The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptx
Investigation of hBN by THz Time Domain Spectroscopy
1. Investigation of Hexagonal Boron Nitride by Terahertz Time-Domain Spectroscopy Jon Leist Momentive Performance Materials, Inc. Mira Naftaly, Richard Dudley National Physical Laboratory
2.
3. The Terahertz Gap 10 GHz 100 GHz 1 THz 10 THz 100 THz 1000 THz 1 cm 1 mm 100 m 10 m 1 m 100 nm Microwave / mm-wave Sub-mm / Terahertz / Far IR mid IR NIR VIS UV Molecular rotation Intramolecular vibration Intermolecular vibration 100 eV 0.8 cm -1 1 meV 8.0 cm -1 10 meV 80 cm -1 100 meV 800 cm -1 1 eV 8000 cm -1 10 eV 80000 cm -1 ELECTRONICS PHOTONICS Terahertz e - transitions
4.
5. Hexagonal Boron Nitride (BN) Hexagonal Structure Analogous to graphite/diamond: Hexagonal crystal structure soft, lubricous Cubic crystal structure hard, abrasive Thermally conductive: Single crystal, in-plane ~300 W/mK Shape, 92% dense ~30-55 W/mK Electrically insulating: volume resistivity ~10 -15 ohm-cm Dielectric constant ~3.9 Loss tangent ~0.0004 at 1 MHz Other Powder Properties: PSD: fine 0.5-50um; coarse 100-400um a b c
6. BN Powders for a Variety of Applications Thermal Conductivity Increases 10X
15. Terahertz Facilities at NPL http://www.npl.co.uk/electromagnetics/terahertz/products-and-services/thz-time-domain-spectroscopy The National Physical Laboratory (United Kingdom) utilizes a powerful collection of equipment to perform Terahertz measurements on a routine basis. Please contact them if you are interested in studying your materials in this frequency range.
22. DISCLAIMER: THE MATERIALS, PRODUCTS AND SERVICES OF MOMENTIVE PERFORMANCE MATERIALS INC., MOMENTIVE PERFORMANCE MATERIALS USA INC., MOMENTIVE PERFORMANCE MATERIALS ASIA PACIFIC PTE. LTD., MOMENTIVE PERFORMANCE MATERIALS WORLDWIDE INC., MOMENTIVE PERFORMANCE MATERIALS GmbH, THEIR SUBSIDIARIES AND AFFILIATES DOING BUSINESS IN LOCAL JURISDICTIONS (collectively “SUPPLIERS”), ARE SOLD BY THE RESPECTIVE LEGAL ENTITY OF THE SUPPLIER SUBJECT TO SUPPLIERS’ STANDARD CONDITIONS OF SALE, WHICH ARE INCLUDED IN THE APPLICABLE DISTRIBUTOR OR OTHER SALES AGREEMENT, PRINTED ON THE BACK OF ORDER ACKNOWLEDGMENTS AND INVOICES, AND AVAILABLE UPON REQUEST. ALTHOUGH ANY INFORMATION, RECOMMENDATIONS, OR ADVICE CONTAINED HEREIN IS GIVEN IN GOOD FAITH, SUPPLIERS MAKE NO WARRANTY OR GUARANTEE, EXPRESS OR IMPLIED, (i) THAT THE RESULTS DESCRIBED HEREIN WILL BE OBTAINED UNDER END-USE CONDITIONS, OR (ii) AS TO THE EFFECTIVENESS OR SAFETY OF ANY DESIGN INCORPORATING SUPPLIERS’ PRODUCTS, MATERIALS, SERVICES, RECOMMENDATIONS OR ADVICE. AFOREMENTIONED EXCLUSIONS OR LIMITATION OF LIABILITY ARE NOT APPLICABLE TO THE EXTENT THAT THE END-USE CONDITIONS AND/OR INCORPORATION CONDITIONS CORRESPOND TO THE RECOMMENDED CONDITIONS OF USE AND/OR OF INCORPORATION AS DESCRIBED BY SUPPLIER IN ITS PRODUCT DATA SHEET AND/OR PRODUCT SPECIFICATIONS. EXCEPT AS PROVIDED IN SUPPLIERS’ STANDARD CONDITIONS OF SALE, SUPPLIERS AND THEIR REPRESENTATIVES SHALL IN NO EVENT BE RESPONSIBLE FOR ANY LOSS RESULTING FROM ANY USE OF ITS MATERIALS, PRODUCTS OR SERVICES DESCRIBED HEREIN. Each user bears full responsibility for making its own determination as to the suitability of Suppliers’ materials, services, recommendations, or advice for its own particular use. Each user must identify and perform all tests and analyses necessary to assure that its finished parts incorporating Suppliers’ products, materials, or services will be safe and suitable for use under end-use conditions. Nothing in this or any other document, nor any oral recommendation or advice, shall be deemed to alter, vary, supersede, or waive any provision of Suppliers’ Standard Conditions of Sale or this Disclaimer, unless any such modification is specifically agreed to in a writing signed by Suppliers. No statement contained herein concerning a possible or suggested use of any material, product, service or design is intended, or should be construed, to grant any license under any patent or other intellectual property right of Suppliers or any of its subsidiaries or affiliates covering such use or design, or as a recommendation for the use of such material, product, service or design in the infringement of any patent or other intellectual property right. Momentive, the M-design logo and ‘the science behind the solutions’ are trademarks of Momentive Performance Materials Inc.
Editor's Notes
The Terahertz region of the spectrum lies between those regions typically associated with electronics and photonics; 20 microns to 2-mm This spectral domain hosts low-frequency crystalline lattice vibrations (phonon modes), hydrogen bonding stretches, and other intermolecular vibrations of molecules.
Terahertz rays interact with and are absorbed by some materials more than others. This interaction is governed by both a material’s density AND structure (for instance, T-rays are heavily absorbed by water and water-bearing materials, whereas they are only weakly absorbed by dry biological materials) The varying interaction of T-rays with matter opens a new imaging modality for applications ranging from airport security to medical diagnostics.
Boron nitride is a synthetic material available in a variety of morphologies analogous to the elemental forms of carbon; for hexagonal BN, the analogous carbon form is graphite It is the layered structure with weak van der Waals bonding, which BN shares with graphite, that is the source of its lubricity and thermal conductivity. The atomic arrangement of adjacent layers in the c-direction, with B and N alternating on the same location, that is the source of its electrical insulator nature, as well as its white color. The strong covalent bonding within layers is the source of its chemical inertness and high temperature oxidation properties. BN is available in a wide range of sizes and morphologies pictured here.
BN powders find application as a common additive in high end cosmetics and anti-chafing products BN finds use in Aerospace applications as a noise reduction and friction stability additive to friction materials, as well as a component of thermally sprayed abradable seals in aircraft engines The thermal conductivity of BN is leveraged in various thermal interface materials in electronics, where it is compounded into plastics and thermal greases (you can see in one application, the TC of the product was increased 10x)
Walk thru grades, ordered by decreasing porosity (first two are binder grades, last two are purified forms of HBN after hot pressing) A molten binder phase forms and allows crystal growth and orientation Uniquely, platelets align their planes with the pressing axis and grow grains of ~ 10 microns
Our measurements of THz transmission were carried out using a Terahertz time-domain spectrometer arrangement. This system consists of a Ti-sapphire pump laser incident on a GaAs Terahertz emitter, optics for manipulating the beam and its properties, a delay stage to allow the Terahertz pulse to be swept in the time domain, and a detector
Here we see an example of the refractive index and loss coefficient data vs frequency for the grade HBN HBN is seen to be birefringent, as expected, with negative birefringence. The slight increase of the refractive index with frequency is attributed to the tail of the lowest phonon resonance for BN at 23 THz
If you compare the four grades, you can see they all display the birefringence noted in the previous slide I’ll highlight that the loss coefficient for the HBT material shows some significant differences between the o-ray and e-ray losses. I’ll discuss this in a few moments.
Let’s examine the effect of porosity on the optical properties of the four grades. I’ve placed the o-ray and e-ray diagrams in the upper right corner for reference. What we see is that with the exception of the HBR grade, the e-ray refractive indices are all constant and the o-ray refractive indices decrease with increasing porosity. It’s not surprising that the refractive index should decrease with porosity, since less material would lie in the optical path. It is notable that the e-ray is nearly constant. We are proposing that this points to the location of the porosity or binder (in the case of the two grades with a binder). BN has a very inert c-plane. The most active sites are at the edge of the platelets. The material grows most quickly in the a and b directions. At the edge, you have a low number of N-H functionalities hanging around. Let’s examine HBN for a moment. Notice that one o-ray measurement is slightly below the trend. HBN contains 4% boric oxide binder, which has a refractive index of 1.75. If you calculate a weighted average of refractive indices for 96% BN at this porosity level and 4% Boric Oxide, you get a refractive index value of 2.072. This is very close to the value seen for the one o-ray measurement. Finally, what about this HBR grade, the grade with a calcium borate binder. First, note that the birefringence values for all the grades lie on a trend line with porosity. So, why do the refractive indices fall below the trend? It is possible that the calcium borate binder donates oxygen to BN to form a variety of boron-oxygen, boron nitrogen oxygen, and calcium nitrogen oxygen compounds. If these have lower refractive indices than BN, this would explain the observed low values in the HBR samples. Unfortunately, we can only speculate at this point, as no data could be found in the literature. Notice that the o-ray and e-ray values are shifted about the same distance below the trends lines, so it seems likely our proposal about the location of the binder phase would apply to HBR as well.
Let’s talk about loss in ceramic materials. In ceramic materials, loss is caused by combined contributions of absorption and scattering. If absorption is the dominant mechanism, then samples with higher porosity will have reduced losses because less material lies in the beam path. Conversely, if scattering predominates, then loss will rise with porosity due to the greater presence of scattering centres. As this figure shows, among the four grades of BN studied, loss increases with porosity, indicating that scattering is the dominant cause (see discussion below). With the exception of grade HBR, there is a linear trend in e-ray losses with porosity. The value extrapolated to fully dense (zero porosity) is close to zero. This indicates that THz absorption in BN is very low. In the HBT grade, losses in the o-ray orientations are much higher than in the e-ray orientation. This is consistent with the view that porosity is segregated at the edges of the platelets, thus causing greater scattering of the beam polarised in the ab-plane. The HBR grade exhibits higher than expected loss. In the discussion above, it was speculated that HBR may contain a significant fraction of compounds incorporating Ca, O, N, and B. All such compounds are polar to a much greater degree than BN, and therefore highly absorbing at THz frequencies, giving rise to absorption losses. Moreover, grains of calcium-containing compounds are known to approach 100 m in size, causing greatly increased scattering from interfaces.
If you’ll bear with me, I want to see if we can get some insight into the size of the scattering elements in the hot pressed BN grades. As a reminder to some, recall two different scattering regimes based on the size of the scattering element with respect to the wavelength of radiation. Mie scatttering pertains to scattering elements on the order of the same size as the wavelength of the radiation, whereas Rayleigh scattering describes the case of very small scattering elements with respect to the wavelength of radiation. Mie scattering follows a f-squared relationship with frequency, whereas Rayleigh scattering follows a f to the 4 th power relationship. The graph on the left shows the loss coefficient data for HBN plotted against the frequency. We have also fitted an equation having Mie and Rayleigh scattering components, each component having an associated coefficient. Now let’s move to the graph on the right. Here we have taken a ratio of the two scattering coefficients as an indicator of which kind of scattering is at work in the materials and plotted this against porosity for the four grades. It is seen that in all grades and orientations the ratio S Mie f 2 / S Rayleigh f 4 > 1, signaling that the f 2 scattering is the dominant mechanism. This suggests that a significant proportion of the scatterers are of the order of >10 m. Indeed, the HBR grade showed only f 2 scattering, marking the absence of small scatterers. This is consistent with the known size of calcium-compound grains in the material, which are known to be of the order of ~100 m [23]. The ratio of scattering coefficients was significantly higher for the pape orientation than for pe or papa , which supports the view that in this orientation the beam interacts with the edges of relatively large platelets. Notably, the ratio decreases with porosity, indicating that in more porous material the size of particles and pores may be smaller than in more highly dense material.
THz time-domain spectroscopy was used to probe four grades of hot pressed boron nitride Major takeaways were: We confirmed the birefringence of hexagonal BN We observed that the refractive index and loss coefficient are dependent on the porosity and binder We discovered that the porosity or binder is located at the edges of the BN c-planes, and we got an idea of the size of the scattering elements. We are currently investigating pyrolytic boron nitride (deposited by CVD methods), as well as BN anti-reflective coatings on Si and BN incorporated into nonpolar polymers
Several
In the top left figure, we see an example of the detector signal vs. time for three cases: no sample in place, a 2.5mm sample, and a 5mm sample. Note the shifted time and decreased amplitude. In the bottom right figure, we see this same data pictured as an amplitude vs. frequency graph for the same three cases. Using two different thicknesses, the loss coefficient and refractive index of the material can be calculated.
The axis directions of the perpendicular-cut samples were determined by observing the time-domain trace of the transmitted THz beam. The sample was positioned in the beam, and was rotated so as to obtain a single-peak trace at one of two delay positions Particular attention was given to minimising the residual features of the other trace. This is because the absence of such features signals good alignment of the THz beam polarisation to the c‑axis, either parallel ( papa ) or orthogonal ( pape ). However, as seen in the figure, it was not possible to eliminate these residues completely, indicating an imperfect mutual alignment of the platelets in the sample.