Abstract
We used a semiclassical model to describe carrier capture into and thermionic escape from GaInNAs/GaAs multiple quantum wells (MQWs) situated within the intrinsic region of a GaAs pin junction. The results are used to explain photocurrent oscillations with applied bias observed in these structures, in terms of charge accumulation and resonance tunnelling.
Keywords:
GaInNAs/GaAs; capture rates; resonant tunnelling; pin multiple quantum wellsBackground
Over the last couple of decades, IIIV compounds containing small quantities of nitrogen (dilute nitrides) have received much attention, both experimentally and theoretically. A number of books and review articles as well as a large number of papers in the field have been published [13]. The interest in this material system started with the discovery of a large bowing parameter upon the addition of small amounts of nitrogen into Ga(In)As. The band gap energy is reduced with increasing nitrogen composition [4]. As a result, it has become possible to fabricate dilute nitridebased lasers, optical amplifiers and photodetectors operating in the 1.3 and 1.55 μm windows of optical communication systems [57] and solar cells in multijunction devices with increased efficiency [8,9].
In the early days of lowdimensional semiconductors, carrier capture into quantum wells of the IIIV compounds was studied with considerable interest aimed at improving the performance of quantum well (QW) lasers [10]. First theoretical calculations of the carrier capture rates were performed by Shichijo [11] and Tang [12]. The mechanism was regarded as a classical process where the carrier capture rate is limited by the optical phonon scattering and the mean free path. Another calculation, presented by Burn and Bastard [13], discovered strong oscillations in electron capture rates as a function of the well width. Babiker and Ridley [14] studied the electron capture rates in GaAs QWs by taking into account the quantum mechanical aspect of the capture process with strong resonances. It has been shown that capture rates strongly depend on structural parameters such as QW and barrier widths, number of wells and the mean free path of the carriers as limited by scattering processes [13,14]. The reason for the choice of dilute nitride quantum wells is because in this study, we aimed at developing a photodetector with a cutoff wavelength of around 1.3 μm that can be lattice matched to GaAs. Therefore, a resonant cavityenhanced photodetector by using GaAs/GaAlAs distributed Bragg reflectors to operate at the 1.3μm communications window would be possible. Obviously, the main disadvantage of dilute nitrides compared to the InPbased material is the poor optical quality in devices with high nitrogen composition. This could be partly overcome by rapid thermal annealing at the expense of blue shifting of the operation wavelength.
In this paper, we present the theoretical analysis of the carrier capture and escape time in a Ga_{0.96}In_{0.04} N_{0.015}As_{0.985}/GaAs multiple quantum wells (MQWs) situated within the builtin field of a GaAs pin structure. Experimentally observed photocurrent oscillations in these structures [15,16], explained in terms of charge accumulation and field domain formation, are shown to be in accord with our theoretical results.
Methods
Capture time and thermionic emission
The semiclassical model used in our analysis provides useful physical insight into carrier transport across and carrier capture into the MQWs. We show that the disparity between the electron and hole capture and reemission times from the quantum wells leads to the accumulation of electrons within the quantum wells. In our samples, the selected In and N concentrations (Ga_{0.96} In_{0.04} N_{0.015} As_{0.985}) in the quantum wells ensure good lattice matching to the GaAs barriers and the substrate [10]. This allows the growth of thicker and highquality layers and making the device suitable for photovoltaic applications where efficient absorption plays a fundamental rule [17].
In the quantum wells with the given composition, electrons are more strongly confined in the QWs (conduction band offset approximately 250 meV), than in the holes (valence band offset approximately 20 meV). The longitudinal optical (LO) phonon energy is ħω_{LO} = 38 meV [16], which is higher than the binding energy of the holes in the QW. Therefore, the holes photogenerated at the GaAs will be captured by the QW via the emission of acoustic phonons. The capture of electrons, however, will involve inelastic scattering with LO phonons which will be very fast compared to the hole capture time and assumed, in our calculations, to be negligible compared to the hole capture rates [18].
Under collisionfree hole transport conditions, we use the following Bethe relation [19,20] to estimate the thermionic capture time for holes reaching the top of the potential barrier Φ (process 1 in Figure 1).
Figure 1. Mechanisms involved in hole capture dynamics into QW.
In this expression, L_{b} is the barrier width, is the heavy hole effective mass, e is the electronic charge, k_{B} is the Boltzman constant, and T is the temperature. The term E_{h} is the kinetic energy of the hole traversing the QW and can be expressed as [20,21]
Here, E_{excess} is the laser excess energy, V_{h} is the depth of the QW in the valence band, and is the electron effective mass in the QW. Since the optical excitation energy above the QW band gap, the laser excess energy term is negligible.
Once the holes have reached the potential barrier edge, they can either traverse the quantum well under the influence of the builtin electric field in the pn junction or be captured into the QW by inelastic scattering with acoustic phonons [22]. These processes are depicted in Figure 1 as processes 2 and 3, respectively. With the hole mean free path l, smaller than the QW width L_{w}, the hole capture time τ_{capture} is reduced by the probability 1  exp(L_{w}/l). The overall capture time of the hole for the GaInNAs/GaAs QW is then equal to:
In the event of not being trapped, the time for holes to traverse the QW is as follows:
Once the hole is captured into the well, it can escape from it via thermionic emission. The thermal escape time τ_{th} from the QW will be determined principally by the height of the barrier discontinuity and can be written as [23]
Where m^{*} is the hole effective mass in the well.
Results and discussion
Using the equations above together with the band anticrossing model [24] and the various material parameters as reported in the literature [3], the analysis of hole τ_{capture} and τ_{cross} has been carried out for the pin GaInNAs/GaAs structure. The results are plotted in Figure 2 as a function of QW width.
Figure 2. The QW width dependence of the hole τ_{capture }(squares) and τ_{cross }(stars) calculated at room temperature.
τ_{capture} decreases exponentially with the QW width, as expected from Equation 3, where as τ_{cross} increases linearly. It is clear that the hole is more likely to traverse the quantum well than to be captured into the QW. In fact, the hole capture time is in the range of 4 to 13 ps, much longer than the 0.1 to 0.4 fs time needed to cross the QW. Thus, we assumed that at low temperatures, the last term [exp (eΦ/k_{B}T)] in Equation 1 would be negligible.
In the current work, however, we took into account the effect of temperature and, therefore, we included this term in our calculation. The temperature dependence of τ_{capture} and τ_{cross} are plotted in Figure 3 for a 10nmthick quantum well.
Figure 3. Temperature dependence of the hole τ_{capture }(squares) and τ_{cross }(stars) calculated for a 10nmthick QW.
The thermal escape time for both electrons and holes are also calculated as a function of temperature, using Equation 5 and plotted in Figure 4. It is clear that the hole escape time is very short, around 0.2 ps at room temperature, due to the small valence band offset. This value is two orders of magnitude shorter than the thermal escape time for electrons (approximately 60 ps). As the temperature decreases, the thermal escape time of electrons rapidly increases while for holes, the time is less than 1 ns up to temperature of T = 30 K, due to a lack of phonons to excite the holes over the potential barrier.
Figure 4. Theoretical thermal escape times for electrons and holes in the 10nmthick QW, as function of temperature.
When the sample is under illumination with photons with energies smaller than the barrier band gap but greater than the quantum wells band gaps, photogenerated electrons will remain in the wells longer than the photogenerated holes. Therefore, accumulation of negative charge in the wells will occur. If the quantum wells are in an electric field, as they are within the builtin filed of GaAs pin structures, in our samples, thermally escaped holes will rapidly be swept away giving rise to a fast component of photocurrent. The accumulated negative charge will contribute to photocurrent via both thermionic emission and resonant tunnelling [25], giving rise to the wellknown photocurrent oscillations as a function of applied voltage as shown in Figure 5, the details of which have already been reported by us elsewhere [26,27].
Figure 5. IV results in dark and light condition, together with the derivative curves.
In Figure 5, the current is plotted against applied voltage for both in darkness and when the sample was illuminated with photons with energies greater than the quantum well band gap.
The photocurrent in Figure 5 has two components; the thermionic current which increases monotonically with applied bias and the oscillatory component which is the resonant tunnelling current [26]. In order to show clearly the oscillatory component, we took the first derivative of the photocurrent. The peak current values correspond to the resonant conditions in the wells adjacent to the anode similar to those as described in references [26,28].
Conclusions
The aim of the work was to explain the photocurrent oscillations as a function of applied voltage that we observed in our earlier studies in GaInNAs/GaAs quantum wells placed in the intrinsic region of a GaAs pin structure. We have shown that hole thermal escape time of photogenerated holes within the quantum wells is very short compared to that of the electrons; therefore, the accumulation of negative charge in the QW may occur and give rise to the photocurrent via thermionic emission and resonant tunnelling. The resonant tunnelling component has an oscillatory behaviour with strong resonances.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
HMK carried out the theoretical works, analysed the data and wrote the paper; NB supervised the project. Both authors read and approved the final manuscript.
Acknowledgements
We would like to thank COST action MP0805 entitled ‘Novel Gain Materials and Devices Based on IIIVN Compounds’ and EPSRC grant EP/P503965/01 for funding.
References

Potter RJ, Balkan N: Optical properties of GaInNAs and GaNAs QWs.
J Phys Condens Matter 2004, 16:33873412. Publisher Full Text

Henini M: Dilute Nitride Semiconductors. Amsterdam: Elsevier Science; 2005.

Erol A: Dilute IIIV Nitride Semiconductor and Material Systems. Berlin: Springer Series; 2008.

Kondow M, Uomi K, Niwa A, Kitatani T, Watahiki S, Yazawa Y: A novel material for long wavelength laser diodes with excellent high temperature performance.
Jpn J Appl Phys 1996, 35:12731275. Publisher Full Text

Jewell J, Graham L, Crom M, Maranowski K, Smith J, Fanning T, Schnoes M: Commercial GaInNAs VCSELs grown by MBE.
Phys Stat Sol 2008, 5:29512956. Publisher Full Text

Jaschke G, Averbeck R, Geelhaar L, Riechert H: Low threshold InGaAsN/GaAs lasers beyond 1500 nm.
J Cryst Growth 2005, 278:224228. Publisher Full Text

Laurand N, Calvez S, Dawson MD, Jouhti T, Konttinen J, Pessa M: 1.3μm continuouslytunable fibercoupled GaInNAs VCSEL.

Hartmann F, Langer F, Bisping D, Musterer A, Höfling S, Kamp M, Forchel A, Worschech L: GaAs/AlGaAs resonant tunnelling diodes with a GaInNAs absorption layer for telecommunication light sensing.
Appl Phys Lett 2012, 100:172113172115. Publisher Full Text

Courel M, Rimada JC, Hernández L: GaAs/GaInNAs quantum well and superlattice solar cell.
Appl Phys Lett 2012, 100:073508073511. Publisher Full Text

Nagarajan R, Fukushima T, Corzine SW, Bowers JE: Effects of carrier transport on highspeed quantum well lasers.
Appl Phys Lett 1991, 59:18351837. Publisher Full Text

Shichijo H, Kolbas RM, Holonyak N, Coleman JJ, Dapkus PD: Calculations in strained quantum wells.
Sol Stat Comm 1978, 27:10291032. Publisher Full Text

Tang JY, Hess K, Holonyak N, Coleman JJ, Dapkus PD: The dynamics of electron hole collection in quantum well heterostructures.
J Appl Phys 1982, 53:60436046. Publisher Full Text

Brum JA, Bastard G: Resonant carrier capture by semiconductor quantum wells.
Phys Rev B 1986, 33:14201423. Publisher Full Text

Babiker M, Ridley BK: Effectivemass eigenfunctions in superlattices and their role in wellcapture.
Superlatt Microstruct 1986, 2:287293. Publisher Full Text

Khalil HM, Mazzucato S, Ardali S, Celik O, Mutlu S, Royall B, Tiras E, Balkan N, Puustinen J, Korpijärvi VM, Guina M: Temperature and magnetic field effect on oscillations observed in GaInNAs/GaAs multiple quantum wells structures.
Mat Sci Engin B 2012, 177:729733. Publisher Full Text

Khalil HM, Mazzucato S, Royall B, Balkan N, Puustinen J, Korpijärvi VM, Guina M: Photocurrent oscillations in GaInNAs/GaAs multiquantum well pin structures.

Van de Walle CG: Band lineups and deformation potentials in the modelsolid theory.
Phys Rev B 1989, 39:18711883. Publisher Full Text

Gupta R, Ridley BK: Elastic scattering of phonons and interface polaritons in semiconductor heterostructures.
Phys Rev B 1993, 48:1197211978. Publisher Full Text

Sze SM: Physics of Semiconductor Devices. 2nd edition. New York: J. Wiley; 1981.

Samuel EP, Talele K, Zope U, Patil DS: Semiclassical analysis of hole capture in Gallium Nitride quantum wells.

Mosko M, Kalna K: Carrier capture into a GaAs quantum well with a separate confinement region.
Semicond Sci Technol 1999, 14:790796. Publisher Full Text

Khalil HM, Mazzucato S, Balkan N: Hole capture and escape times in pin GaInNAs/GaAs MQW structures.

Fox M, Miller DAB, Livescu G, Cunningham JE, Jan WY: Quantum well carrier sweep out: relation to electroabsorption and exciton saturation.
IEEE J Quantum Electron 1991, 27:22812295. Publisher Full Text

Shan W, Walukiewicz W, Ager JW, Haller EE, Geisz JF, Friedman DJ, Olson JM, Kurtz SR: Band anticrossingin GaInNAs alloys.
Phys Rev Lett 1999, 82:12211224. Publisher Full Text

Grahn HT, Balkan N, Ridley BK, Vickers AJ: Negative Differential Resistance and Instabilities in 2D Semiconductors. New York: NATO ASI Series; 1993:189202.

Royall B, Balkan N, Mazzucato S, Khalil HM, Hugues M, Roberts JS: Comparative study of GaAs and GaInNAs/GaAs multiquantum well solar cells.
Phys Stat Sol B 2011, 248(5):11911194. Publisher Full Text

Khalil HM, Royall B, Mazzucato S, Balkan N: Photoconductivity and photoluminescence under bias in GaInNAs/GaAs MQW pin structures.
Nanoscale Res Lett 2012, 7:539542. PubMed Abstract  BioMed Central Full Text  PubMed Central Full Text

Laikhtman B: Current–voltage instabilities in superlattices.
Phys Rev B 1991, 44:1126011265. Publisher Full Text