In low-frequency electronics, a circuit board substrate is treated as a passive, purely insulating anchor. It simply holds copper traces in place. However, as wireless communications push into the millimeter-wave (mm-Wave) and Terahertz (THz) regimes—driven by the demands of 5G and 6G networks—this assumption completely breaks down. At ultra-high frequencies, the substrate active participates in, and often degrades, signal propagation.
To understand why, we must look at the behavior of dielectric materials under rapidly alternating electromagnetic fields.
The Intuition Behind Dielectric Loss
An alternating current (AC) signal traveling through a transmission line generates a rapidly changing electric field. This field penetrates the surrounding substrate material. At the microscopic level, the substrate contains bound charges and molecular dipoles. When exposed to an electric field, these dipoles shift and realign themselves to match the field's polarity.
At megahertz frequencies, this alignment happens effortlessly. But when frequencies climb to gigahertz (GHz) or terahertz (THz) levels, the electric field reverses direction billions or trillions of times per second. The substrate's molecular dipoles attempt to keep pace with this ultra-fast rhythm. Because no physical material can polarize instantaneously, a phase lag develops between the applied electric field and the material's polarization.
This lag acts like microscopic friction. Instead of cleanly storing and releasing electrical energy, the substrate absorbs a portion of the electromagnetic wave and dissipates it as heat. This phenomenon is known as dielectric loss, and its severity is governed by the material's loss tangent ($\tan \delta$) and relative permittivity ($\varepsilon_r$). As operating frequencies increase, this molecular friction worsens, leading to severe signal attenuation, reduced radiation efficiency, and degraded quality factors ($Q$) in resonant components.
The Failure of Standard Substrates
For decades, the electronics industry has relied on cost-effective substrates like FR-4 (a flame-retardant fiberglass-epoxy laminate) and standard silicon. While highly economical, these materials pose a fundamental barrier to high-frequency hardware design.
As shown by Mingqing Wang et al. in Design of Miniaturized Tri-band Wearable Antenna Based on Characteristic Mode Theory, standard FR-4 exhibits a sharp increase in dielectric loss at frequencies above 5 GHz. This trend severely limits wideband performance. By the time signals reach the upper bands of Ultra-Wideband (UWB) or early mm-Wave spectrum, the material constraints become severe. Meng Wang et al. demonstrated in A butterfly-like slot UWB antenna with WLAN band-notch characteristics for MIMO applications that the high dielectric loss of FR-4 directly caps the peak efficiency and gain of wideband antenna systems.
At even higher frequencies, such as the 24 GHz threshold for consumer mm-Wave devices, the energy dissipation becomes catastrophic. Bumhyun Kim et al. established in Performance Enhancement of mm-Wave Phased Arrays for Mobile Terminals Through Grounded Coplanar Waveguide Feeding Networks With via Fences that despite clever architectural routing, the physical properties of FR-4 cause continuous signal leakage and attenuation. In hardware-centric sensing and radar systems, this loss destroys phase precision. Andrea M Tonello and Sarmad Ahmed Shaikh observed in Performance Analysis of 180° HRR Coupler Used for Direction Finding with an Antenna Array that standard materials become entirely unsuitable around 10 GHz due to tangent loss, necessitating a transition to specialized low-loss alternatives to maintain signal integrity.
A similar crisis occurs inside integrated circuits. Standard silicon substrates possess a relatively high conductivity. Xin Hu et al. showed in Analysis of full-wave conductor system impedance over substrate using novel integration techniques that at high frequencies, this conductivity allows currents to propagate through the substrate itself. This introduces unwanted noise, causes electromagnetic interference (EMI), and degrades the self-resonance frequency of critical components.
Advanced Material Alternatives
To bypass these physical limitations, researchers are shifting toward specialized materials engineered with ultra-low loss tangents and low relative permittivity.
[Standard Silicon / FR-4] --> High Loss Tangent --> High Signal Dissipation (Heat)
[Quartz / High-Resistivity] --> Low Loss Tangent --> Efficient High-Frequency Propagation
For terahertz-range integrated circuits, replacing standard silicon with high-resistivity silicon substrates drastically minimizes substrate-induced energy dissipation. As detailed by Fengjuan Wang et al. in TSV-based hairpin bandpass filter for 6G mobile communication applications, leveraging high-resistivity silicon alongside vertical Through-Silicon Vias (TSVs) allows circuits to maintain remarkably low insertion losses even at frequencies as high as 0.4 THz.
For discrete components, quartz has emerged as an excellent alternative. Haoran Zhao et al. demonstrated in Micro-Electromechanical System-Based Parasitic Patch Antenna on Quartz Substrate for High Gain that quartz’s low dielectric constant and minimal loss tangent provide a superior platform for Ku-band RF applications, avoiding the performance degradation typical of silicon. Similarly, Yutian Lu et al. focused on synthesizing next-generation microwave and terahertz dielectric ceramics, such as enstatite, in High- Q enstatite microwave/terahertz dielectric ceramics modulated by phase transition and lattice distortion. By tuning phase transitions and lattice distortions, the authors achieved the ultra-low relative permittivity required to prevent signal delay and degradation in 5G and 6G infrastructure.
Even when exploring modern manufacturing techniques like additive manufacturing, dielectric constraints remain paramount. Ying Sun et al. noted in Data-Driven Bayesian Optimization Framework for Rapidly Developing Novel Wideband, Low-Profile Dipole Antenna With 3-D-Printed Technology that while 3D printing offers immense design flexibility, standard printing filaments suffer from high loss tangents that cause severe drops in total radiation efficiency at high frequencies. Consequently, maintaining performance across wide bandwidths requires the careful selection and optimization of specialized low-loss printing compounds.
Geometric and Structural Workarounds
Because specialized substrates like sapphire or silicon carbide can be prohibitively expensive for mass-market consumer devices (Kai Yang in SV-SAW RF filters based on low-cost 128°Y LiNbO3/SiO2/poly-Si/Si substrate for 6G cmWave wireless communications), researchers have developed geometric and structural techniques to manipulate how electromagnetic fields interact with the substrate.
One compelling approach is to alter the nature of the wave itself. Yuxi Lu et al. demonstrated in Multi-layer transmission line of spoof surface plasmon polaritons that engineers can use Spoof Surface Plasmon Polaritons (SSPPs)—engineered surface-wave structures—to tightly confine electromagnetic fields to the interface of a multi-layer transmission line. This confinement prevents the fields from deeply penetrating the underlying substrate, bypassing bulk dielectric losses and closing the gap between idealized simulations and physical prototypes.
Another strategy involves rethinking the material architecture entirely. Instead of relying on traditional metal patches over a substrate, Zhiwei Song and Hongxing Zheng developed a Dielectric Resonator Antenna (DRA) in Design of One-Eighth Spherical Dielectric Resonator Antenna for 5G Applications. By utilizing a hollow, 3D-printed dielectric shell rather than a solid metal radiator, the design eliminates the severe conduction and skin-effect losses that plague conventional metal antennas at high frequencies.
Where metal conductors must be used, their geometry can be optimized to combat related high-frequency phenomena like the skin effect, where current concentrates at the outer boundaries of a conductor. Yi Feng et al. showed in Characterization of Inkjet Printed Coplanar Waveguides for Flexible Electronics that a "variable ink-layer thickness" approach—thickening only the edges of printed silver nanoparticle tracks where current density is highest—significantly mitigates conductor losses on flexible, non-traditional substrates.
The Path Forward: Foundational Characterization
Ultimately, designing hardware for the next generation of wireless networks requires moving away from trial-and-error engineering and toward rigorous material science. As Ali Abedi et al. emphasized in Introduction to Special Issue on New Paradigms in Intelligent Spectrum Management, the transition to the Terahertz spectrum introduces signal attenuation challenges that cannot be solved without foundational dielectric characterization.
By experimentally measuring and modeling how various thermoplastics, ceramics, and substrates behave under ultra-high-frequency polarization, engineers can precisely calculate signal attenuation. This foundational data allows for the co-design of materials and geometries, ensuring that the substrates of tomorrow act as efficient conduits for high-frequency signals rather than energy sinks.
Key Foundational Work
Analysis of full-wave conductor system impedance over substrate using novel integration techniques (2005) — contributed 2 concept elements to this direction.
Characterization of Inkjet Printed Coplanar Waveguides for Flexible Electronics (2011) — contributed 1 concept elements to this direction.