Many gamers around the world, who made the Xbox 360 era, are very familiar with the situation when their console turned into a frying pan in which they could fry eggs. A similar sad situation occurs not only with game consoles, but also with phones, laptops, tablets and much more. In principle, almost any electronics can experience thermal shock, which can lead not only to its breakdown and upset feelings of its owner, but also to a “bad-boom” of the battery and serious injuries. Today we will get acquainted with a study in which scientists from Stanford University, like Nick Fury from comics, created a shield that protects heat-sensitive electronic parts from overheating and, as a result, prevents them from breaking. How did scientists manage to create a thermal shield, what are its main components and how effective is it? We learn about this and not only from the report of the research group. Go.
Research basis
The problem of overheating has been known for a very long time, and scientists solve it in a variety of ways. One of the most popular is the introduction of glass, plastic and even layers of air, which serve as a kind of insulators of thermal radiation. In modern realities, this method can be improved by reducing the thickness of the protective layer to a few atoms without losing its thermal insulation properties. This is exactly what the researchers did.
We are talking, of course, about nanomaterials. However, their use in thermal insulation was previously complicated by the fact that the wavelength of heat carriers (phonons*) is much shorter than that of electrons or photons.
Phonon* — a quasiparticle, which is a quantum of the oscillatory motion of crystal atoms.
In addition, due to the bosonic nature of phonons, it is impossible to control them by voltage (as is done with charge carriers), which generally complicates the control of heat transfer in solids.
Previously, the control of the thermal properties of solids, as the researchers remind us, was carried out by means of nanolaminate films and superlattices due to structural disorder and high density of interfaces, or by means of silicon and germanium nanowires due to strong phonon scattering.
To a number of thermal insulation methods described above, scientists are confidently ready to attribute two-dimensional materials, the thickness of which does not exceed several atoms, which makes it easy to control them on an atomic scale. In their study, they used van der Waals (vdW) assembly of atomically thin 2D layers to achieve very high thermal resistance throughout their heterostructure.
Van der Waals forces* - forces of intermolecular / interatomic interaction with an energy of 10-20 kJ / mol.
The new technique made it possible to obtain thermal resistance in a vdW heterostructure 2 nm thick, comparable to that in a SiO2 (silicon dioxide) layer 300 nm thick.
In addition, the use of vdW heterostructures made it possible to control thermal properties at the atomic level by layering heterogeneous two-dimensional monolayers with different atomic mass densities and vibrational modes.
So, let's not pull the cat by the whiskers and proceed to consider the results of this amazing study.
Results of the study
First of all, let's get acquainted with the microstructural and optical characteristics of the vdW heterostructures used in this study.
Image #1
On the image 1a shows a cross-sectional diagram of a four-layer heterostructure consisting of (from top to bottom): graphene (Gr), MoSe2, MoS2, WSe22 and a SiO2/Si substrate. To simultaneously scan all layers, use Raman laser* with a wavelength of 532 nm.
Raman laser* is a type of laser in which Raman scattering is the main mechanism of light amplification.
Raman scattering of light, in turn, is the inelastic scattering of optical radiation by substance molecules, which is accompanied by a significant change in the radiation frequency.
To confirm the microstructural, thermal and electrical homogeneity of heterostructures, several methods were used at once: scanning transmission electron microscopy (STEM), photoluminescence spectroscopy (PL), Kelvin probe microscopy (KPM), scanning thermal microscopy (SThM), as well as Raman spectroscopy and thermometry .
Picture 1b shows us the Raman spectrum of the Gr/MoSe2/MoS2/WSe22 heterostructure on the SiO2/Si substrate in the place marked with a red dot. This graph shows the signature of each monolayer in the array of layers as well as the signature of the Si substrate.
On the 1c—1f shows dark-field STEM images of the Gr/MoSe2/MoS2/WSe22 heterostructure (1s) and Gr/MoS2/WSe22 heterostructures (1d—1f) with different lattice orientations. STEM images show gaps atomically close to vdW without any contamination, which makes it possible to fully see the total thickness of these heterostructures. The presence of interlayer bonding was also confirmed over large scan areas by means of photoluminescence (PL) spectroscopy (1g). The photoluminescent signal of individual layers inside the heterostructure is significantly suppressed compared to the signal of an isolated monolayer. This is explained by the process of interlayer charge transfer due to the close interlayer interaction, which becomes even stronger after annealing.
Image #2
In order to measure the heat flux perpendicular to the atomic planes of the heterostructure, an array of layers was structured in the form of four-probe electrical devices. The top layer of graphene is in contact with palladium (Pd) electrodes and is used as a heater for Raman thermometry measurements.
This electrical heating method provides an accurate quantification of the power input. Another possible heating method, optical, would be more difficult to implement due to ignorance of the absorption coefficients of individual layers.
On the 2a the diagram of a four-probe measurement is shown, and 2b shows a top view of the structure under test. Schedule 2s shows the measured heat transfer characteristics for three devices, one containing only graphene and two containing arrays of Gr/WSe22 and Gr/MoSe2/WSe22 layers. All variants demonstrate the ambipolar behavior of graphene, which is associated with the absence of a band gap.
It was also found that current conduction and heating occur in the upper layer (in graphene), since its electrical conductivity is several orders of magnitude higher than that of MoS2 and WSe22.
To demonstrate the homogeneity of the devices under test, Kelvin probe microscopy (KPM) and scanning thermal microscopy (SThM) measurements were taken. On the chart 2d KPM measurements are displayed with the identification of a linear potential distribution. The results of SThM analysis are shown in 2. Here we see a map of electrically heated Gr/MoS2/WSe22 channels, as well as the presence of uniformity in surface heating.
The scanning techniques described above, in particular SThM, confirmed the homogeneity of the investigated structure, that is, its homogeneity, in terms of temperatures. The next step was to quantify the temperature of each of the constituent layers using Raman spectroscopy (i.e., Raman spectroscopy).
All three devices were tested, the area of each of which was ~40 µm2. In this case, the heater power was changed by 9 mW, while the absorbed laser power was lower than ~5 μW for a laser spot area of ~0.5 μm2.
Image #3
On the chart 3a one can see an increase in the temperature (∆T) of each layer and substrate as the heater power increases in the Gr/MoS2/WSe22 heterostructure.
The slopes of the linear function for each material (layer) indicate the thermal resistance (Rth=∆T/P) between the individual layer and the heat sink. Considering the uniform distribution of heating over the area, it is quite easy to analyze thermal resistances from the lower to the upper layer, during which their values are normalized by the channel area (WL).
L and W are the channel length and width, which significantly exceed the SiO2 substrate thickness and the lateral thermal heating length, which is ~0.1 µm.
Therefore, we can derive the formula for the thermal resistance of the Si substrate, which will look like this:
Rth,Si ≈ (WL)1/2 / (2kSi)
In this situation kSi ≈ 90 W m–1 K–1, which is the expected thermal conductivity of such a highly doped substrate.
The difference between Rth,WSe2 and Rth,Si is the sum of the thermal resistance of 2 nm thick SiO100 and the boundary thermal resistance (TBR) of the WSe2/SiO2 interface.
Putting together all the above aspects, we can establish that Rth,MoS2 - Rth,WSe2 = TBRMoS2/WSe2, and Rth,Gr - Rth,MoS2 = TBRGr/MoS2. Therefore, from the graph 3a you can extract the TBR value for each of the WSe2/SiO2, MoS2/WSe2 and Gr/MoS2 interfaces.
Next, the scientists compared the total thermal resistance of all heterostructures, measured by Raman spectroscopy and thermal microscopy (3b).
Two-layer and three-layer heterostructures on SiO2 demonstrated effective thermal resistance in the range from 220 to 280 m2 K/GW at room temperature, which is equivalent to the thermal resistance of SiO2 with a thickness of 290 to 360 nm. Despite the fact that the thickness of the studied heterostructures does not exceed 2 nm (1d—1f), their thermal conductivity is 0.007–0.009 W m–1 K–1 at room temperature.
Image #4
Figure 4 shows the results of measurements of all four structures and the boundary thermal conductivity (TBC) of their interfaces, which allows us to evaluate the degree of influence of each of the layers on the previously measured thermal resistance (TBC = 1 / TBR).
The researchers note that this is the first ever measurement of TBC for atomically close interfaces between separate monolayers (2D/2D), in particular between WSe2 and SiO2 monolayers.
The TBC of the monolayer WSe2/SiO2 interface is lower than that of the multilayer WSe2/SiO2 interface, which is not surprising, since there are significantly fewer bending phonon modes available for transmission in the monolayer. Simply put, the TBC of the interface between the 2D layers is lower than the TBC of the interface between the 2D layer and the 3D SiO2 substrate (4b).
For a more detailed acquaintance with the nuances of the study, I recommend looking at и to him.
Finale
This study, according to the scientists themselves, gives us knowledge that can be applied in the implementation of atomic thermal interfaces. This work showed the possibility of creating heat-insulating metamaterials, the properties of which are not found in nature. In addition, the study also confirmed the possibility of making the most accurate measurements of the temperature of such structures, despite the atomic scale of the layers.
The heterostructures described above can become the basis for ultra-light and compact thermal "shields" capable, for example, of removing heat from hot spots in electronics. In addition, this technology can be used in thermoelectric generators or in thermally controlled devices, increasing their performance.
This study once again confirms that modern science is seriously carried away by the principle of "efficiency in a thimble", which cannot be called a stupid idea, given the limited resources of the planet and the ongoing growth in demand for all kinds of technological innovations.
Thank you for your attention, stay curious and have a great week everyone! 🙂
Thank you for staying with us. Do you like our articles? Want to see more interesting content? Support us by placing an order or recommending to friends, 30% discount for Habr users on a unique analogue of entry-level servers, which was invented by us for you: (available with RAID1 and RAID10, up to 24 cores and up to 40GB DDR4).
Dell R730xd 2 times cheaper? Only here in the Netherlands! Dell R420 - 2x E5-2430 2.2Ghz 6C 128GB DDR3 2x960GB SSD 1Gbps 100TB - from $99! Read about
Source: habr.com