Podcast Summary: "Why Diamond Transistors Are So Hard"

Podcast: Asianometry
Host: Jon Y
Episode Date: January 22, 2026

Episode Overview

In this episode, Jon Y explores the tantalizing prospects and fundamental challenges of using diamond as a material for transistors in power electronics. He contrasts diamond’s remarkable theoretical advantages against the immense scientific and practical hurdles involved in making diamond-based transistors a reality, offering listeners an engaging blend of semiconductor science, manufacturing realities, and industry context.

Key Discussion Points & Insights

1. Silicon’s Strengths and Limitations

• Bandgap:
  • Silicon’s bandgap is 1.12 electron volts (eV), which is relatively narrow.
  • Advantages: Lower voltage needed for switching.
  • Drawbacks: More susceptible to heat-induced leakage currents and “negative feedback” loops increasing unwanted heat.
  • [02:05] "With more heat, charge carriers can punch through the gate and generate a small leakage current. This current dissipates more heat, which worsens the leakage and makes even more heat—a negative feedback loop."
• Breakdown Field:
  • Silicon’s low breakdown field means it can’t easily handle high voltages and strong electric fields.
  • This limits its use in high-power, high-temperature settings like EV power inverters and high-frequency RF amplifiers.
  • [03:45] "So silicon isn't good to use in situations of high heat and strong electric fields... like, for example, the inverters in EVs that convert DC to AC power."

2. Exploring Alternative Wide Bandgap Materials

• Candidates:
  • Indium phosphide (1.35 eV) and gallium arsenide (1.42 eV): only slightly better than silicon and are toxic or rare.
  • Silicon carbide (3.26 eV) and gallium nitride (3.4 eV): popular but with drawbacks like low carrier mobility or instability at high frequencies.
  • [04:30] "Moreover, they're both toxic to humans. And on the rarer side, silicon carbide has relatively low carrier mobility..."

3. Diamond: The ‘Ultimate’ Semiconductor?

• Material Superiority:
  • Ultra-wide bandgap (5.5 eV), extremely high breakdown field (up to 10 MV/cm—33x silicon), and unmatched thermal conductivity (2,200 W/mK).
  • Carrier mobility: 3x that of silicon for electrons, up to 9x for electron holes (in pure diamond).
  • Radiation resistance and potential for miniaturization.
  • [07:20] "Perhaps the best thing is how diamond combines all of these properties. High thermal conductivity, high breakdown voltage, and radiation hardness and high carrier mobility in a single package."
• Applications:
  • Ideal for power electronics in extreme environments and for devices requiring very high density and reliability.
  • Some current use as a heat sink in high electron mobility transistors.
  • [08:15] "A diamond transistor has that unparalleled heat sink built right in."

4. The Manufacturing Challenge

• Crystal Growth Issues:
  • Unlike silicon (grown by Czochralski pulling from melt), diamond can’t be melted and reformed due to physical constraints—molten diamond turns into graphite unless under immense pressure.
  • High-pressure high-temperature (HPHT) method: Yields small, often impure crystals; expensive and size-limited.
  • [10:30] "We cannot produce a vat of diamond melt unless you put it under untenable high pressure conditions. Molten diamond turns into graphite, which we can uphold."
• Chemical Vapor Deposition (MPCVD):
  • Main modern method: Grows diamond on seed crystals, but growth is slow and large single-crystal seeds are hard to come by.
  • Issues with growth rate, uniformity, and cost (a 10mm wafer costs up to 10,000x more than silicon).
  • [13:20] "A two-carat gem quality diamond might take about two weeks to produce... single crystal wafers right now just 10 millimeters wide can cost up to 10,000 times of equivalent sized silicon."
• Methods to Scale up:
  • Growing diamond laterally from seeds, mosaic method (joining small seeds), and heteroepitaxy (growing on iridium) are all under exploration, but no breakthrough yet.
  • [15:00] "The mosaic method... the diamond grows over the mosaic, fusing together to create a single larger piece..."

5. The Doping Dilemma

• Why Doping Matters:
  • Allows for creation of N-type and P-type material, enabling the flow of charge carriers and device switching.
• Diamond’s Activation Energy Problem:
  • Boron (P-type): 0.36 eV activation energy (vs. 0.045 eV in silicon).
  • Phosphorus (N-type): 0.57 eV (even worse!).
  • Result: At room temperature, very few dopant atoms supply charge carriers, so little current flows.
  • [17:40] "With silicon there's plenty of water, meaning charge carriers available... but the threshold for doped diamond is multiples higher."
• Attempts to Overcome:
  • Heavy doping (leads to metallic—not semiconducting—behavior).
  • Delta doping (layers), but ion implantation damages the crystal lattice.
  • Fundamental issue: Can’t create efficient and stable N-type diamond transistors.
• A Serendipitous Discovery:
  • Hydrogen termination (late 1980s): Attaching hydrogen to surface bonds dramatically boosts surface conductivity by “terminating” dangling bonds.
  • Allowed creation of P-type conductive layers; first MESFET demonstrated in 1994.
  • [21:10] "If we terminated the dangling carbon bonds on the diamond surface, we can produce a conductive surface a trillion times higher than expected."
• Limitations:
  • Only enables surface conduction; modern transistors need vertical conduction (through the bulk).
  • Doesn’t fix the N-type doping problem.
  • [24:15] "In the diamond MESFET, power flows only through the thin oxygen or hydrogen terminated layer at the surface... Changing this means fixing the doping problem again."

6. Industry Outlook and Future Possibilities

• Use Today:
  • Diamond is mainly used as a heat sink for other semiconductors rather than an active device material.
• Practical Barriers:
  • Manufacturing scalability, defect control, doping reliability, and cost.
  • Even if the technical issues are solved, market adoption lags (as with silicon carbide: 4-inch wafers in 1999, but commercial devices only in 2011).
  • [27:10] "Superior material properties is just a small part of the equation. It needs to be manufacturable, scalable and economic."
• Long-term Vision:
  • Diamond’s unmatched properties keep it in research pipelines and niche applications; market needs could drive breakthroughs in the future.
  • [28:25] "Nevertheless, the market's growing needs and the material's own inherent allure will continue to drive more research in the years to come."

Notable Quotes & Memorable Moments

• On the science-fiction plot twist:

  • [00:20] “Blofeld should have instead concentrated his criminal organization’s resources on developing diamond transistors. Had he succeeded, he might have reaped far greater profits than with some silly space laser.” — Jon Y

• On silicon’s limitations:

  • [02:35] “With more heat, charge carriers can punch through the gate and generate a small leakage current. This current dissipates more heat, which worsens the leakage and makes even more heat—a negative feedback loop.” — Jon Y

• On diamond’s theoretical supremacy:

  • [07:20] “Perhaps the best thing is how diamond combines all of these properties. High thermal conductivity, high breakdown, voltage and radiation, hardness and high carrier mobility in a single package. No other material is like this.” — Jon Y

• On the challenge of making large diamonds:

  • [12:55] “Single crystal wafers right now just 10 millimeters wide can cost up to 10,000 times of equivalent sized silicon. We are far from ready.” — Jon Y

• On doping frustrations:

  • [17:45] “If we’re to use industry lingo, that room temperature works, but the threshold for doped diamond is multiples higher. Or…I mean, like I said, the correct lingo is to say deeper.” — Jon Y

• On the serendipitous hydrogen-termination discovery:

  • [21:15] “A chance experiment with CVD diamond discovered that if we terminated the dangling carbon bonds on the diamond surface, we can produce a conductive surface a trillion times higher than expected.” — Jon Y

• On the enduring hope for diamond:

  • [28:25] “The market's growing needs and the material's own inherent allure will continue to drive more research in the years to come.” — Jon Y

Timestamps for Important Segments

• [00:02–02:50]: Opening, silicon’s advantages and core limitations.
• [03:00–06:15]: Overview of wide bandgap semiconductor alternatives.
• [06:20–09:45]: Diamond physical properties, why it’s theoretically superior.
• [10:00–16:00]: History and methods for artificial diamond synthesis; why scaling up is an engineering nightmare.
• [16:20–20:15]: Fundamental doping challenges and the physics behind them.
• [20:30–23:50]: Surface conduction workaround and early diamond transistor research.
• [24:00–28:00]: Analogies to current uses, industry context, and concluding thoughts.

Conclusion

Jon Y delivers a compelling narrative explaining why diamond, often imagined as a “holy grail” for semiconductor performance, remains a distant goal in the transistor world despite its amazing physical properties. The episode closes on a pragmatic yet optimistic note: while diamond power electronics remain years (or decades) out, diamond’s unique properties continue to drive ongoing research and future possibilities in high-performance electronics.