In quantum well photodetectors, the electric field across a reverse-biased junction can make the active region less effective at absorbing light. This effect, known as the Quantum-Confined Stark Effect (QCSE), changes the energy structure of a quantum well and can shift electron and hole wave functions apart. A study of GaInNAs/GaAs structures explores how composition gradients and step-like wells could counter that separation, with the design matched to the direction of the field in the junction.
Why the Quantum-Confined Stark Effect matters
Photodetector performance depends in part on how effectively the active region absorbs incident radiation. Quantum wells can provide useful optical and electronic properties, but their active thickness is limited. In a p-n, p-i-n or Schottky junction, the built-in potential and applied reverse bias produce an electric field across the active region. That field changes the potential profile within a quantum well.
The resulting change in the electron and hole wave functions is central to the QCSE. If the carriers’ wave functions become spatially separated, their overlap decreases. Because the probability of an electron-hole transition is proportional to the square of this overlap, reduced overlap can lower the well’s absorption efficiency. The effect depends not only on the presence of an electric field, but also on the structure of the well and the field’s direction.
Using GaInNAs composition to shape carrier states
GaInNAs is a multicomponent semiconductor alloy containing gallium, indium, nitrogen and arsenic. Its growth can produce nonuniform distributions of indium and nitrogen, in part because alloy composition and growth processes can lead to segregation. Such inhomogeneity can introduce structural disorder and affect the potential landscape and carrier confinement within a quantum well.
Rather than treating composition variation only as an imperfection, the analysis considers whether a deliberately shaped distribution could influence carrier wave functions in a useful way. The researchers examined measured GaInNAs/GaAs structures and also modeled alternative quantum-well designs to investigate their response to electric fields.
For the measured structure, three GaInNAs wells were grown by atmospheric-pressure metal organic vapour phase epitaxy. High-resolution X-ray diffraction and secondary ion mass spectrometry were used to characterize the structure and composition. The SIMS profiles showed that the indium and nitrogen signals peaked at different depths, with their maxima separated by 38 seconds of sputtering time. Using these measurements alongside diffraction analysis, the authors estimated a 12 nm well with indium content varying from 3 to 11 percent and nitrogen from 0.6 to 0 percent, with composition gradients toward the surface.
This characterization establishes that composition can vary through the growth direction. The subsequent comparison of well designs was a numerical analysis, not a demonstration that a finished photodetector had achieved a particular performance improvement.
Comparing graded and step-like quantum wells
The modeled structures used wells 10 nm thick, with average indium and nitrogen contents of 15 and 2 percent, respectively. Four cases were compared:
- Uniform well: indium and nitrogen were distributed uniformly through the layer.
- Convergent grading: indium increased from 10 to 20 percent while nitrogen increased from 1 to 3 percent.
- Opposed grading: indium increased from 10 to 20 percent while nitrogen decreased from 3 to 1 percent.
- Step-like well: a layered GaInAs/GaNAs/GaAs arrangement created abrupt changes in composition.
For each design, electron and hole states were calculated without an electric field and with fields applied in opposite directions. The assumed field was 80 kV/cm, and the calculations considered ground-state transitions. The opposite field directions represent different junction orientations, such as n-i-p and p-i-n structures.
In the uniform well, the squared overlap of the electron and hole wave functions fell from approximately 0.999 without a field to 0.279 under either field direction. This illustrates how strongly an applied field can reduce overlap in a symmetric design. The graded structures behaved differently depending on the field direction: their overlap could be higher in one orientation and lower in the other.
Matching well design to junction orientation
The opposed-gradient and step-like designs illustrate why the junction’s field direction matters. In these structures, the composition profile shifts the electron and hole states toward different sides of the well even without an applied field. A field in one direction can increase that separation, while a field in the opposite direction can reduce it.
For the opposed-gradient well, the calculated squared overlap was 0.3976 with no field, 0.0858 under the n-i-p field direction and 0.6438 under the p-i-n direction. For the step-like well, the corresponding values were 0.3047, 0.5765 and 0.1329. These results show that neither design is universally beneficial: the direction that increases overlap for one structure reduces it for the other.
The practical design implication is to consider the quantum-well profile and junction polarity together. A composition gradient or step-like arrangement may compensate for field-induced separation when its internal potential profile is oriented appropriately. The modeled results point to the opposed-gradient well for a p-i-n field direction and the step-like well for an n-i-p field direction.
Implications for photodetector development
For engineers and device developers, the work highlights composition control as a potential design variable in dilute-nitride quantum wells. It also underscores the importance of combining material characterization with electronic-structure calculations. In a multicomponent alloy, assuming a uniform layer can miss composition variations that affect carrier confinement and optical transitions.
The findings are based on structural measurements and theoretical calculations. They identify design approaches for compensating the adverse effect of QCSE on absorption, but do not establish the performance of a fabricated detector. Further device-level validation would be needed to determine how a particular well design performs in an operating photodetector.
Conclusion
Reverse-bias electric fields can reduce absorption in quantum well photodetectors by separating electron and hole wave functions. In GaInNAs/GaAs structures, graded composition and step-like well profiles can reshape those states, but their effect depends on field direction. For practical design, the key is to treat alloy profile, junction orientation and operating electric field as connected choices. This offers a considered route for exploring QCSE compensation in quantum-well-based photodetectors.