Backward Diode

1. Theory

Definition: A backward diode is a tunnel-type PN junction doped just up to the edge of degeneracy, so it conducts more easily in reverse than in forward for small voltages.
Near-degenerate doping: The N-side conduction-band edge and P-side valence-band edge line up almost exactly at zero bias, with no band overlap.
Thin junction: The heavy doping gives a very narrow depletion region, allowing band-to-band tunnelling under even a small reverse bias.
No tunnelling peak: Without band overlap there is no forward tunnelling current peak or negative-resistance region, unlike a tunnel diode.
“Backward” I-V curve: Near the origin the curve looks like a mirrored conventional diode — strong reverse conduction, weak forward conduction until thermal diffusion takes over.

2. Working Principle

Small reverse bias: Filled P-side valence states move opposite empty N-side conduction states, and electrons tunnel across, giving a steep reverse current.
Small forward bias: No states line up for tunnelling, so only a tiny current flows until normal diffusion current starts at larger forward voltage.
Zero-bias curvature: The sharpest bend of the I-V curve sits right at $V = 0$ , so it rectifies small RF signals without any DC bias.
Low zero-bias resistance: Strong tunnelling at the origin gives a low dynamic resistance that matches easily to $50 Ω$ lines.
Temperature stability: Tunnelling depends weakly on temperature, so detector sensitivity varies less than in thermionic (Schottky) detectors.
No stored charge: Conduction is by tunnelling, not minority-carrier diffusion, so response is limited only by the RC time.

3. Applications

Zero-bias microwave and mm-wave detectors: Unbiased, low-noise, temperature-stable square-law power detection.
Passive mm-wave imaging: Heterojunction (InAs/AlSb) backward diodes serve as sensitive detectors in imaging arrays.
Low-level mixers: Mixing very small RF signals with low local-oscillator drive.

Name	Formula	Related Formulae	Variables/Constants
Near-degenerate doping condition	$N_{D} \approx N_{C}, N_{A} \approx N_{V} (just below full degeneracy)$	$contrast: Tunnel Diode page uses N_{D} > N_{C}, N_{A} > N_{V} (fully degenerate)$	$N_{C}, N_{V} : e f f ec t i v e d e n s i t y o f s t a t es (see B a s i c S e mi co n d u c t or E q u a t i o n s p a g e)$
Reverse (and near-zero-bias) tunneling current	$I (V) \approx I_{0}^{'} T (E) (V ≲ 0)$	$T \approx exp (- \frac{4 2 m ^{*} E _{g}^{3/2}}{3 q ℏ E}) (see Zener Diode page)$	$I (V) : c u r r e n t d o mina t e d b y ban d - t o - ban d t u nn e l in g a t an d b e l o w z er o bia s$ $I_{0}^{'} : t u nn e l in g c u r r e n t p r e f a c t or$
Total diode current	$I = I_{T} (V) + I_{t h} (V)$	$(see Tunnel Diode page; no I_{e x c} - d o mina t e d valley or forward peak forms here)$	$I_{T} : t u nn e l in g c u r r e n t co m p o n e n t$ $I_{t h} : or d ina r y t h er ma l d i f f u s i o n c u r r e n t, d o minan t o n l y a t l a r g er f or w a r d bia s$
Zero-bias dynamic resistance	$R_{0} = (\frac{d I}{d V})_{V = 0}^{- 1}$		$R_{0} : s ma l l - s i g na l r es i s t an ce a t z er o bia s, t y p i c a l l y l o w d u e t o s t r o n g t u nn e l in g co n d u c t i o n r i g h t a t t h e or i g in$
Curvature coefficient (detector sensitivity)	$γ = \frac{I ^{''} ( 0 )}{I ^{'} ( 0 )}$	$β_{I} = \frac{γ}{2} = \frac{I ^{''} ( 0 )}{2 I ^{'} ( 0 )}$	$γ : c u r v a t u r e coe f f i c i e n t o f t h e I - V c u r v e a t V = 0$ $β_{I} : c u r r e n t r es p o n s i v i t y (d e t ec t or se n s i t i v i t y)$ $I^{'} (0), I^{''} (0) : f i r s t an d seco n d d er i v a t i v es o f I (V) a t V = 0$
Detected (rectified) current for a small RF signal	$Δ I \approx \frac{1}{4} I^{''} (0) v_{0}^{2}$	$v (t) = v_{0} cos (ω t)$	$Δ I : D C (d e t ec t e d) o u tp u t c u r r e n t$ $v_{0} : R F s i g na l am pl i t u d e$
Switching/recovery time (negligible, tunneling-dominated)	$t_{s} \approx few \times R_{0} C_{j}$	$t_{s}^{T u nn e l} \approx \frac{C _{j} ( V _{v} - V _{p} )}{I _{p}} (see Tunnel Diode page)$ $t_{r r}^{P N / P I N} \approx τ ln (1 + \frac{I _{F}}{I _{R}}) (see PN Junction Diode / PIN Diode pages)$	$t_{s} : R C - l imi t e d r es p o n se t im e; n o t a s p ec t y p i c a l l y q u o t e d f or ba c k w a r d d i o d es, w h ose main a ppl i c a t i o n i s R F d e t ec t i o n r a t h er t han d i g i t a l s w i t c hin g$

Important Questions

Ques: What actually distinguishes a backward diode from a tunnel diode, given they’re both degenerately-doped devices built from the same junction physics?

A tunnel diode is doped well past full degeneracy on both sides, so there is a wide band of energies where filled n-side conduction-band states align with empty p-side valence-band states under forward bias, producing the pronounced tunneling current peak and negative-resistance region described on the Tunnel Diode page. A backward diode is doped only up to the degeneracy threshold, so that the conduction-band edge on the n side and the valence-band edge on the p side are nearly aligned at zero bias (supporting strong tunneling current under small reverse bias), but not enough band overlap to ever produce a true forward tunneling peak. The practical result is a diode whose I-V curve looks roughly like a mirror image of a conventional diode near the origin: it conducts far more readily in reverse than in forward for small voltages, which is the origin of the name “backward” diode. Reference: S.M. Sze and Kwok K. Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007.

Ques: Why is the curvature coefficient $γ$ so central to the backward diode’s role as a zero-bias microwave detector?

Square-law detection relies on the nonlinearity (curvature) of a diode’s I-V curve: a small RF signal centered at a given bias point produces a rectified DC output proportional to $I^{''}$ at that point. Conventional PN or Schottky diodes need forward bias of several tenths of a volt before their exponential I-V curve develops useful curvature, which means they require a bias current (and the shot noise that comes with it) to function as sensitive detectors. A backward diode’s curve already has strong curvature exactly at $V = 0$ , because the onset of tunneling conduction happens right at the origin rather than requiring forward turn-on — allowing it to detect very small RF signals with no DC bias at all, reducing detector noise and simplifying circuit design. This is the historical reason backward diodes were prized for low-level microwave and millimeter-wave power detection, before zero-bias Schottky detector diodes became the more common choice. Reference: S.M. Sze and Kwok K. Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007.

Ques: Does a backward diode have a meaningful reverse recovery time?

Not in the PN/PIN sense, and for the same reason as the Tunnel Diode: conduction near the operating region is tunneling-dominated rather than diffusion-dominated, so there is no significant minority-carrier charge stored to discharge on switching. What little switching response time exists is simply the $R_{0} C_{j}$ charging time of the junction capacitance through the (already low) zero-bias dynamic resistance. In practice this isn’t a parameter backward diode datasheets typically emphasize, since the device’s role is almost always as a passive RF/microwave detector operating continuously at small signal levels, not as a digital switching element — unlike the tunnel diode, which was historically used specifically for its fast, charge-limited switching behavior. Reference: S.M. Sze and Kwok K. Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007.

Ques: Why does a low zero-bias dynamic resistance $R_{0}$ matter for detector performance?

Microwave transmission lines and antennas typically present a standardized, low characteristic impedance (commonly $50 Ω$ ), and efficient power transfer into a detector diode requires the diode’s own small-signal resistance to be reasonably well matched to that impedance. A backward diode’s strong tunneling conduction right at the origin naturally gives it a low $R_{0}$ without any external bias network needed to pull it into a favorable operating point — unlike a conventional diode, which would need forward bias current (and an associated bias network) just to reach a comparably low dynamic resistance. This natural, unbiased impedance match is part of why backward diodes integrate simply into passive detector circuits. Reference: S.M. Sze and Kwok K. Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007.

Quartz 5

Explorer

Backward Diode

1. Theory

2. Working Principle

3. Applications

Important Questions

Ques: What actually distinguishes a backward diode from a tunnel diode, given they’re both degenerately-doped devices built from the same junction physics?

Ques: Why is the curvature coefficient $γ$ so central to the backward diode’s role as a zero-bias microwave detector?

Ques: Does a backward diode have a meaningful reverse recovery time?

Ques: Why does a low zero-bias dynamic resistance $R_{0}$ matter for detector performance?

Graph View

Table of Contents

Backlinks

Quartz 5

Explorer

Backward Diode

1. Theory

2. Working Principle

3. Applications

Important Questions

Ques: What actually distinguishes a backward diode from a tunnel diode, given they’re both degenerately-doped devices built from the same junction physics?

Ques: Why is the curvature coefficient γ so central to the backward diode’s role as a zero-bias microwave detector?

Ques: Does a backward diode have a meaningful reverse recovery time?

Ques: Why does a low zero-bias dynamic resistance R0​ matter for detector performance?

Graph View

Table of Contents

Backlinks

Ques: Why is the curvature coefficient $γ$ so central to the backward diode’s role as a zero-bias microwave detector?

Ques: Why does a low zero-bias dynamic resistance $R_{0}$ matter for detector performance?