Access Official Technical PDF

Share Summary:LinkedInEmail

AlGaN solar-blind photodiodes are designed to detect ultraviolet radiation below approximately 280 nm, a region in which atmospheric ozone absorbs nearly all incoming solar radiation. This naturally low background can support applications such as flame monitoring, corona detection, UV environmental monitoring, and secure non-line-of-sight communication. A back-illuminated metal-insulator-semiconductor (MIS) design offers a way to address a key device-engineering challenge: producing effective high-aluminum-content p-type AlGaN for conventional p-i-n photodiodes.

Why consider an MIS structure for solar-blind UV detection?

AlGaN is suited to ultraviolet detection because its bandgap can be adjusted through alloy composition. Higher aluminum content shifts the material’s absorption toward shorter wavelengths, including the solar-blind region. The alloy also has a sharp absorption cutoff and chemical and thermal stability.

Back-illuminated p-i-n photodiodes are attractive for their low dark current and fast response, and their configuration can support integration with silicon readout integrated circuits using flip-chip techniques. However, high-aluminum-content p-AlGaN is difficult to dope effectively: magnesium acceptors have high activation energy, making it challenging to obtain high hole concentration and low resistivity. Lattice and thermal-expansion mismatch can also contribute to cracks in epitaxial layers, while forming a good ohmic contact to p-AlGaN can hinder carrier transport.

An MIS structure replaces the high-aluminum p-type layer with a high-work-function metal electrode and an insulating layer at the metal-semiconductor interface. This avoids dependence on p-AlGaN while retaining a back-illuminated device architecture. The approach is relevant to solar-blind photodetectors, where adding a p-GaN layer to ease transport can introduce near-ultraviolet response that is undesirable for strict solar-blind operation.

AlGaN solar-blind photodiodes: the role of layer design

The reported device was grown on a c-plane sapphire substrate using metal-organic chemical vapor deposition. Its epitaxial stack included an AlN buffer, AlGaN/AlN superlattices, and an undoped AlGaN layer before the MIS structure. Two features within the design were intended to improve the material and carrier behavior.

A graded n-AlGaN transition layer

A thin n-AlGaN layer with aluminum content decreasing from 54% to 45% was placed between the n-AlGaN layer and the undoped AlGaN absorption region. X-ray reciprocal-space mapping indicated that the graded layer helped regulate strain in the subsequent epitaxial layers. The resulting absorption layers were reported as fully relaxed, addressing strain associated with the differences between the substrate and the AlGaN epilayers.

An n-AlGaN interlayer in the absorption region

A homogeneous n-AlGaN interlayer was placed between undoped AlGaN layers in the light-absorption region. The purpose was to redistribute the electric field within that region. A stronger field can assist in separating photo-generated electrons and holes and transporting them toward the contacts, reducing the opportunity for recombination and supporting photocurrent extraction.

Device simulations compared the electric-field profiles with and without this interlayer. The structure containing it showed a higher field in the absorption region both at zero bias and under reverse bias. This illustrates how layer placement and doping can be used to shape the internal field rather than relying only on the basic Schottky junction field.

Reported photodiode response

The device showed a bandpass spectral response from approximately 255 to 280 nm, within the intended solar-blind range. At zero bias, its peak responsivity was 0.115 A/W at 270 nm, corresponding to an external quantum efficiency (EQE) of 53%. The ultraviolet-to-visible rejection ratio exceeded three orders of magnitude. Under a reverse bias of 3 V, peak responsivity rose to 0.154 A/W, corresponding to an EQE of 70.6%.

Transient response measurements used a 266 nm pulsed laser. The reported 10–90% rise time was around 1.4 ns, while the measured 90–10% decay time was about 55 microseconds. Fitting the decay with a first-order exponential function gave a decay time of approximately 24 microseconds. These figures describe different ways of characterizing the decay response and should not be treated as interchangeable.

The device’s dark current density was about 2.4 × 10-7 A/mm2 under 3 V reverse bias. Taken together, the measurements demonstrate solar-blind spectral selectivity alongside substantial responsivity at zero and reverse bias. They do not remove the need to evaluate operating conditions and response requirements for a particular detection system.

Practical implications for UV detector design

For engineers evaluating ultraviolet detectors, the MIS approach highlights several linked design considerations:

  • Contact and doping constraints: Replacing high-aluminum p-AlGaN can avoid a difficult p-type doping and contact problem.
  • Strain management: A composition-graded n-AlGaN layer can help manage strain in the epitaxial absorption stack.
  • Carrier collection: An n-type interlayer can alter the absorption-region electric field to favor carrier separation and transport.
  • Spectral selectivity: The window and absorption layers together determine the short- and long-wavelength cutoffs, so the full stack matters for solar-blind operation.

Back illumination and MIS construction are not just packaging choices; they are part of the device architecture used to balance materials limitations, electric-field distribution, and spectral response. The reported results show how a graded transition layer and an absorption-region interlayer can work together to address these concerns. For industrial and technical applications, the practical takeaway is to assess the detector’s wavelength range, bias conditions, dark current, and transient response against the requirements of the sensing or monitoring system.

Access Official Technical PDF

About The Author