graphene, the quality, the intactness and the amount of remaining organic contaminants can hamper the quality of measurements when used as support in the TEM.
The graphene intactness and purity are strongly dependent on this transfer method, which eventually determines the quality and is therefore a limiting factor for some applications. Again, it is noted that the same concerns may be applicable to many other fields of potential application and not merely limited to TEM analysis, even though the need for high-quality sample support surfaces is clearly very pronounced in this field.
For example, EP 3135631 A1 discloses a method of transferring graphene onto a target substrate having cavities (or holes) or comprising a water soluble layer. A graphene monolayer is grown on a metal foil or a metal thin film on silicon to form a stack. A protective layer is applied onto the stack, e.g. poly(methyl methacrylate) (PMMA), cellulose nitrate (NC), cellulose acetate butyrate (CAB), poly(lactic acid) (PLA), poly(phthalaldehyde) (PPA), poly(bisphenol A carbonate) (PC), or a combination or derivative thereof. A frame is furthermore attached to the protective layer, which comprises a thermal release adhesive polymer layer. This frame provides integrity and allows handling of the sample, while the metal foil or metal thin film (and silicon substrate) are detached or removed. Then, the sample can be dried and deposited onto the target substrate, and the frame can be cut off or thermally released.
EP 2679540 A1 discloses another approach known in the art for, in this case, providing a graphene monolayer on insulating substrates. An adhesive tape is applied to a graphene layer deposited at the bottom (and simultaneously at the top side) of a copper foil in a CVD graphene synthesis process. The adhesive tape is, together with the bottom graphene layer, released from the copper foil by heat at a temperature 1° C. to 5° C. higher than the release temperature of the adhesive tape. A sacrificial protective layer, e.g. PMMA, is then coated onto the top graphene layer that remains on the foil, the foil is etched away, and the graphene is transferred to the insulating substrate, finally removing the sacrificial layer via dissolution or thermal treatment.
Suk et al, “Transfer of CVD-Grown Monolayer Graphene onto Arbitrary Substrates,” ACS NANO 5(9), discloses a wet graphene transfer method using a protective layer of PMMA and ammonium persulfate as etching solution. After removal of the PMMA sacrificial layer by acetone, the transferred graphene layer is furthermore cleaned by annealing at 400° C. in an argon-hydrogen (Ar—H2) atmosphere for an hour.
U.S. Pat. No. 9,388,048 B1 discloses a method for synthesizing a layer of graphene by flowing hydrogen and carbon feedstock over a catalytic layer on a substrate.
EP 3 367 423 A1 discloses a method for obtaining multilayer graphene, in which a first graphene monolayer with a protective layer on top thereof is deposited on a sample comprising a second graphene monolayer grown on a metal foil.
It is an object of embodiments of the present disclosure to provide in good, efficient and/or cheap methods of transferring a graphene layer onto a substrate or support structure, the substrate and/or support structure with graphene thus obtained and a graphene sample support for use in (e.g. transmission) electron microscopy.
It is an aspect of embodiments of the present disclosure that a good, reliable, clean, uniform, and/or homogeneous graphene (mono) layer can be provided on a target substrate or structure.
It is an aspect of embodiments of the present disclosure that a good (e.g. uniform; continuous; large area) coverage and/or high flatness of the graphene layer can be achieved on such target substrate or structure.
It is an aspect of embodiments of the present disclosure that the graphene can be transferred to the target substrate or structure such as to remain intact or substantially intact.
It is an aspect of embodiments of the present disclosure that the graphene can be transferred to the target substrate or structure without or with little contamination, such as contamination by polymers or other organic compounds used in the preparation process.
It is an aspect of embodiments of the present disclosure that such target structure having a clean and high-quality graphene (mono) layer deposited thereon can be particularly useful as a sample support in (e.g. transmission) electron microscopy, e.g. in a TEM grid. For example, the graphene can provide a layer having an aspect of low thickness, high thermal conductivity, high electrical conductivity and/or reproducible (good) uniformity.
It is an aspect of embodiments of the present disclosure that such target structure can be particularly useful in other applications as well, such as graphene sensors and/or MEMS devices.
It is an aspect of embodiments of the present disclosure that a (e.g. TEM) sample support is provided that can reduce wastage of resources, time and/or samples in electron microscopy and similar applications due to failed experiments.
It is an aspect of embodiments of the present disclosure that a background signal obtained when imaging a specimen with TEM or related techniques can be desirably reduced and/or homogenized, thus allowing a simple numerical subtraction.
It is an aspect of embodiments of the present disclosure that artefacts during charged particle (e.g. electron) microscopy to a non-conductive specimen can be reduced or avoided. This may, for example, enable or improve the study of specimens that are extremely beam sensitive, e.g. including (but not limited to) soft-hard interfaces and/or damage-sensitive (in) organic compounds, e.g. capping ligands.
It is an aspect of embodiments of the present disclosure that clean and thin graphene supports are provided that enable or improve the investigation of soft materials, such as surface ligands, or other low-contrast materials by electron microscopy.
A method in accordance with embodiments of the present disclosure achieves the above objective.
In a first aspect, the present disclosure relates to a method of transferring a graphene layer onto a target substrate or support structure. The method comprises obtaining a metal foil onto which the graphene layer is provided, stabilizing the graphene layer by applying a layer of a cellulose-based polymer onto the graphene layer, and placing the metal foil having respectively the graphene layer and the cellulose-based polymer layer stacked thereon in or on an etching solution to dissolve the metal foil supporting the graphene layer. The method further comprises diluting and/or neutralizing the etching solution after the metal foil has been dissolved, and depositing the graphene layer directly onto the target substrate or structure by placing the target substrate or structure underneath the graphene layer floating in or on the diluted and/or neutralized etching solution and removing the diluted and/or neutralized etching solution until the graphene layer settles onto the target substrate or structure. The method comprises dry cleaning the target substrate or structure with the graphene layer deposited thereon to remove the cellulose-based polymer layer by bringing the target substrate or structure with the graphene layer in direct contact with activated carbon, e.g. by embedding the target substrate or structure with the graphene layer in the activated carbon and/or by applying a layer of activated carbon onto the target substrate or structure with the graphene layer (i.e. such that a substantial portion of the cellulose-based polymer layer is in contact with the activated carbon), and heating the activated carbon.
Furthermore, the activated carbon and the target substrate or structure with the graphene layer (being in direct contact with the activated carbon) are heated to a temperature at least 5° C. higher than the melting temperature of the cellulose-based polymer and said temperature is maintained until the cellulose-based polymer is substantially removed from the graphene layer by the activated carbon.
In a method in accordance with embodiments of the present disclosure, bringing the target substrate or structure with the graphene layer thereon into direct contact with the activated carbon may comprise or consist of embedding the target substate or structure with the graphene layer in the activated carbon.
In a method in accordance with embodiments of the present disclosure, bringing the target substrate or structure with the graphene layer thereon into direct contact may comprise or consist of at least covering the cellulose-based polymer, e.g. at least 50% of the area of the cellulose-based polymer layer, e.g. at least 75%, e.g. at least 90%, e.g. at least 95%, e.g. substantially entirely, with a layer, e.g. film, of activated carbon.
In a method in accordance with embodiments of the present disclosure, obtaining the metal foil onto which the graphene layer is provided may comprise growing the graphene layer onto the metal foil by chemical vapor deposition.
A method in accordance with embodiments of the present disclosure may comprise mechanically flattening the metal foil with the graphene layer provided thereon before applying the cellulose-based polymer layer.
In a method in accordance with embodiments of the present disclosure, applying the cellulose-based polymer layer may comprise coating the metal foil having the graphene layer attached thereto with a solution of the cellulose-based polymer using a dip-coating and/or spin-coating method.
In a method in accordance with embodiments of the present disclosure, the cellulose-based polymer layer may consist of cellulose acetate butyrate (CAB).
In a method in accordance with embodiments of the present disclosure, the cellulose-based polymer layer may have a thickness in the range of 10 nm to 500 μm, e.g. in the range of 15 nm to 30 nm, e.g. a thickness of 20 nm.
In a method in accordance with embodiments of the present disclosure, applying the cellulose-based polymer layer may comprise removing the cellulose-based polymer that was dip-coated directly onto the metal foil on the side opposite of the side where the graphene layer is provided such that the metal foil is exposed and the graphene layer remains covered by the cellulose-based polymer.
In a method in accordance with embodiments of the present disclosure, placing the metal foil having respectively the graphene layer and the cellulose-based polymer layer stacked thereon in or on the etching solution may comprise placing the metal foil in or on the etching solution with the metal foil directed downward, such that the graphene layer can settle onto the target substrate or structure without inversion (in the step of removing the diluted and/or neutralized solution).
In a method in accordance with embodiments of the present disclosure, the etching solution may comprise ammonium persulfate or may consist of a solution of ammonium persulfate.
In a method in accordance with embodiments of the present disclosure, the etching solution may be a solution of ammonium persulfate in ultrapure water at a concentration in the range of 0.3 g to 5 g per 100 mL solvent, e.g. in the range of 3 g to 4 g per 100 mL solvent.
In a method in accordance with embodiments of the present disclosure, the step of diluting and/or neutralizing may comprise (or consist of) exchanging the etching solution with water, or ultrapure water (UPW).
In a method in accordance with embodiments of the present disclosure, the target substrate or support structure may comprise an (transmission) electron microscopy sample support grid.
A method in accordance with embodiments of the present disclosure may comprise drying the target substrate or structure with the graphene layer deposited thereon before the step of dry cleaning.
In a method in accordance with embodiments of the present disclosure, the activated carbon and the embedded target substrate or structure with the graphene layer may be heated to a temperature at least 5° C. higher than the melting temperature of the cellulose-based polymer and this temperature may be maintained for at least 30 minutes, e.g. at least 1 hour, e.g. at least 2 hours, e.g. at least 4 hours. However, as noted further hereinbelow, even a time of only about a minute may suffice in a low-pressure environment, e.g. a vacuum environment.
In a method in accordance with embodiments of the present disclosure, the activated carbon and the embedded target substrate or structure with the graphene layer may be heated to a temperature in the range of 250° C. to 300° C. and this temperature may be maintained for at least 8 hours, e.g. at least 10 hours, e.g. at least 12 hours, e.g. 15 hours or longer. It is noted, however, that the treatment time may also be less than 8 hours, e.g. at least 1 minute, or at least 5 minutes, or at least 10 minutes, or at least 20 minutes, or at least 30 minutes, or at least 1 hour, or at least 2 hours, or at least 4 hours; e.g. as already indicated hereinabove. For example, the (minimum) required treatment time may be substantially lower when applying the heating in a vacuum (or low-pressure) environment (without limitation thereto).
In a method in accordance with embodiments of the present disclosure, the step of dry cleaning may comprise (rapidly) cooling down the activated carbon, having the target substrate or structure with the graphene layer still embedded therein, after the heating of the activated carbon, using liquid nitrogen.
In a method in accordance with embodiments of the present disclosure, the heating step of the dry cleaning may be performed under low ambient pressure and/or in vacuum. It is an aspect of this approach that the heating can be performed under a desirably low temperature (e.g. slightly above the melting point of the polymer at the pressure of the low-pressure or vacuum environment, e.g. in the range of 210° C. to 220° C., e.g. 215° C. to 240° C., or potentially even 225° C. to 300° C.) and requires little time, e.g. at least one minute, or at least 2 minutes, e.g. at least 5 minutes, e.g. a time in the range of 1 minute to 10 minutes, or in the range of 2 minutes to 30 minutes. A relatively low temperature (e.g. compared to conventional dry cleaning without the aid of activated carbon) and/or a short exposure time can desirably reduce the risk of damage to the delicate graphene layer.
A method in accordance with embodiments of the present disclosure may comprise a further wet cleaning step, before the dry cleaning step (and after the step of dissolving the metal foil, e.g. after removing the diluted/neutralized etching solution and optionally drying the intermediate product thus obtained), to remove (or reduce) metal nanoparticles (e.g. copper nanoparticles, e.g. remnants of the metal foil that remain on the graphene layer after the foil is dissolved and the etching solution is removed).
In a method in accordance with embodiments of the present disclosure, the graphene layer may be made hydrophilic by inducing a static charge thereon, e.g. in a step of depositing (e.g. drop-casting) a sample onto the graphene layer after said graphene layer has been deposited onto the target substrate or support structure, e.g. inducing a static charge on the graphene layer while drop-casting the sample onto a prepared graphene grid (as obtained by the method steps discussed hereinabove).
In a second aspect, the present disclosure relates to an electron microscopy sample support grid, comprising a substrate or support structure, and a graphene layer, forming a sample support, positioned on the substrate or support structure, in which the graphene layer has a thickness of less than 2.0 nm. For example, the electron microscopy sample support grid may be obtained and/or obtainable by a method in accordance with embodiments of the first aspect of the present disclosure. The sample support grid is furthermore characterized by a ratio of surface area of clean graphene over the total area in a bright-field transmission electron microscopy image of the sample support grid at a magnification sufficiently high to exclude the substrate or support structure from the image, in which said ratio is at least 60%, e.g. at least 65%, when the pixels of the image thus obtained are segmented into (exactly) two components corresponding to a clean homogeneous graphene layer component and a remainder component (containing any type of inhomogeneities, impurities or other factors that would differ from a homogeneous, and substantially constant thickness, graphene layer background) by a pixel-wise segmentation algorithm.
The independent and dependent claims describe specific and features of the disclosure. Features of the dependent claims can be combined with features of the independent claims and with features of other dependent claims as deemed appropriate, and not necessarily only as explicitly stated in the claims.
The drawings are schematic and not limiting. Elements in the drawings are not necessarily represented on scale. The present disclosure is not necessarily limited to the specific embodiments of the present disclosure as shown in the drawings.
Notwithstanding the exemplary embodiments described hereinbelow, is the present disclosure only limited by the attached claims. The attached claims are hereby explicitly incorporated in this detailed description, in which each claim, and each combination of claims as allowed for by the dependency structure defined by the claims, forms a separate embodiment of the present disclosure.
The word “comprise,” as used in the claims, is not limited to the features, elements or steps as described thereafter, and does not exclude additional features, elements or steps. This therefore specifies the presence of the mentioned features without excluding a further presence or addition of one or more features.
In this detailed description, various specific details are presented. Embodiments of the present disclosure can be carried out without these specific details. Furthermore, well-known features, elements and/or steps are not necessarily described in detail for the sake of clarity and conciseness of the present disclosure.
The present disclosure relates to a method of transferring a graphene layer, e.g. a graphene monolayer, onto a target substrate or support structure, such as a support frame or electron microscopy sample grid, e.g. for TEM applications. For example, a single layer of ultra-clean graphene may be provided on a TEM support, which may show improved coverage ratio and graphene cleanness after transfer onto the support when compared to prior-art approaches. Furthermore, a low graphene roughness can be achieved and a good flatness, e.g. avoiding or reducing wrinkles of the layer. It will be understood by the person skilled in the art that this method and its potential aspects are not limited to TEM supports, e.g. may find application in many other fields where a high-quality graphene layer on a substrate or support structure is of use.
Referring to
The method comprises obtaining 2 a metal foil 12 onto which a graphene (mono) layer 11 is provided. The graphene layer 11 may be a monolayer, e.g. having large single-crystalline domains. The graphene layer 11 may be grown by chemical vapor deposition (CVD) onto the metal foil 12, e.g. as known in the art. For example, the metal foil with graphene layer may be a monolayer CVD graphene on copper (Cu), which is for example commercially available from Graphenea Inc., Cambridge (USA). The same or similar product may be available from other high-quality graphene providers for research and/or high-tech applications. For example, such graphene foils may be available in a size of about 1 cm2, e.g. 10 mm by 10 mm. Other formats may be commercially available as well, e.g. 5 cm by 5 cm of even larger sizes, e.g. 15 cm by 15 cm. However, embodiments of the present disclosure are obviously not limited to specific choices of supplier, graphene carrier (e.g. metal foil), foil size, and/or other such straightforward variations.
Obtaining the metal foil with the graphene (mono) layer thereon may comprise forming a graphene layer (e.g. monolayer) on a metal foil, e.g. using chemical vapor deposition. Many variations are known in the art to synthesize a graphene layer onto a metal substrate, e.g. using CVD. For example, methane gas can be used as carbon source for such deposition process, and alternatives may include e.g. petroleum asphalt. Hydrogen may be used in the process to promote carbon deposition (e.g. using a combination of methane and hydrogen at a suitably tuned flow ratio). The hydrogen may desirably corrode amorphous carbon and improve the quality of the deposited graphene, but should be used in moderation to avoid corrosion of the graphene and/or damage to the formed crystal lattice. The carrier gas may comprise an inert gas, e.g. argon, for example in combination with hydrogen. Optionally, a catalyst may be used, such as iron nanoparticles, nickel foam and/or gallium vapor. Not only the gas flow ratio is typically tuned to optimize results, but also the ambient pressure, temperature and chamber material may need to be considered to achieve high-quality layers. Quartz may be a suitable material for the chamber and auxiliary tubing. These deposition processes are well-known in the art, and the skilled person is capable of determining such parameters with straightforward experimentation and the knowledge readily available in the field.
The method 1 may comprise mechanically flattening 3 the metal foil (supporting the graphene layer), e.g. by compressing the foil (with graphene layer) between two flat surfaces 21. For example (without limitation thereto), the metal foil may be placed between two clean microscope slides, e.g. previously cleaned with ethanol and acetone to avoid contamination of the graphene layer 11. While the step of mechanically flattening 3 is not strictly necessary, it may substantially yield a lower surface roughness of the graphene layer in the end result.
The method comprises stabilizing the graphene layer 11 by applying 4 a layer 13 of a cellulose-based polymer, e.g. cellulose acetate butyrate (CAB), onto the graphene layer 11. Other examples of cellulose-based polymers may include ethyl cellulose (EC), cellulose acetate (CA), cellulose acetate propionate (CAPr), amongst others, such as comprising formvar (polyvinyl formal), butvar (polyvinyl butyral resin) and/or glycerol. For example, the graphene layer may be coated by a CAB layer, or in another illustrative example by a thin layer of a glucose solution. e.g. a solution of about 1.8 g per 10 ml water (without limitation thereto).
The layer 13 may be a thin layer, e.g. having a thickness in the range of 10 nm to 500 μm, in the range of 15 nm to 1 μm, in the range of 15 nm to 100 nm, or about 20 nm. For most purposes, a layer of 20 nm may be considered sufficient, even though higher values are not necessarily detrimental. However, the final step of cleaning away the cellulose-based polymer may be easier, faster and/or less prone to leaving contamination behind when a layer thickness in the lower ranges mentioned hereinabove is selected.
The step of applying the cellulose-based polymer layer onto the graphene layer may comprise coating the metal foil having the graphene layer attached thereto with a solution of the cellulose-based polymer, e.g. using a dip-coating and/or spin-coating method.
For example, the metal foil with graphene may be dip-coated with a coating solution, e.g. a solution of 0.25 g CAB (e.g. having an average molecular weight Mn of about 30000; e.g. as commercially available from Merck KGgA; CAS no. 9004-36-8) in 100 mL ethyl acetate or another suitable solvent (e.g. a solution of 2.5 g/L CAB in ethyl acetate). Other suitable cellulose-based polymer solutions, e.g. in a suitable volatile solvent, may be considered as well. A homogenous solution can be achieved by stirring or another type of agitation, e.g. using a magnetic stirrer. The solution of CAB may have a concentration in the range of 1 g/L to 10 g/L, e.g. in the range of 1.5 g/L to 5 g/L, e.g. in the range of 2 g/L to 3 g/L, e.g. in the range of 2.3 g/L to 2.7 g/L of CAB in ethyl acetate or another suitable solvent. The CAB may have a molecular weight (Mn) in the range of 12*10∧3 to 70*10∧3, e.g. 12*10∧3, e.g. 30*10∧3, e.g. 70*10∧3.
Referring to
The layer of the cellulose-based polymer may be applied such as to reach a layer thickness (of the cellulose-based polymer) in the range of 10 nm to 100 nm, in the range of 15 nm to 30 nm, or about 20 nm. It is an aspect of a dip-coating and/or spin-coating process, that the thickness of the layer can be easily and reproducibly controlled in this range, e.g. to achieve a layer thickness of about 20 nm. Applying the cellulose-based polymer layer may furthermore comprise removing the cellulose-based polymer 22, e.g. CAB, from the metal foil 12 on the side that is opposite of the side covered by the graphene layer 11. For example, dip-coating may produce a stack of CAB-graphene-Cu-CAB, wherein the CAB layer 22 directly on the metal foil is removed to obtain substantially a stack of CAB-graphene-Cu, e.g. such that the metal foil (Cu in this example) can be effectively and efficiently etched away. If, alternatively, a spin-coating process is used, it will be understood that a stack of CAB-graphene-Cu (or, generally, with any type of metal foil in the final layer) can be obtained directly, i.e. without requiring the removal of a superfluous CAB layer 22 on the opposite side.
The latter CAB layer 22 (or another cellulose-based polymer) may, for example, be removed from the backside of the metal foil by a suitable solvent, such as acetone, ethanol or isopropanol. For example, the CAB may be removed using a drop 24 of acetone, which is applied carefully to the side to be cleaned to avoid spreading to the topside of the CAB-graphene-Cu stack. A drop of acetone may be deposited on a clean surface, e.g. a glass slide, and the backside of the stack may be brought carefully into contact with the acetone with due care to avoid dispersion to the other side.
Thus, the graphene layer to be stabilized and protected is covered by a cellulose-based polymer layer, while the other side of the metal foil is substantially cleaned of this polymer, such that it can be exposed to an etchant.
However, embodiments are not limited to this illustrative process for applying the polymer, and it will be understood that another suitable method, as known in the art, for applying a cellulose-based polymer onto the graphene layer may be used (alternatively or additionally), e.g. a spin-coating technique.
The method further comprises placing 5 the metal foil 12, having respectively the graphene layer 11 (directly) and the cellulose-based polymer layer 13 (indirectly) provided on top thereof, in or on an etching solution 25 to dissolve the metal 12 supporting the graphene layer 11. For example, the metal foil 12 may be placed downside onto the surface of the etching solution to further avoid or minimize damage to the sensitive graphene layer, e.g. allowing the foil to float in or on the etching solution.
The etching solution 25 may comprise ammonium persulfate ((NH4)2S2O8) or may consist of a solution of ammonium persulfate, e.g. in (substantially pure) water. It is an aspect that, thanks to the stabilization by the polymer, a relatively high concentration, e.g. 3.4%, of ammonium persulfate can be used, which thus contributes to a speedy process. This also has the aspect that the graphene is protected from damage and/or wrinkling.
For example, the dissolution of the metal foil can be carried out in less than 4 hours, e.g. less than 2 hours, e.g. about an hour or even less. Even though the etching time may, generally, depend on the concentration of the etching solution, the stabilizing effect of the cellulose-based polymer enables the use of a relatively high concentration, and thus a rapid etching.
For example, the etching solution 25, e.g. an ammonium persulfate solution, may be heated to a temperature in the range of 30° to 50° C., e.g. in the range of 35° C. to 45° C., e.g. about 40° C., e.g. to further increase the efficiency of the chemical etching process. For example, the solution may be prepared on a hotplate (e.g. at 40° C.) while, in embodiments, being continuously stirred, e.g. using a magnetic stirrer (e.g. at 650 rpm). However, it is to be noted that heating the etching solution is not necessarily required, e.g. the same etching can be achieved at room temperature, even though a faster etching can be achieved at elevated temperature.
The etching solution may be a solution of ammonium persulfate of 3.42 g in 100 ml (e.g. 34.2 g/L) in a suitable solvent, water, or ultrapure water (UPW), e.g. according to the standards defined in ASTM D5127 and/or SEMI F63. The concentration of the ammonium persulfate may be in the range of 0.3 g to 5 g per 100 mL solvent, e.g. in the range of 0.3 g to 4 g per 100 mL solvent, e.g. in the range of 2 g to 4 g per 100 mL solvent, e.g. in the range of 3.0 g to 3.5 g per 100 mL solvent, e.g. in the range of 3.3 g to 3.4 g per 100 mL solvent. While excessive etching concentrations could potentially damage the graphene and/or introduce undesirable residues, it is noted that higher concentrations may fasten the etching process considerably. However, if the exposure time is adjusted accordingly, a copper foil can be etched away with concentrations of etchant ((NH4)2S2O8) as low as, e.g., 0.34 g/100 mL UPW.
In order to prevent contamination by the dissolved metal of the metal foil, the method comprises diluting and/or neutralizing 6 the etching solution 25 before removing the graphene layer, which is now only attached to (and stabilized by) the remaining cellulose-based polymer layer. For example, this dilution may comprise a gradual dilution (and/or neutralization) using water, e.g. ultrapure water, to remove the metal, e.g. copper, residues. For example, clean solvent, e.g. ultrapure water, may be added repeatedly, e.g. while siphoning off excess solution (e.g. using a pipette).
The etching solvent may thus be removed by exchanging it, e.g. in steps, with water (or UPW), for example until the pH has reached a value sufficiently close to pure water, e.g. in the range of 5.7 to 7.1, e.g. in the range of 5.8 to 6.2, e.g. in the range of 5.8 to 5.9, e.g. about 5.85 (e.g. as measured for UPW for a resistivity of 18.2 MΩcm). For example, the etching solvent may be exchanged with water repeatedly until the pH value is increased from an acid environment (e.g. about 3) to neutral (e.g. about 7).
This step avoids damage to the target substrate or structure by the remaining etchant, e.g, particularly in case the target substrate or structure is sensitive to oxidation. Moreover, even if the target substrate or structure is not oxidation-sensitive, contamination by residues of the dissolved metal is reduced or prevented.
While the etching solvent may be alternatively exchanged with another solvent, such as isopropanol and/or acetone, experiments have shown that these alternatives could potentially lead to a higher contamination and/or could pose problems, e.g. of the graphene tending to attach more easily to the wall of the recipient. Nonetheless, embodiments using another solvent than water for dilution and/or exchange of the etching solution are not necessarily excluded.
The method further comprises depositing 7 the graphene layer 11 (directly) onto a target substrate or structure 20. It is an aspect of a method in accordance with embodiments that the graphene layer, floating in or on the diluted etching solution, can be easily scooped onto the target substrate or structure.
The method may thus also comprise providing a target substate or structure onto which the graphene layer is to be transferred.
Referring to
The target substrate or structure may, for example, be placed 27 on top of a piece of filter paper 26 or other suitable protective (e.g. shock absorbing) material at the bottom of the recipient containing the solution 25, and the graphene layer 11 can thus be softly lowered on top of the substrate or structure by gently removing the solution without risk (or with reduced risk) of damage. It will also be appreciated that such filter paper 26 or similar carrier material can be used to easily manipulate the substrate or structure, e.g. to remove the construct from the recipient once the graphene layer has been deposited thereon.
For example, the target substrate or structure 20 may comprise one or more transmission electron microscopy sample grids. As an example, for a 10 mm by 10 mm graphene layer, 9 TEM grids, having a diameter of 3.05 mm, may be arranged in a grid pattern of 3 rows by 3 columns on the filter paper inside an area of 10 mm by 10 mm, such that the graphene layer can be deposited on top of this pattern of TEM grids.
For example, the target substrate or structure 20 may comprise a bare 2000 M gold TEM grid. Other grid materials are not necessarily excluded, e.g. copper, nickel, molybdenum, platinum, rhodium, silicon nitride etc. (or alloys or mixtures of multiple materials). Other mesh sizes, e.g. 1000, 1500, 3000, etc., are not necessarily excluded, and the pitch of the grid may equally vary. The TEM grid size may be a standard size of 3.05 mm diameter, even though embodiments are not necessarily limited thereto (e.g. another example being 2.3 mm). Such TEM grid may comprise a holey CAB layer onto which the graphene layer, after transfer, may be supported. The TEM grid may also comprise an amorphous carbon layer.
Even though reference is made to the ‘solution’ for the sake of clarity, it will be understood that the ‘solution’ at this stage in the process may be substantially nothing more than water, e.g. if the aforementioned step of removing 6 (by diluting and/or neutralizing) the etching solvent was carried out effectively.
It will also be understood that the position of the graphene layer, while lowering it on top of the substrate or structure can be controlled during this step, e.g. by gentle nudges and/or by the controlling the (direction of the) flow by which the solution is being removed, such that a good alignment of the graphene layer on the target substrate or structure can be easily obtained.
Experiments have demonstrated that gently lowering the graphene layer onto the target, by removing the liquid medium, improves the obtained coverage and roughness of the final construct. For example, when compared to a similar approach in which the graphene layer is scooped from above onto the target, a higher coverage and lower roughness are obtained by letting the graphene layer settle on top of the target.
The steps described hereinabove can be automated in a straightforward manner, such that even better precision in alignment and reproducibility of the manufactured item (target substrate or structure having the graphene layer provided thereon) can be achieved and/or such that a high processing speed can be achieved. It will also be clear that this process can be easily scaled up to manufacture many of such items simultaneously. Moreover, the limiting step in the efficiency of the process described hereinabove (disregarding any prior steps for manufacturing the target substrate/structure and/or creating the graphene layer, e.g. by CVD) appears to be the etching time, which, as already mentioned can be performed relatively quickly, e.g. at least compared to some prior art approaches that rely on such chemical etching step.
The method may further comprise drying 9 the target substrate or structure 20 with the graphene layer 11 deposited thereon. For example, by placing the construct on a hotplate, e.g. at a temperature in the range of 25° C. to 50° C., e.g. in the range of 30° C. to 40° C., e.g. about 35° C. It is to be noted that, when placed on a filter paper or other suitable carrier material, it is not only easy to remove the construct from the recipient that contained the diluted etching solution (e.g. substantially water) for drying, but that the construct may also be protected from direct heat in the drying process. It is also noted that the drying may also be carried out in situ, e.g. without removing from the recipient, to avoid any damage due to manipulation.
Optionally, the method may also comprise a wet cleaning step, using a (further) etching solution, e.g. before the dry cleaning step discussed further hereinbelow (e.g. after drying the target substrate or structure with the graphene layer deposited thereon, e.g. for 35 minutes at 35° C.). For example, a drying step as generally discussed hereinabove may be applied to the target substrate or structure with the graphene layer deposited thereon and then placed in the further etching solution, e.g. ammonium peroxodisulphate. For example, the intermediate product may be placed at the bottom of a petri dish with a solution of 3 g ammonium peroxodisulphate per 100 ml water. An agitator may be used to ensure good exposure of the intermediate product (and potential impurities thereon) to the etching solution, e.g. by stirring using a stirring bar at 100 rpm for 2 hours. Then, the solution may be gently neutralized, e.g. to the pH (5.85) of ultrapure water (18.2 MΩcm). The (neutralized) solution may then be removed (or, obviously, the intermediate product may be removed from the vessel containing the neutralized solution), and the intermediate product may be (optionally) washed and/or rinsed, e.g. using ultrapure water (e.g. in a fresh bath, to remove nanoparticles or other contaminations carried over, from the solution, on the intermediate product).
After this optional wet cleaning to remove impurities, such as copper nanoparticles, the structure may be dried again, e.g. in accordance with the drying 9 step already discussed hereinabove or simply at room temperature. It is noted that this optional wet cleaning step may be adapted to remove impurities, such as metal nanoparticles, without being too harsh, e.g. to avoid damage to the graphene structure, since the sacrificial cellulose-based polymer is substantially removed by the subsequent dry cleaning step as discussed further hereinbelow.
The method further comprises a dry cleaning 10 (a dry heating treatment) to remove the cellulose-based polymer layer 13. In this step, the target substrate or structure 20 with the graphene layer 11 and the remaining cellulose-based polymer 13 are brought in direct contact with activated carbon 31, e.g. embedded in activated carbon. For example, the activated carbon may be provided in a recipient (e.g. a petri dish) or pile, a pit may be formed in the activated carbon, and the construct may be buried in this pit, substantially covering the construct from all sides by the activated carbon (i.e. closing the pit). For example, a heap of 0.7 cm activated carbon may be provided, and the construct may be buried at a depth of about 0.5 cm in this heap. However, it will be understood that these values are only illustrative, e.g. such heap size and embedding depth can vary. For example, the construct (i.e. the intermediate product) may be buried at a depth in the range of 0.25 to 10 cm (without necessarily limiting to this range), e.g. in the range of 0.5 to 5 cm, e.g. at a depth of 0.5, 1, 2 or 3 cm.
Alternatively (or additionally), a layer of activated carbon may be applied to the target substrate or structure 20 with the graphene layer 11 and the remaining cellulose-based polymer 13. For example, a film of activated carbon may be applied to the intermediate product such that the activated carbon is in direct contact with the intermediate product, e.g, particularly in direct contact with the exposed cellulose-based polymer and/or any contaminants in/on the graphene. However, care should be taken to avoid activated carbon films that comprise binding agents or other products that are unsuitable, e.g; have a low melting point. For example, it may be desirable to avoid activated carbon films that rely on Teflon to support and/or bind the film structure. However, procedures of preparing a suitable (e.g. substantially pure) activated carbon film are known in the art, e.g. the example disclosed in Liu et al, “Self-assembled activated carbon nanoparticles for reliable time-discretized quantitative surface-enhanced Raman spectroscopy,” Analytical Methods 9 (47), pp. 6622-6628, and/or in Esteban et al, “Tunable supercapacitor materials derived from hydrochar/gold nanograpes,” ACS Applied Energy Materials 3 (9), pp. 9348-9359.
The activated carbon 31 in which the construct is embedded and/or having the construct in direct contact with the activated carbon may then be heated, e.g. to a temperature in the range of 210° C. to 300° C., e.g. in the range of 250° C. to 300° C., for example at a rate of 5° C. per minute (embodiments not necessarily limited to this illustrative temperature range and/or heating rate). However, a gradual increase of the temperature may not be strictly necessary. For example, the construct in/with activated carbon may be heated essentially instantaneous to 210° C. and the temperature may be gradually increased up to 250° C. at a rate of (e.g.) 5° C. per minute.
Specifically, the target temperature may be chosen to be higher than the melting temperature of the cellulose-based polymer, but conservatively selected to avoid damage to the graphene and/or the target substrate or structure. For example, the target temperature may be at least 5° C. or at least 10° C., above (e.g. a lower or upper limit of) the melting temperature. This target temperature is maintained for a sufficient time, e.g. at least 4 hours, e.g. at least 8 hours, e.g. for a time in the range of 10 hours to 24 hours, e.g. 15 hours. After cooling down (and removing 32 of the construct from the carbon, or vice versa), the target substrate or structure 20 remains with the graphene layer, while the polymer is cleaned away. However, experiments have confirmed that good results can be obtained in even shorter times, e.g. at least one hour of heat treatment. Generally, a time of at least 30 minutes may be considered as sufficient, in which tests can be used to adjust the time to determine a suitable trade-off between quality (e.g. cleanness) and use of resources (e.g. processing time). As discussed further hereinbelow, use of a vacuum may even further reduce the required time.
By heating in activated carbon, e.g. for 15 h at 300° C., the cellulose-based polymer, e.g. CAB, and potentially other contaminants can be efficiently removed. The melting temperature of CAB is between 170° C. and 240° C., such that a temperature slightly above the melting temperature of the polymer can liquefy the material and allow it to be absorbed by the carbon. It is to be noted that other cellulose-based polymers may equally be suitable for a method in accordance with embodiments, and the temperature of the thermal treatment can be adjusted accordingly.
Without the aid of the activated carbon, substantially higher temperatures may be required for sufficiently cleaning the polymer away. For example, a conventional vacuum thermal treatment may need a temperature in the range of 700° C. to 900° C., (i.e. when the treatment is not aided by activated carbon, in accordance with embodiments of the present disclosure) to achieve an acceptable level of cleanness (e.g. at least for application of the product as TEM support), which can lead to the destruction of the sensitive layer, e.g. resulting in a lower coverage, as well as corrugation.
Optionally, this heating step may be performed under low ambient pressure and/or in vacuum. This may be useful to reduce the time needed for dry cleaning the construct (I.e. the intermediate product) and/or to allow an efficient cleaning at a lower temperature (which may reduce the risk of damage to the graphene layer). For example, in (substantially) vacuum, an efficient dry cleaning can be achieved at a minimum temperature of 215° C. (e.g. for removal of a CAB polymer layer), as confirmed by in-situ experiments with a heating holder, with a substantially complete removal of the polymer in a few minutes (e.g. at least one minute, or at least 2 minutes, e.g. at least 5 minutes). It is noted that it is an aspect of this approach, that the time required for such vacuum treatment is substantially lower than would be required for a conventional vacuum (or in air) cleaning step (i.e. without the use of activated carbon), such that damage to the graphene (e.g. due to air exposure) can be reduced or even entirely avoided.
It will be understood that the (absolute values of) temperatures mentioned hereinabove may be different when another polymer material is used, e.g. the values given may be generally suitable for a CAB layer but might need adjustment (or, as the case may be, may not) for other materials, e.g. a dextrose film, aniline, formvar, glycerine, polyvinylpyrrolidone, cellulose triacetate and/or other polymers that may be contemplated.
The dry cleaning step 10 may also comprise cooling down the target substrate or structure 20 with the graphene layer 11 (e.g. while still embedded in the activated carbon 31) after the heat treatment. By quickly cooling down the activated carbon pile, degassing of residues that were absorbed by the activated carbon during the cleaning process can be avoided or reduced. Thus, the cooling may be performed at a high rate of decreasing temperature, such as at least 10° C./s (i.e. the temperature may change with a rate≤−10° C./s). For example, the activated carbon 31 with the embedded target substrate or structure 20 (and the graphene layer 11 thereon) may be exposed to (e.g. placed in; not necessarily directly exposed to) liquid nitrogen after the aforementioned heating. Thus, the temperature (e.g. 300° C.) of the activated carbon pile (and the embedded construct) after the heat treatment may be decreased rapidly. Direct contact between the liquid nitrogen and the activated carbon (and/or the embedded construct) may be avoided. For example, the activated carbon pile may be contained in a (e.g. shallow) recipient, such as a petri dish, and this recipient may be placed in a larger container containing the liquid nitrogen, wherein care is taken that the volume of the liquid nitrogen in the larger container is below a predetermined threshold to avoid overflowing of the liquid nitrogen from the larger container into the recipient. In view of the large temperature difference between the heated activated carbon and the liquid nitrogen, a relatively short cooling time may suffice, e.g. in the range of 5 to 20 seconds or in the range of 10 to 15 seconds. The duration of this cooling step may be determined such as to be long enough to sufficiently cool down the activated carbon (to avoid said degassing), but also to be not too long, to prevent the (potentially damaging) formation of ice in the construct.
After transferring the graphene onto a TEM support (for example), a low graphene roughness can be achieved, e.g. down to 0.2 nm (potentially even less).
Furthermore, in embodiments where the target structure is a TEM grid, e.g. an Au or Cu grid, this target structure may already comprise a holey CAB layer (as known in the art). However, this holey CAB layer on the primary grids can be substantially protected during the treatment by the amorphous carbon layer of such TEM grids. During the heating treatment, these initial layers are not being removed and act (as intended) as a further support for the graphene.
Experiments have shown that this approach results in a clean graphene layer, for which a signal to background ratio in TEM measurements can be achieved that can be, for example, six times higher than observed for comparable TEM graphene grids manufactured by a method as known in the art. Furthermore, substantially less contamination is observed during TEM analysis compared to commercially available TEM graphene grids.
Optionally, the graphene layer and/or the target substrate or structure 20 may be made hydrophilic, e.g. by inducing a static charge thereon, for example in a step of depositing (e.g. drop-casting) a sample (e.g. dispersion) onto the graphene layer (e.g. onto a TEM graphene grid as obtainable by the steps discussed hereinabove). For example, a plastic material (e.g. a rod) may be rubbed with a suitable cloth, e.g. wool, to induce a negative charge on the plastic material. By bringing this plastic material in close proximity (e.g. about 1 cm or less) to the target (e.g. to the TEM graphene grid or other target substrate/structure with graphene thereon as obtained by the steps discussed hereinabove), a static charge can be induced on the graphene layer. For example, the plastic material may be held as close as possible (e.g. within 1 cm; e.g. while still avoiding direct contact) above or below the TEM graphene grid (or other target substrate/structure with graphene layer) while drop-casting the sample onto the graphene layer, e.g. underneath the plastic material. The plastic material (or other static charge inducer) may be kept close while removing (e.g. evaporating and/or actively removing, e.g. by holding an edge of the grid to a filter paper) the solvent, e.g. such that the induced hydrophilic property is maintained.
In a second aspect, the present disclosure relates to an electron microscopy sample support grid, comprising a substrate or support structure, and a graphene layer positioned on this substrate or support structure. The graphene layer thus forms a sample support onto which a sample can be placed for analysis. Furthermore, the graphene layer has a thickness of less than 2.0 nm, e.g. when measured by EELS (cf. measurement methodology discussed further hereinbelow).
Such sample support grids have already been discussed hereinabove, and are furthermore known as such in the art. Features of such sample support grid, e.g. as discussed hereinabove or otherwise well-known in the field, are considered to correspond to (at least optional) features of an electron microscopy sample support grid in accordance with embodiments of the present disclosure. For example, the sample support grid may comprise metal wires and/or metal layer(s) to form a support structure (primary grid), by which the graphene layer is supported. Examples of such metal materials include copper and gold, among others. The device may also comprise a holey CAB layer, or suitable alternative thereof. For example, a (e.g. CAB) foil or layer, supported by the primary grid, may be perforated to form a collection of holes with certain size, shape and/or arrangement. For example, these holes may range between 0.5 μm and 5 μm. It is an aspect of such holey layer that openings (holes) can be achieved having a diameter in the micrometer or even sub-micrometer range, whereas, for example, holes formed by crossing metal grid lines may typically have larger dimensions, e.g. in the 5 μm to 200 μm range.
The support substrate or structure may thus comprise holes (through-holes) which are covered by the graphene layer (or a plurality of such graphene layers, e.g. different segments of graphene may cover different holes). These holes may for example be formed by grid openings in a grid pattern of the support structure, e.g. openings between crossing grid lines of a primary grid, or by holes in a holey (e.g. CAB) layer, or a combination thereof.
Further optional features are well-known to the skilled person, and therefore not discussed in detail. For example, the grid may comprise a frame structure (e.g. formed from a substrate, such as a semiconductor substrate or a metal around the rim of the grid to allow a reliable, stable and sufficiently strong mechanical mounting of the grid in an electron microscopy apparatus.
By applying a method of manufacture in accordance with embodiments of the first aspect of the present disclosure, as discussed hereinabove, a quality of the graphene sample support can be achieved that is substantially better than what can be achieved using prior-art methods, at least in so far as is known and could be ascertained by the inventors.
Among other quality metrics, such as improved cleanness, uniformity of thickness, flatness, etc., a consistent and consistently small thickness of the graphene layer, after transfer onto the supporting grid structure or substrate, of less than 2.0 nm can be achieved, whereas prior art methods failed to provide such thin graphene layers in experimental tests. Even though, theoretically, a monoatomic layer of graphene has a definite thickness (e.g. 0.34 nm if defined in terms of the interlayer distance between graphene sheets in a graphite structure), it will be understood that the actually observed (i.e. observable; measurable, empirical) thickness can vary due to various quality-related factors. To give a few examples, the observed thickness can increase due to pollutants, or the observed thickness can increase due to wrinkling, folding or other deviations from a (ideally) flat surface, or extra graphene layers.
The electron microscopy sample support grid in accordance with embodiments of the present disclosure can furthermore be characterized by a ratio of clean graphene over the total area in a bright-field transmission electron microscopy image of the sample support grid at a magnification of 7100×, wherein said ratio is greater than 60%, e.g. greater than 65%. Pixels of the image can be classified as either clean graphene or not, e.g. containing residues of polymer and/or hydrocarbon contaminants, by a Random Forest Classifier algorithm. For the sake of robustness, this measurement can be averaged over a number of different sampling locations on the sample support grid, e.g. a central point and four points around the center along two in-plane perpendicular axes, e.g. x and y directions of in-plane translation.
The electron microscopy sample support grid in accordance with embodiments of the present disclosure can furthermore be characterized by a ratio of grid holes covered by graphene to total number of grid holes, in which said ratio is greater than 85%. This ratio can be determined by acquiring a scanning electron microscopy image of the sample support grid, and segmenting the image into three components, corresponding to a grid hole covered by graphene, a grid hole uncovered or insufficiently covered by graphene (“empty” hole), and the support grid structure itself. For this purpose, a Three-Component Random Forest Classifier algorithm can be used. The ratio is thus determined by the number of pixels in the graphene-covered image segment over the total number of pixels.
In order to determine the thickness of the graphene layer in an unambiguous and definite way, the following protocol can be used. In this protocol, the local thickness is calculated from electron energy loss spectroscopy (EELS) spectra, which provides an indication for the presence of hydrocarbons and the number of graphene layers. As known in the art, EELS can be used to determine the local thickness of a wide range of materials in a reliable and repeatable manner.
In embodiments, EELS spectra are collected in five random locations (sample points) over the surface area of the graphene layer. This approach was applied in the illustrative comparison (see further hereinbelow) of a sample support in accordance with embodiments of the present disclosure (i.e. obtained by a method as discussed hereinabove) to similar prior-art sample support devices comprising a graphene layer, as commercially available. However, to avoid any possible ambiguity in testing a device, a sample support grid in accordance with embodiments of the present disclosure can be detected by measuring the thickness (using the EELS log-ratio absolute method referred to hereinbelow) of the graphene layer in a single point that corresponds to the geometrical center of the graphene layer, or to the geometrical center of that part of the graphene layer that is suitable and clearly intended for effective use as the sample support and observation area.
The local thickness can be calculated from the EELS data by the log-ratio absolute method, i.e. in accordance with the following equation:
in which t refers to the thickness to be determined, and λ refers to the mean free path of electron inelastic scattering. The ratio of intensity for zero-loss electrons (I0) to the total transmitted intensity (It) can be directly obtained by the EELS measurement. Appropriate values of this mean free path parameter λ are generally known for most materials of interest, including carbon-based materials, such as graphene. This equation is based on the assumption of a Poisson distribution of the electron energy loss. This approach to determine an absolute measure of the specimen thickness is considered to be well-known and, furthermore, is implemented in many (or most) standard electron microscopy software packages, such as Digital Micrograph (DM).
The application of this technique to a vacuum region (in a hole) allows a background signal to be determined during the measurement. In the illustrative comparison discussed hereinbelow, the average value that was observed for such vacuum hole was 0.78 nm, which was subtracted from the calculated values. Thus, in a test protocol to measure the thickness of the graphene layer, a background signal (thickness measurement) may be subtracted from the calculated thickness (e.g. for the geometrical center of the layer), in which the background signal is a thickness value calculated on the basis of a vacuum region.
The total error ΔX of the (averaged) thickness measurement,
can be estimated as ΔX=√{square root over (ES2+Eα2)}, where Ea is the statistical error
with s being the empirical standard deviation and n the number of measurements (e.g. 5). Es represents the accuracy of the method, which can be estimated to be 20% of the measured value.
The following example demonstrates the quality achievable for a graphene layer transferred onto a TEM grid for use as a sample support.
Referring to
Cleaner graphene grids enable the visualization of crystalline structure, and the ability to better distinguish organic residues from the transfer process and the organic compounds being investigated. For example, for carbon-based materials, e.g. carbon nanotubes, this may be highly desirable.
The thickness of the support also influences the quality. For example, to visualize surface ligands or polymer coatings, a high-quality and clean graphene layer may be required.
Furthermore, the high thermal and electrical conductivity of graphene offer further aspects. For example, metal halide perovskites are highly beam-sensitive materials; lead clusters are nearly instantly formed during TEM analysis.
A high electric conductivity also avoids charging effects as would be present on insulating samples or supports.
The ADF STEM images were segmented using an algorithm, implemented in Matlab, with multilevel thresholds using Otsu's method. The segmentation result shows clean graphene areas as black regions and regions comprising different structures or materials, e.g. polymer, hydrocarbons, copper nanoparticles, etc., as various grey intensity values. The average percentage of clean graphene area is then calculated (area of ‘black’ segmentation component over total area) and averaged over the ten different ADF STEM images. The average percentages of ‘clean’ graphene that were thus obtained are summarized in the table hereinbelow. This estimate area percentage of clean graphene is, in a limited sense, related to the ‘coverage’, a property that is commonly referred to in publications and/or commercial information regarding such graphene grids. However, it is to be noted that this measure is indicative of the relative area covered by clean graphene, whereas ‘coverage’ refers to the area covered by graphene regardless of it being clean or not. In other words, the estimated area percentage can be considered as a measure of cleanness.
The grid 21710-25 is described by the manufacturer Ted Pella as a pelco single layer graphene TEM support film on lacey carbon (300 M Cu).
The Graphenea sample support is described as suspended graphene on TEM grid (Quantifoil gold).
The sample grids from the supplier ACS Materials relate to graphene on a lacey carbon TEM grid of respectively 300 M Cu (E23CL105) and 2000 M Cu (E32C0111), and to graphene on a silicon nitride TEM grid with 2 μm holes (EOBSN105).
The alleged graphene area coverage was, for all of the mentioned prior-art examples, at least 75%, and, in the case of the Graphenea TEM grid, even greater than 95%.
In a further experiment, bright field transmission electron microscopy (BF-TEM) images were acquired at a magnification of 7100×, at five reference positions: four positions between the centre and the rim of the grid (north, east, south, west), and in the centre of the grid. The acquired BF-TEM images were segmented using a Random Forest Classifier. The Random Forest Classifier enables the segmentation of different regions: clean areas of graphene and areas containing residues of polymer and/or hydrocarbon contamination. Three to five regions of the different target areas (clean area and areas with residues) were manually selected to train the classifier, such that the algorithm is trained to recognize the different areas in the full image. Once trained, the classifier can be applied to other images. The results are shown in FIG. 37. From left to right, a graphene grid image used as input, the segmented clean graphene region and the segmented area with residues are shown.
From the obtained segmented images, the number of pixels in the clean graphene and the residue-containing regions can be calculated. This results in an integrated density from which cleanness metrics can be calculated as the ratio of clean graphene over total area and the ratio of area with residues over total area. An overview of results obtained for a few different brands currently available is given in the table hereinbelow:
Diffraction patterns can also be acquired. A diffraction pattern is two-dimensional cut, obtained from the intersection of the Ewald sphere and the reciprocal lattice. If the graphene is tilted, the Ewald sphere intersects the cones at a certain height, resulting in blurring of the diffraction points and stretching of the overall pattern perpendicular to the tilt axis. In addition, the maximum intensity of the diffraction spot decreases due to this blurring. Both effects can be measured and provide a quantitative measure of the sample roughness Rrms (see e.g. Thomsen et al, “Suppression of intrinsic roughness in encapsulated graphene,” Phys. Rev. B 96, 014101, and Kirilenko et al, “Measuring the corrugation amplitude of suspended and supported graphene,” Phys. Rev. B 84, 235417).
with l the maximum intensity of a diffraction spot and g the distance of a diffraction spot in reciprocal space from the zero-order position. g is measured as half the distance between a spot and its complementary spot on the opposite side of the tilt axis. This formula shows that a measure for Rims can be extracted from the slope of In (l) versus g2. An illustrative diffraction pattern (with indications of/and g) is shown in
Electron energy loss spectroscopy (EELS-STEM) datasets were furthermore acquired to evaluate the cleanness and flatness of the graphene. The spectra are acquired using a TEM at a voltage of 80 kV to avoid damage in the structure of graphene.
Carbon structural variations in graphene can be investigated using the carbon edge (284 eV) in an EEL spectrum, due to the distinctive variation in the spectra resulting from differences in the electronic structure. When the crystallinity of the sample decreases or the amount of hydrocarbon contamination is higher (i.e. an increase in the degree of sp3 bonding, indicated by the sigma peak), the intensity of the π* peak consequently decreases. For example, the crystallinity and cleanness can, to some extent, be illustrated by the ratio of the intensities of the π* and σ* peaks (π*/σ*).
Secondary electron scanning electron microscopy (SE-SEM) images were acquired to evaluate the coverage. The images were acquired at the five standard positions already discussed hereinabove: 4 positions between the centre and the rim of the grid (north, east, south, west), and one in the centre of the grid. The images cover the area of a full 200M grid window, i.e. a 90 μm by 90 μm region.
The SEM images are segmented based on a Random Forest Classifier. For example, this Random Forest Classifier is implemented in Fiji—ImageJ in the plugin Trainable Weka Segmentation (DOI 10.5281/zenodo.59290). This plugin combines a collection of machine learning algorithms with a set of manually selected image features to produce pixel-based segmentations. This algorithm enables the segmentation of different regions: ‘graphene’ (holes covered by graphene), ‘empty’ (holes not covered or partially broken) and ‘grid’ (the supporting grid or holey film on top of the grid) in the graphene grids. For training the algorithm, three to five regions of the different target areas (graphene, empty and grid) are manually selected. With this input, the algorithm is trained to recognize the different areas in the full image. Afterwards, the trained classifier can be applied to other images. Results thus obtained are shown in
From the obtained segmented images, the number of pixels for the graphene and empty regions were calculated. This results in an integrated density from which the coverage ratio can be calculated as the ratio of graphene-covered over total area (or number, which is essentially equivalent if the size of the holes is assumed constant) of the holes (e.g. using the sum of the graphene-covered and empty hole areas as the fraction's denominator). An overview of the different products is given in the table hereinbelow:
It is furthermore to be noted that the coverage ratio for the device in accordance with embodiments of the present disclosure was based on a prototype device, which may allow for further optimization. It can reasonably be expected that the coverage can be improved even further, e.g. by improving the uniformity of the polymer layer when being applied, e.g. by spin coating.
The present application is the National Phase entry of International Patent Application No. PCT/IB2022/052893 filed Mar. 9, 2022.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/IB2022/052893 | 3/29/2022 | WO |
