1. Metamaterials and metasurfaces (the short version)

• Dimensions, not composition, set the behaviour: when structures much smaller than the wavelength (metal patterns, resonators, cavities, thin films with lattices) are arranged periodically, the whole behaves as one medium whose effective permittivity and permeability (for electromagnetic waves) or effective density and bulk modulus (for sound) can take values the composition alone cannot give. The review by Smith et al. states that artificially constructed metamaterials can show electromagnetic characteristics unlike those of any conventional material, such as artificial magnetism and negative refractive indexSourced [Source 1]
• A metasurface makes a wavefront with a thin sheet: a two-dimensional array whose elements impose different phase delays sets the direction of reflection or refraction. Yu et al. showed that abrupt phase changes over a wavelength-scale distance allow the laws of reflection and refraction to be generalisedSourced [Source 4]
• The difficulty lies on the deployment side: narrow bandwidth, loss in metals and dielectrics, the components that switch the elements, and the field comparison (against existing repeaters) create the distance between a clean laboratory curve and the field. The 2024 roadmap on active metamaterials likewise says that their added complexity makes progress hard, particularly beyond theory and laboratory experimentsSourced [Source 20]

2. Principle — structures smaller than the wavelength act as an effective medium

A wave does not “see” the individual details of structures much smaller than its wavelength; it feels them as an averaged medium. Designing this averaged (effective) medium through the shape, size and arrangement of the structure is the idea behind metamaterials. The review by Smith et al. describes metamaterials showing electromagnetic properties that conventional materials lack (artificial magnetism, negative refractive index) and the value of expanding our view of what constitutes a materialSourced [Source 1].

On the theory side, Pendry showed that a slab of negative-refractive-index material can focus all Fourier components of an image, including non-propagating ones (a “superlens”). He stated that this could be realised in the microwave band with then-current technology, and simulated that a thin slab of silver could resolve objects only a few nanometres across at visible frequenciesSourced [Source 2] (the optical side is covered in our explainer on metalenses and flat optics). On the experimental side, Schurig et al. reported in 2006 a microwave-band experiment in which a metamaterial cloak “hid” a copper cylinder, reducing scattering and shadow. They presented it as the first practical realisation of a cloak following a theory that held such a cloak possible, at least over a narrow frequency bandSourced [Source 3].

3. Metasurfaces — setting the phase place by place

Yu et al. showed experimentally that a two-dimensional array imposing abrupt phase changes at the wavelength scale (metallic antennas on silicon) produces anomalous reflection and refraction, in excellent agreement with generalised laws derived from Fermat’s principle. They also showed that such phase discontinuities can generate optical vorticesSourced [Source 4]. Conventional optical components build up phase gradually as the wave propagates; a metasurface instead imposes the phase abruptly on a thin surface, so the thickness can be kept to about a wavelengthSourced [Source 4].

As a way to switch electromagnetic phase electrically, Cui et al. proposed the coding metamaterial, in which two kinds of unit cell with 0 and π phase response are treated as “0” and “1”. With a biased diode switching each cell between 0 and 1, and a field-programmable gate array (FPGA) changing the sequence, a single surface becomes a “programmable metamaterial” that manipulates waves in different ways; they confirmed this by numerical simulation and experimentSourced [Source 5]. Di Renzo et al. sort RIS implementations into large arrays of inexpensive antennas, usually spaced half a wavelength apart, and metamaterial-based surfaces whose scattering elements are much smaller than the wavelength. They state that RISs need the largest number of scattering elements, but each is backed by the fewest and least costly components, and that no power amplifiers are usually neededSourced [Source 9].

4. Our calculation: the wavelength at 28 GHz, and the number of elements

Let us check numerically that dimensions scale with wavelength. Taking the speed of light as 299,792,458 m/s, the wavelength at 28 GHz is about 10.7 mm and at 39 GHz about 7.7 mmOur calculation. If elements are packed at half-wavelength pitch, a 50 cm square holds about 93 elements per side at 28 GHz (about 8,650 elements) and about 130 per side at 39 GHz (about 16,900 elements)Our calculation.

In their announcement of 28 February 2025, Japan Display (JDI), KDDI and KDDI Research presented a portable liquid-crystal metasurface reflector for the 28 GHz band that is about 50 cm square and about 8 kg, switched on and off from a smartphone, with a changeable scattering pattern (manufacturers’ stated values)Sourced [Source 18]. The element pitch, the number of elements and how voltage is applied to the liquid crystal could not be confirmed from this announcement.

5. Radio-wave absorbers and stealth — the limits of bandwidth and thickness

One of the first practical applications of metamaterials is the radio-wave absorber. Landy et al. showed a design in which two resonators, coupling separately to the electric and magnetic fields, are placed in a single unit-cell layer to absorb nearly all incident radiation. The fabricated sample had a predicted absorbance of 96% and a measured peak absorbance above 88% at 11.5 GHz. Because the structure consists solely of metallic elements, the substrate can be chosen for other properties, they noteSourced [Source 6].

An absorption figure always comes with a frequency, an angle of incidence and a polarisation. The Landy value is a measurement at one frequency, 11.5 GHz, and does not demonstrate absorption over a wide bandSourced [Source 6]. There is also a fundamental limit between thickness and bandwidth for an absorber backed by a metal plate. From the analytic properties of the reflection coefficient of a metal-backed multilayer slab, Rozanov derived the lower bound on the thickness-to-bandwidth ratio of a physically realisable radar absorber, and showed that the least thickness of a 10 dB broadband dielectric radar absorber is 1/17 of the largest operating wavelengthSourced [Source 7].

Let us apply the 1/17 to some frequencies (taking the wavelength in vacuum, and reading the “largest operating wavelength” as the wavelength at the lowest frequency at which the absorber operates; that reading is our interpretation and contains an inference). Working from 1 GHz upward gives about 17.6 mm; from 3 GHz, about 5.9 mm; from 10 GHz, about 1.8 mm; and from 28 GHz, about 0.63 mmOur calculation. These numbers show that absorbing broadly down to low frequencies requires either giving up thinness or using another means such as magnetic materials (the bound is for the dielectric case; the abstract indicates that a different relation applies when magnetic materials are used)Sourced [Source 7].

6. A materials engineer’s view 1: loss, and tunable materials

7. RIS for 6G and radio-wave lenses for window glass — from laboratory to field

Millimetre waves at 28 GHz and above travel in straight lines, so places that radio waves do not reach (coverage holes) arise easily in the shadow of buildings and trees. The motivation for RIS and metasurface reflectors is to deliver the signal into such places by reflecting or focusing it with a surface, instead of adding base stationsSourced [Sources 15, 17]. Di Renzo et al. organise RIS as an enabling technology for smart radio environments (SREs). Compared with phased arrays, multi-antenna transmitters and relays, they say, RISs need many elements but few and cheap components per element and usually no power amplifiers, so they could be built with low cost, size, weight and power (C-SWaP)Sourced [Source 9].

The published experiments and demonstrations (Japanese examples)

The numbers and the “world first” claims are all those of the announcing companies and institutions. Please do not rank experiments that differ in conditions (measurement point, area, environment).

Standardisation, and critical views

On the standards side, ETSI (the European Telecommunications Standards Institute) announced on 25 April 2025 two new reports: “Implementation and Practical Considerations” (GR RIS 004, covering RIS unit-cell design, switching methods, RIS types, the RIS controller, costs, field trials and measurement methods) and “Diversity and Multiplexing of RIS-aided Communications” (GR RIS 005). It also said that it had updated its three earlier reports: GR RIS 001 (Use Cases, Deployment Scenarios and Requirements), 002 (Technological Challenges and Impact on Architecture and Standards) and 003 (Communication Models, Channel Models, Channel Estimation, and Evaluation Methodology)Sourced [Source 13]. These are group reports, not standards themselves.

In 3GPP, the network-controlled repeater (NCR), a repeater with added network control, was studied in Release 18. Technical report TR 38.867 V18.0.0 (September 2022) assumes the NCR to be an in-band RF repeater (a relay that amplifies the received band and re-transmits it) for the FR1 and FR2 bands, single-hop and stationary, and transparent to the UE, and notes that cost efficiency is a key consideration. The study covers side control information (beam, timing, UL-DL TDD configuration, ON-OFF, power control), its signalling, and repeater managementSourced [Source 14]. For RIS, a preprint by Åström et al. states that RIS was proposed as a study item in both 3GPP Releases 18 and 19 but was not continued in either and was left for possible future specification, and it sets out practical issues as the rationale. The same paper compares RIS with the network-controlled repeater standardised in Release 18Sourced (a non-peer-reviewed preprint; we did not check the 3GPP meeting records themselves) [Source 12].

Within research, views are divided. Björnson et al. sorted out misunderstandings that spread amid the rising expectations for RIS, and examined three myths: that current network technology can only control the transmitter and receiver, not the environment in between; that RIS achieves a better asymptotic array gain than conventional beamforming; and that the path loss is the same as with anomalous mirrors. They named two questions that must be answered for RIS to succeed: a convincing use case, and how to estimate channels and control an RIS in real timeSourced [Source 10]. Tang et al. organised the free-space path loss of RIS as a function of the distances to the transmitter and receiver, the RIS size, near-field and far-field effects, and the radiation patterns of antennas and unit cells, and reported that measurements of three fabricated RISs in a microwave anechoic chamber matched the models wellSourced [Source 11]. The difference between an anechoic chamber and a reflective field site is set out in Chapter 10.

8. Acoustic and elastic-wave metamaterials — sound insulation and vibration control

An ordinary wall blocks sound better the heavier it is (the mass law). Keeping weight low while still blocking sound is one motivation for acoustic metamaterials. In 2000, Liu et al. fabricated “locally resonant” sonic crystals and showed spectral gaps, ranges of frequencies where sound does not pass, with a lattice constant two orders of magnitude smaller than the relevant wavelength. The composite behaves as a material with effective negative elastic constants, and a 2 cm slab broke the conventional mass-density law of sound transmission by one or more orders of magnitude at 400 Hz, they reportSourced [Source 19]. Yang et al. built a simple membrane-type metamaterial, a thin film with a small mass attached, divided by a rigid grid, and report breaking the mass-density law by about 200 times in the 100–1000 Hz range. The eigenfrequencies can be tuned by placing a small mass at the centre of the membrane, and near-total reflection occurs at a frequency between two eigenmodes, where the in-plane average of the normal displacement is zeroSourced [Source 21]. The review by Ma and Sheng traces acoustic metamaterials over about 15 years, from an academic curiosity to an active field, from local resonances to phase manipulation, superabsorption, active control, and wave transport in moving fluids, elastic and mechanical metamaterialsSourced [Source 23].

View the original SVG schematic and exact annotations

On the practical side, in an announcement dated 6 January 2020 (US time), Nissan Motor showed at CES 2020 an “acoustic meta-material” in which a combination of a lattice structure and plastic film controls air vibrations to limit the transmission of wide-band noise, such as road and engine noise, in the 500–1200 Hz range. The company says it weighs one-fourth as much as the heavy rubber board mostly used to isolate this band while providing the same degree of sound isolation, and that its mass-production cost competitiveness is almost the same as, or possibly better than, current materials (manufacturer’s stated value; the basis of comparison, test method and effect in real vehicles are not given in the announcement)Sourced [Source 22]. Nissan also says it began research on meta-material technology around 2008Sourced [Source 22].

Working out the sound wavelengths: about 0.69 m at 500 Hz and about 0.29 m at 1200 Hz (taking the speed of sound in air as 343 m/s), and about 0.86 m at 400 Hz for Liu et al.Our calculation. A 2 cm slab is about 1/43 of the wavelength at 400 HzOur calculation. The relationship is clear: a thin plate controls a wavelength orders of magnitude longer than the 10.7 mm wavelength of electromagnetic waves at 28 GHz.

For elastic waves (vibration in solids), Hussein et al. organise “phononic” media, which have periodicity in the constituent materials, the internal geometry or the boundary conditions, as systems describable by a frequency band structure like an electronic band diagram, and survey a field that runs from trusses and ribbed shells to phononic crystals and metamaterialsSourced [Source 24]. Examples of vibration-damping products could not be confirmed in primary sources, so none are given here.

9. A materials engineer’s view 2: resonances drift

10. Laboratory performance versus deployment

Here we sort the above into “what was shown in the laboratory or demonstration” and “what deployment asks”. The right-hand column gives the sources in the reference list.

11. Limits, and what could not be confirmed

(1) How far it has been put into practice

What has been published are results at the stage of laboratory measurement, demonstration and field trial. The scale of commercial deployment, prices and long-term reliability could not be confirmed from published primary sources.

(2) Details of element structure and materials

The element pitch, element count and millimetre-wave dielectric loss of the liquid-crystal layer in the JDI/KDDI reflector, and the materials and element dimensions of the NTT DOCOMO/AGC metasurface lens, could not be confirmed from the announcements. The lattice dimensions, film material and measurement conditions of Nissan’s acoustic metamaterial also could not be confirmed from its announcement.

(3) What 3GPP and ETSI contain

We checked the text of 3GPP TR 38.867 V18.0.0. How RIS came not to become a study item in Releases 18 and 19 rests on the statement in the preprint by Åström et al.; we did not check the 3GPP meeting records themselves. We also did not check the text of the ETSI reports (GR RIS 001–005) and limit ourselves to ETSI’s announcement of 25 April 2025.

(4) The texts of standards

We have not checked the texts (which are paid) of IEC 62431, JIS R 1679, ISO 10140-2, ISO 354, ISO 10534-2, ISO 717-1, IEEE 149 and others, except the foreword, scope and contents of the JIS R 1679 preview. Test procedures, pass criteria and calibration are not described.

(5) Other applications

Optical metalenses (see our explainer on metalenses and flat optics), plasmonics (see our explainer on plasmonics and nanophotonics), thermal-radiation and radiative-cooling metasurfaces (see our explainer on thermal control materials and radiative cooling), real stealth systems and mass-produced active metamaterials are not covered here.

12. How to read evaluations and standards

What is measured: for radio waves, reflectivity (dB), absorption, the phase and amplitude of reflection and transmission, and antenna gain and radiation pattern, measured in a free-space set-up (horn antennas and a network analyser, an anechoic chamber). For sound, transmission loss (dB), sound absorption coefficient and surface impedance, measured in a reverberation room, in a pair of laboratory rooms with the specimen mounted between them, or in an impedance tube. We found no ISO, IEC or IEEE standard that names metamaterials in general. In practice the existing standards for each use (radio-wave absorbers, antennas, sound insulation of building elements, sound absorbers) are usedSourced (a statement of what could not be confirmed).

Which standard we match to which item is our own arrangement (Our calculation). Numbers and titles were checked on the official ISO page for ISO, the IEC Webstore for IEC, the IEEE SA page for IEEE, and the JISC JIS search for JIS. The two ASTM standards were checked on non-ASTM pages; the ASTM website was not opened.

How the standards differ in standing

• ISO, IEC and IEEE are international or internationally used standards; JIS is Japan’s national standard. The foreword of JIS R 1679 says that when it was established (2007) there was no corresponding international standard [27]. IEC 62431 was published in July 2008, but the relationship between the two and the current edition of the JIS were not checked here
• ASTM is a US-centred standards organisation. A property with the same name does not necessarily use the same method or conditions. ASTM E1050 and ISO 10534-2 are both normal-incidence impedance-tube methods, but we did not confirm that editions or details match
• ETSI and 3GPP are telecommunications standards bodies, and their documents are not material test standards. The RIS documents are technical reports or study reports, to be kept distinct from specifications [13, 14]

Cautions in reading the standards

• A standard is a test method, not a guarantee of performance. IEC 62431 and JIS R 1679 are methods for measuring reflectivity; we did not check whether they go as far as evaluating the operation of a surface whose reflection direction is switched. ISO 354 states that it is not intended for measuring the absorption characteristics of weakly damped resonators, so you need to confirm which standard a resonant structure was measured with (whether a resonant structure counts as a “weakly damped resonator” is our inference and has not been checked)
• The same named property can differ with the standard and its edition. The normal-incidence absorption coefficient of ISO 10534-2 is not comparable with the random-incidence coefficient of ISO 354 [32]. A single-number rating (ISO 717-1) merges band-by-band values, so the narrow peaks and dips of Fig. 6 are hard to see in a single number (our commentary)
• Read a manufacturer’s value together with the standard, edition and test conditions it follows. Nissan’s “the same insulation at one-fourth the weight” has no standard or conditions in the announcement [22]. The “above 24 dB” of NTT DOCOMO and AGC is received power at the focus with float glass taken as 0 dB, under normal incidence at 27.6 GHz [16]
• The gap between a standard’s range and conditions of use: laboratory standard results do not include flanking transmission or the reflective environment of the field [29]

13. Glossary

14. References (primary sources)

1. Smith D. R., Pendry J. B., Wiltshire M. C. K. “Metamaterials and Negative Refractive Index”, Science 305, 788–792 (2004), DOI 10.1126/science.1096796 — doi.org/10.1126/science.1096796
2. Pendry J. B. “Negative Refraction Makes a Perfect Lens”, Phys. Rev. Lett. 85, 3966–3969 (2000), DOI 10.1103/PhysRevLett.85.3966 — doi.org/10.1103/PhysRevLett.85.3966
3. Schurig D. et al. “Metamaterial Electromagnetic Cloak at Microwave Frequencies”, Science 314, 977–980 (2006), DOI 10.1126/science.1133628 — doi.org/10.1126/science.1133628
4. Yu N. et al. “Light Propagation with Phase Discontinuities: Generalized Laws of Reflection and Refraction”, Science 334, 333–337 (2011), DOI 10.1126/science.1210713 — doi.org/10.1126/science.1210713
5. Cui T. J. et al. “Coding metamaterials, digital metamaterials and programmable metamaterials”, Light: Science & Applications 3, e218 (2014), DOI 10.1038/lsa.2014.99 — doi.org/10.1038/lsa.2014.99
6. Landy N. I. et al. “Perfect Metamaterial Absorber”, Phys. Rev. Lett. 100, 207402 (2008), DOI 10.1103/PhysRevLett.100.207402 — doi.org/10.1103/PhysRevLett.100.207402
7. Rozanov K. N. “Ultimate thickness to bandwidth ratio of radar absorbers”, IEEE Trans. Antennas Propag. 48, 1230–1234 (2000), DOI 10.1109/8.884491 — doi.org/10.1109/8.884491
8. Khurgin J. B. “How to deal with the loss in plasmonics and metamaterials”, Nat. Nanotechnol. 10, 2–6 (2015), DOI 10.1038/nnano.2014.310 — doi.org/10.1038/nnano.2014.310
9. Di Renzo M. et al. “Smart Radio Environments Empowered by Reconfigurable Intelligent Surfaces: How It Works, State of Research, and The Road Ahead”, IEEE J. Sel. Areas Commun. 38, 2450–2525 (2020), DOI 10.1109/JSAC.2020.3007211 — doi.org/10.1109/JSAC.2020.3007211
10. Björnson E., Özdogan Ö., Larsson E. G. “Reconfigurable Intelligent Surfaces: Three Myths and Two Critical Questions”, IEEE Commun. Mag. 58(12), 90–96 (2020), DOI 10.1109/MCOM.001.2000407 — doi.org/10.1109/MCOM.001.2000407
11. Tang W. et al. “Wireless Communications With Reconfigurable Intelligent Surface: Path Loss Modeling and Experimental Measurement”, IEEE Trans. Wireless Commun. 20, 421–439 (2021), DOI 10.1109/TWC.2020.3024887 — doi.org/10.1109/TWC.2020.3024887
12. Åström M. et al. “RIS in Cellular Networks — Challenges and Issues”, arXiv:2404.04753 (preprint, version 2 of 7 November 2024, not peer-reviewed) — arxiv.org/abs/2404.04753
13. ETSI “ETSI releases new Reports on Reconfigurable Intelligent Surfaces” (25 April 2025) — etsi.org/newsroom/news/2522
14. 3GPP “TR 38.867 V18.0.0 (2022-09) Study on NR network-controlled repeaters (Release 18)” — 3gpp.org/ftp/Specs/archive/38_series/38.867
15. Nippon Telegraph and Telephone Corporation and NTT DOCOMO News release on the demonstration of user-tracking metasurface control (12 November 2021; in Japanese) — group.ntt/jp/newsrelease/2021/11/12/211112b.html
16. NTT DOCOMO and AGC Press release on turning window glass into a radio-wave lens with metasurface technology (26 January 2021, with the annex of experimental results; in Japanese) — agc.com/news/pdf/20210126.pdf
17. Japan Display and KDDI Research Press release on a liquid-crystal metasurface reflector whose reflection direction can be changed (7 October 2021; in Japanese) — j-display.com/news/release/detail/20211007000000.html
18. Japan Display, KDDI and KDDI Research Press release on reflecting millimetre waves between outdoor buildings with a liquid-crystal reflector (28 February 2025; in Japanese) — j-display.com/news/release/detail/20250227161354.html
19. Liu Z. et al. “Locally Resonant Sonic Materials”, Science 289, 1734–1736 (2000), DOI 10.1126/science.289.5485.1734 — doi.org/10.1126/science.289.5485.1734
20. Active Metamaterials Roadmap (multiple authors) “The 2024 Active Metamaterials Roadmap”, arXiv:2411.09711 (preprint) — arxiv.org/abs/2411.09711
21. Yang Z. et al. “Membrane-Type Acoustic Metamaterial with Negative Dynamic Mass”, Phys. Rev. Lett. 101, 204301 (2008), DOI 10.1103/PhysRevLett.101.204301 — doi.org/10.1103/PhysRevLett.101.204301
22. Nissan Motor Co., Ltd. “Nissan shows new, lightweight noise-reducing technology at CES” (6 January 2020) — global.nissannews.com/releases/nissan-shows-new-lightweight-noise-reducing-technology-at-ces
23. Ma G., Sheng P. “Acoustic metamaterials: From local resonances to broad horizons”, Sci. Adv. 2, e1501595 (2016), DOI 10.1126/sciadv.1501595 — doi.org/10.1126/sciadv.1501595
24. Hussein M. I., Leamy M. J., Ruzzene M. “Dynamics of Phononic Materials and Structures: Historical Origins, Recent Progress, and Future Outlook”, Appl. Mech. Rev. 66, 040802 (2014), DOI 10.1115/1.4026911 — doi.org/10.1115/1.4026911
25. IEC “IEC 62431:2008” product page (IEC Webstore) — webstore.iec.ch/en/publication/7006
26. Japanese Industrial Standards Committee (JISC) JIS search listing for “JIS R 1679” — jisc.go.jp (JIS R 1679)
27. Japanese Standards Association “JIS R 1679:2007” preview (foreword, scope, contents; in Japanese) — webdesk.jsa.or.jp/preview/pre_jis_r_01679_000_000_2007_j_ed10_ch.pdf
28. IEEE SA “IEEE 149-2021” page — standards.ieee.org/ieee/149/6667
29. ISO “ISO 10140-2:2021” page — iso.org/standard/79487.html
30. ISO “ISO 717-1:2020” page — iso.org/standard/77435.html
31. ISO “ISO 354:2003” page — iso.org/standard/34545.html
32. ISO “ISO 10534-2:2023” page — iso.org/standard/81294.html
33. Japanese Industrial Standards Committee (JISC) JIS search listing for “JIS A 1416” — jisc.go.jp (JIS A 1416)
34. Japanese Industrial Standards Committee (JISC) JIS search listing for “JIS A 1409” — jisc.go.jp (JIS A 1409)
35. Japanese Industrial Standards Committee (JISC) JIS search listing for “JIS A 1405” (Parts 1 and 2) — jisc.go.jp (JIS A 1405)
36. ANSI Webstore (a non-ASTM sales page) Listing giving the title and subject of ASTM E90 — webstore.ansi.org/standards/astm/astme9099
37. Intertek (a non-ASTM explanatory page) Explanation of the title and scope of ASTM E1050 — intertek.com/building/standards/astm-e1050

15. Claim-to-source audit

Last updated 10 October 2026. Sources are limited to primary material (peer-reviewed papers, official pages of 3GPP, ETSI, ISO, IEC, IEEE and JISC, and announcements by the companies and institutions concerned). The texts of standards are paid and could not be checked, apart from the foreword, scope and contents of the JIS R 1679 preview. The ASTM International website was not opened. Because the article includes structural readings and materials-design interpretations, those are marked as Commentary and kept separate from sourced fact. The performance of real stealth systems, details of element structures and materials, the 3GPP meeting records, the text of the ETSI reports and long-term field durability could not be confirmed and are not stated here. Figures from announcements are the stated values of the companies and institutions that made them, and are not meant for ranking experiments that differ in conditions. All figures are explanatory concept graphics. Figs. 2, 3, 4 and 6 are vector drawings; Figs. 1 and 5 are AI-generated images, and none of them shows a real photograph, measured data or a physical product.