GaN Power Diode

1. Theory

Definition: A GaN power diode is a high-voltage rectifier made of gallium nitride, built either as a Schottky diode or as a PN diode.
Wide bandgap: GaN’s bandgap ( $\approx 3.4$ eV) gives a critical field of about 3.3 MV/cm, roughly ten times that of silicon.
Drift region: A lightly doped layer supports the blocking voltage; the higher critical field allows a thinner, more heavily doped drift layer for the same breakdown voltage.
Baliga figure of merit: Because on-resistance scales with $1/ E_{cr i t}^{3}$ , GaN’s figure of merit is two to three orders of magnitude above Si’s.
Device structures: Vertical GaN-on-GaN diodes follow the classic drift-region picture; lateral GaN-on-Si diodes conduct through the AlGaN/GaN 2DEG used in HEMTs.

2. Working Principle

Forward conduction (Schottky type): Majority electrons cross the metal-GaN barrier by thermionic emission, with no minority-carrier storage.
Forward conduction (PN type): Bipolar injection gives conductivity modulation but a higher turn-on voltage ( $\approx 3$ V) because of the wide bandgap.
Reverse blocking: The depletion region spreads through the drift layer (vertical) or along the 2DEG with field plates shaping the surface field (lateral).
Fast switching: Schottky-type GaN diodes have almost no reverse recovery charge, so switching loss per cycle is very low.
Breakdown: Vertical GaN PN diodes can reach true avalanche breakdown; lateral devices are usually limited by surface or buffer breakdown instead.

3. Applications

Monolithic GaN power ICs: Lateral diodes integrate on the same chip as GaN HEMT switches in compact converters.
High-frequency power conversion: Low-loss rectification in MHz-range switching converters and wireless power receivers.
High-voltage vertical rectifiers: GaN-on-GaN PN diodes offer avalanche-rugged blocking above 1 kV (still an emerging technology).

Name	Formula	Related Formulae	Variables/Constants
Breakdown voltage (one-sided drift region)	$B V \approx \frac{ε _{s} E _{cr i t}^{2}}{2 q N _{D}}$	$(see PN Junction Diode / Avalanche Diode pages)$	$B V : b r e ak d o w n v o l t a g e$ $E_{cr i t} : cr i t i c a l f i e l d (\approx 3.3 MV/cm f or G a N, v s \approx 0.3 MV/cm f or S i)$ $N_{D} : d r i f t r e g i o n d o p in g co n ce n t r a t i o n$
Minimum drift region thickness	$W_{d r i f t} \geq \frac{2 ε _{s} B V}{q N _{D}}$	$(see PN Junction Diode page, general W (V) form)$	$W_{d r i f t} : minim u m d r i f t l a y er t hi c k n ess n ee d e d t o s u pp or t B V w i t h o u t t h e d e pl e t i o n r e g i o n p u n c hin g t h r o ug h$
Specific on-resistance (drift region)	$R_{o n, s p} = \frac{W _{d r i f t}}{q μ _{n} N _{D}}$		$R_{o n, s p} : a r e a - n or ma l i z e d d r i f t r e g i o n r es i s t an ce$ $μ_{n} : e l ec t r o n m o bi l i t y in t h e d r i f t r e g i o n$
Unipolar specific on-resistance limit (Baliga’s limit)	$R_{o n, s p} = \frac{4 B V ^{2}}{ε _{s} μ _{n} E _{cr i t}^{3}}$	$derived by eliminating W_{d r i f t}, N_{D} between the two rows above$	$(fundamental tradeoff curve between breakdown voltage and on-resistance for a given material)$
Baliga figure of merit (material)	$B F O M = ε_{s} μ_{n} E_{cr i t}^{3}$	$R_{o n, s p} = \frac{4 B V ^{2}}{B F O M}$	$B F O M : ma t er ia l f i g u r e o f m er i t f or u ni p o l a r p o w er d e v i ces; G a N^{'} s B F O M i s r o ug h l y tw o t o t h r ee or d er s o f ma g ni t u d e hi g h er t han S i^{'} s$
Thermionic emission current (Schottky-type GaN power diode)	$I = A A^{*} T^{2} exp (- \frac{q ϕ _{B}}{k _{B} T}) [exp (\frac{V}{n V _{T}}) - 1]$	$(see Schottky Diode / Metal-Semiconductor Contacts pages)$	$(same thermionic emission physics, evaluated with GaN’s own A^{*}, ϕ_{B}, and ε_{s})$ $k_{B} : B o l t z mann co n s t an t$

Important Questions

Ques: Why does GaN achieve such a dramatically better breakdown-voltage-vs-on-resistance tradeoff than silicon?

The unipolar specific on-resistance limit scales as $R_{o n, s p} \propto B V^{2} / E_{cr i t}^{3}$ , so the critical field’s cube dominates the comparison between materials far more than any other parameter. GaN’s critical field is roughly an order of magnitude higher than silicon’s, and since that factor enters cubed, the Baliga figure of merit for GaN comes out roughly two to three orders of magnitude higher than silicon’s — meaning a GaN device can be designed for the same breakdown voltage as an equivalent silicon device with dramatically lower on-resistance (or conversely, achieve far higher breakdown voltage for the same on-resistance), which is the entire economic case for wide-bandgap power devices. Reference: B.J. Baliga, “Semiconductors for high-voltage, vertical channel field-effect transistors,” J. Appl. Phys. 53, 1759 (1982); S.M. Sze and Kwok K. Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007.

Ques: Why do many practical GaN power diodes use a Schottky barrier rather than a PN junction, even though GaN PN junctions are entirely possible?

A Schottky barrier is a majority-carrier (unipolar) device, which — as established on the Schottky Diode page — has no significant minority-carrier storage and therefore negligible reverse recovery time, unlike a PN junction’s bipolar conduction. In power switching applications, reverse recovery charge directly translates into switching energy loss every single cycle, so for a device meant to switch at high frequency with minimal loss, the Schottky structure’s fast, storage-free turn-off is often worth more than the lower forward voltage drop a PN junction might otherwise offer, which is why Schottky (and Schottky-like) structures dominate commercial GaN power diode offerings. Reference: S.M. Sze and Kwok K. Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007.

Ques: How does a real, lateral GaN-on-Si power diode (built on a HEMT-like epitaxial structure) differ from the vertical drift-region picture this page’s formulas assume?

Many commercially available GaN power diodes are not built as classical vertical devices with a uniformly doped drift region at all — they are lateral structures grown on the same AlGaN/GaN epitaxial stack used for GaN HEMTs, conducting current through the polarization-induced 2DEG described on the GaN HEMT page rather than through a doped semiconductor drift layer. These devices trade the vertical drift-region tradeoff curve derived in this table for a different one, governed instead by 2DEG sheet resistance and lateral electric field management (often using field plates to shape the surface field and push breakdown voltage higher without a thicker drift region). This is why the simple unipolar figure-of-merit curve here is best understood as the fundamental physical limit for a vertical, uniformly doped device, while real commercial lateral GaN power diode datasheets reflect a related but structurally distinct set of engineering tradeoffs. Reference: O. Ambacher et al., J. Appl. Phys. 85, 3222 (1999) (see GaN HEMT page); S.M. Sze and Kwok K. Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007.

Quartz 5

Explorer