## Abstract

The spin diffusion and drift at different excitation wavelengths and different temperatures have been studied in undoped InGaAs/AlGaAs multiple quantum well (MQW). The spin polarization was created by optical spin orientation using circularly polarized light, and the reciprocal spin Hall effect was employed to measure the spin polarization current. We measured the ratio of the spin diffusion coefficient to the mobility of spin-polarized carriers. From the wavelength dependence of the ratio, we found that the spin diffusion and drift of holes became as important as electrons in this undoped MQW, and the ratio for light holes was much smaller than that for heavy holes at room temperature. From the temperature dependence of the ratio, the correction factors for the common Einstein relationship for spin-polarized electrons and heavy holes were firstly obtained to be 93 and 286, respectively.

## Background

Much attention has been given to semiconductor spintronics for the promising applications in information technology [1, 2]. One of the fundamental issues on semiconductor spintronics is the spin transport and its manipulation, including spin-related diffusion and drift. The anomalous circular photogalvanic effect (ACPGE) [3–5] and anomalous Hall effect (AHE) [6–10], which are derived from the same spin-orbit coupling (SOC) mechanisms (intrinsic or extrinsic) based on the reciprocal spin Hall effect (RSHE), open avenues to the study of the relationship between the diffusion and the drift of photoinduced spin-polarized electrons. According to [11], the ratio of the diffusion coefficient to the mobility of the photoinduced spin-polarized electrons has been measured to be 0.08 V in AlGaN/GaN heterostructure with the excitation wavelength of 1064 nm at room temperature. In this work, we focused on the spectrum and temperature dependence of the transport properties corresponding to interband transitions in an undoped InGaAs/AlGaAs multiple quantum well (MQW) in which a strong Rashba SOC had been demonstrated in previous studies [10, 12, 13].

The intensity of the circularly polarized light has a Gauss profile, thus an inhomogeneous spin density will be excited on the sample plane. One can assume that the effective spin density (*N*
_{eff}) flowing through the region of *x*=0 has a Gaussian distribution [11, 13], i.e., \(N_{\text {eff}}=g\tau _{s}\frac {c}{\sigma }e^{-x^{2}/\sigma ^{2}}\), where *g*,*τ*
_{
s
},*c*, *x*, and *σ* are the generation rate of spin-polarized carriers, the spin relaxation time, an arbitrary constant, the spot coordinate along the *x* axis, and the standard deviation of the Gaussian distribution, respectively. On the one hand, under normal incidence, the gradient of the spin density will induce a diffused spin polarization current (SPC): \(\textit {\textbf {q}}_{r}^{z}=-{D_{s}}\nabla _{r}\textit {n}^{z}(\textit {r})\), where *D*
_{
s
} is the spin diffusion coefficient of the photoinduced carriers, *n*
^{z}(*r*) the spin density along the *z* direction, and *r* the radial direction in the *x*−*y* plane. According to the RSHE [3, 14], a transverse electric current (density) perpendicular to both the direction of the SPC and the direction of the spin polarization is produced, which can be expressed as \(\textit {j}=\gamma {e}\textit {\textbf {q}}_{r}^{z}\times \hat {\textit {z}}\), where *e*, *γ* are the elementary charge, the spin-orbit interaction coefficient based on RSHE, respectively. As a result, a swirling electric current will be induced around the light spot, which will further generate an observed ACPGE current *j*
_{ACPGE}/*e*=−*γ*
*D*
_{
s
}∇*N*
_{eff} (shown in Fig. 1
c). On the other hand, when a circularly polarized light irradiates vertically on the sample, the flow of the spin-polarized carriers driven by the longitudinal electric field will also lead to a transverse AHE current which is also derived from the RSHE, as shown in Fig. 1
b [6–10]. And the AHE current can be expressed as *j*
_{AHE}/*e* = *γ*
*μ*
_{
s
}
*E*
*N*
_{eff} [11], where *μ*
_{
s
} is the spin mobility of the photoinduced carriers, and *E* is the external electric field. Thus, the total spin-related photoinduced current along the two circle electrodes can be expressed as

where *j*
_{total} is equal to *j*
_{ACPGE+AHE}.

## Methods

The sample studied here is an undoped In _{0.15}
*Ga*
_{0.85} As/Al _{0.3}
*Ga*
_{0.7}As MQW grown by molecular beam epitaxy. A 200-nm buffer layer is initially deposited on (001) SI-GaAs substrate, followed by ten periods of 100-Å In _{0.15}
*Ga*
_{0.85}As/ 100-Å Al _{0.3}
*Ga*
_{0.7}As QWs. Then, a 500 Å Al _{0.3}
*Ga*
_{0.7}As layer and 100 Å GaAs cap layer are deposited. The sample is cleaved into a narrow strip along the GaAs [1\(\bar {1}\)0] direction with a width of 4 mm and a length of 12 mm. The geometry has been shown in Fig. 1
a, where two circle ohmic electrodes (whose radius are both 0.25 mm) with a distance of 2.5 mm and two strip ohmic electrodes (whose size are both 0.5 × 3 mm) with a distance of 10 mm were made along *y* and *x* direction, respectively, by indium deposition and annealed at about 420 °C in nitrogen atmosphere.

The experimental setup is described as follows. A modelocked Ti:sapphire laser with a repetition rate of 80 MHz serves as the radiation source. The incident light goes through a polarizer and a photoelastic modulator (PEM), of which the peak retardation is set to be *λ*/4, to yield a modulated circularly polarized light with a fixed modulating frequency at 50 KHz. By using an optical chopper with the rotation frequency of 223 Hz, the spectra of common photoinduced currents (*I*
_{PC}) are also measured for comparison, which show clearly the energy positions corresponding to 1hh-1e (the first valence subband of heavy holes to the first conduction subband) and 1lh-1e (the first valence subband of light holes to the first conduction subband) transitions (see Fig. 1
d). The Gaussian profile light beam irradiates vertically on the sample with a diameter of about 1.7 mm at the perpendicular bisector of the two circle electrodes. The external electric field applies to the strip electrodes. The photogalvanic currents are collected through the two circle electrodes by two lock-in amplifiers with the synchronization frequencies set to be 50 KHz and 223 Hz, respectively.

## Results and discussion

In the experiment, the total spin-related currents at electric fields of 0, + 4, and −4 V/cm are measured as a function of the spot location at three different wavelengths. As shown in Fig. 2 a, under the electric field of 0 V/cm, the total spin-related currents which are actually contributed only from ACPGE currents reverse the sign from the left to right side. The electric fields selected to be + 4 and −4 V/cm ensured that the AHE currents were comparable with the ACPGE currents. The total spin-related currents were fitted very well in Eq. 1 (shown in Fig. 2 b, c). The AHE current was extracted by subtracting the ACPGE current from the total spin-related photocurrent. The extracted curves in Fig. 2 d exhibit symmetric Gaussian-like distribution, and the directions corresponding to electric fields of + 4 and −4 V/cm are opposite, being consistent with the mechanism of AHE.

Next, we extracted the extreme values of ACPGE and AHE currents from Fig. 2 a, d, respectively. To our surprise, the extracted curve shape in Fig. 3 a is quite different from that in Fig. 3 b. In concrete terms, there are no obvious peaks corresponding to 1hh-1e and 1lh-1e in Fig. 3 a, while there are distinct peaks corresponding to 1hh-1e and 1lh-1e in Fig. 3 b, which are due to that the diffusion coefficient and the mobility of the photoinduced spin-polarized carriers have different relationships with excitation wavelength. According to [11], by observing the two extreme points of total spin-related current, the ratio of the spin diffusion coefficient to the mobility of the photoinduced spin-polarized electrons can be calculated directly. However, in our experiment, the extreme positions are not always clear to be figured out especially at small wavelengths (see Fig. 2 b, c). Despite this, one can still get the ratio of the spin diffusion coefficient to the mobility of the photoinduced spin-polarized carriers by fitting the experiment data using Eq. 1.

From the fitted data in Fig. 4, we can see that the ratio is independent of external electric field, and the ratio remains almost a constant around 1hh-1e while it decreases sharply when decreasing the wavelength to 1lh-1e. In order to understand the possible reasons, we have to take the spin current from holes into consideration. According to the experiments of ISHE from [6, 15], we can deduce that the directions of the spin transverse force for electrons and holes are the same if the directions of spin polarization are the same and the directions of spin current are also the same; as a result, the directions of the spin-polarized electric currents for electrons and holes are opposite. Therefore, we can depict the detailed spin diffusion electric currents and spin drifting electric currents corresponding to 1hh-1e and 1lh-1e, respectively (see Fig. 5). For 1hh-1e transition, the effective spin density can be simplified as *N*
_{eff}=*τ*
_{
e
}
*N*
_{0} for electrons and *N*
_{eff}=*τ*
_{
hh
}
*N*
_{0} for heavy holes, where \(N_{0}=g_{0}\frac {c}{\sigma }e^{-x^{2}/\sigma ^{2}}\) is a constant (where *g*
_{0} is the generation rate for 1hh-1e transition), *τ*
_{
e
} is the spin relaxation time of electrons, and *τ*
_{
hh
} is the spin relaxation time of heavy holes. For 1lh-1e transition, *N*
_{
eff
} can be simplified as *N*
_{eff}=*τ*
_{
e
}
*N*
_{1} for electrons and *N*
_{eff}=*τ*
_{
lh
}
*N*
_{1} for light holes, where \(N_{1}=g_{1}\frac {c}{\sigma }e^{-x^{2}/\sigma ^{2}}\) is a constant (where *g*
_{1} is the generation rate for 1lh-1e transition), and *τ*
_{
lh
} is the spin relaxation time of light holes. So, the total spin-polarized electric current corresponding to 1hh-1e and 1lh-1e can be expressed as

and

respectively.

Thus, the ratios of the spin diffusion coefficient to the mobility of the photoinduced spin-polarized carriers can be defined as

for 1hh-1e transition and

for 1lh-1e transition. The Rashba effect which is stronger for the first valence band than that for the first conduction band and is stronger for the first light hole subband than that for the first heavy hole subband in *p*-type quantum wells has been demonstrated by several authors [16–18]. Supposing the situation for undoped MQW is similar to that for the *p*-type quantum wells. Thus, the reciprocal spin Hall coefficient (which is proportional to Rashba effect) of light holes (*γ*
_{
lh
}) is probably far larger than that of electrons (*γ*
_{
e
}). Assuming *μ*
_{
lh
} is comparable with *μ*
_{
e
}, while *D*
_{
lh
} is far smaller than *D*
_{
e
}, and *τ*
_{
e
} is at the same order with *τ*
_{
lh
}, only then can we get *μ*
_{
lh
}
*γ*
_{
lh
}
*τ*
_{
lh
}≫*μ*
_{
e
}
*γ*
_{
e
}
*τ*
_{
e
}. As a result, \(\left (\frac {D_{s}}{\mu _{s}}\right)_{1lh-1e}\) is smaller than \(\left (\frac {D_{s}}{\mu _{s}}\right)_{1hh-1e}\) (see Fig. 4). Around the 1hh-1e transition, the contribution to the ratio of the spin diffusion coefficient to the mobility of spin-polarized carriers is mainly from the 1hh-1e transition. With wavelength further decreasing, the contribution to the ratio from 1hh-1e transition became more and more weak, while the contribution from 1lh-1e became dominant. Therefore, the ratio remains almost a constant around 1hh-1e while it decreases sharply when decreasing the wavelength to 1lh-1e.

At last, we studied the temperature dependence of ACPGE and AHE currents. From Fig. 6, we can see that the ACPGE and AHE currents are significantly enhanced with temperature decreasing, especially at 1hh-1e transition. The spectra of AHE currents at different temperatures in Fig. 6
c are obtained by subtracting the data in Fig. 6
b from the corresponding data in Fig. 6
a. If we just consider the 1hh-1e transition, by contrasting the value of ACPGE current with that of AHE current at different temperatures, we can further deduce the temperature dependence of \(\frac {D_{s}}{\mu _{s}}\). As shown in Fig. 7, we have found that the ratio first increased and then decreased when the temperature decreased from room temperature. In these quantum wells, electron spin relaxation time is considered to vary approximately as *T*
^{−2}, i.e., *τ*
_{
e
}=*A*
_{
e
}
*T*
^{−2}, where *A*
_{
e
} is a constant, which is quiet different from the assumption of a narrow quantum wells in [2, 10, 19]; and the hole spin relaxation time is roughly proportional to *T*
^{−1}, i.e., *τ*
_{
hh
}=*A*
_{
hh
}
*T*
^{−1}, where *A*
_{
hh
} is a constant [20]. Assuming that \(\frac {D_{e}}{\mu _{e}}=\frac {\chi _{e}{k_{B}{T}}}{e}\) and \(\frac {D_{\textit {hh}}}{\mu _{\textit {hh}}}=\frac {\chi _{\textit {hh}}{k_{B}{T}}}{e}\), where *χ*
_{
e
} and *χ*
_{
hh
} are correction factors for the common Einstein relationship for the spin-polarized electrons and holes, respectively. At high temperatures (≥80 K), the mobility of GaAs/AlGaAs two-dimensional electron and hole gases are respectively proportional to *T*
^{−2.4} and *T*
^{−2} given by former studies [21, 22]. For spin-polarized electrons and holes, in order to simplify the discussion, we suppose the ratio of spin mobility between spin-polarized electrons and spin-polarized holes is proportional to *T* (there were few relevant reports on the temperature dependence of spin mobility), i.e., \(\frac {\mu _{e}}{\mu _{\textit {hh}}}=\lambda _{0}T\), where *λ*
_{0} is a constant. Then, Eq. 4 can be further expressed as

Ignoring the temperature dependence of *γ*
_{
e
} and *γ*
_{
hh
}, one can use Eq. 6 to fit the data in Fig. 7; and the correction factors for Einstein relationship for spin-polarized electrons and heavy holes are fitted as *χ*
_{
e
}=93 and *χ*
_{
hh
}=286, respectively. The factor of spin-polarized heavy holes is almost 3 times larger than that of spin-polarized electrons. According to the Einstein relationship for electron transport, the value of \(\frac {D}{\mu }\) is estimated to be about 0.026 V at room temperature, which is much smaller than the value of \(\frac {D_{s}}{\mu _{s}}\) we obtained in this work. We would like to clarify that \(\frac {D_{s}}{\mu _{s}}\) for spin transport is not necessarily the same as \(\frac {D}{\mu }\) for electron transport. The essential difference is that the Einstein relationship is derived based on the conservation law of electrons; however, spin is not conservative [11]. We believe that the Einstein relationship for spin should be different for different semiconductor materials.

## Conclusions

In conclusion, the spin diffusion and drift at different wavelengths and different temperatures have been studied in undoped InGaAs/AlGaAs MQW. By using the AHE and ACPGE which are all derived from RSHE, we obtained the ratio between the spin diffusion coefficient and the mobility of spin-polarized carriers. From the wavelength dependence of the ratio, we found that the spin diffusion and drift of holes became as important as electrons in this undoped MQW, and the ratio for light holes was much smaller than that for heavy holes at room temperature. From the temperature dependence of the ratio corresponding to the 1hh-1e transition, we believed the ratio is contributed by the combined effect of spin-polarized electrons and spin-polarized heavy holes. The correction factors for the common Einstein relationship for spin-polarized electrons and heavy holes are firstly obtained to be 93 and 286, respectively. It is worth noting that the AHE and ACPGE measurements used in this study are conducted under ambient conditions with a simple setup and operation, which provides a good method for the study of spin-related diffusion and drift.

## References

žutić I, Fabian J, Sarma SD (2004) Spintronics: fundamentals and applications. Rev Mod Phys 76(2): 323.

Wu MW, Jiang JH, Weng MQ (2010) Spin dynamics in semiconductors. Phys Rep 493(2): 61.

He XW, Shen B, Chen YH, Zhang Q, Han K, Yin CM, Tang N, Xu FJ, Tang CG, Yang ZJao (2008) Anomalous photogalvanic effect of circularly polarized light incident on the two-dimensional electron gas in Al

_{ x }*Ga*_{1−x }N/GaN heterostructures at room temperature. Phys Rev Lett 101(14): 147402.Mei FH, Tang N, Wang XQ, Duan JX, Zhang S, Chen YH, Ge WK, Shen B (2012) Detection of spin-orbit coupling of surface electron layer via reciprocal spin Hall effect in InN films. Appl Phys Lett 101(13): 132404.

Duan JX, Tang N, Ye JD, Mei FH, Teo KL, Chen YH, Ge WK, Shen B (2013) Anomalous circular photogalvanic effect of the spin-polarized two-dimensional electron gas in Mg

_{0.2}*Zn*_{0.8}O/ZnO heterostructures at room temperature. Appl Phys Lett 102(19): 192405.Saitoh E, Ueda M, Miyajima H, Tatara G (2006) Conversion of spin current into charge current at room temperature: inverse spin-Hall effect. Appl Phys Lett 88(18): 182509.

Miah MI (2007) Observation of the anomalous Hall effect in GaAs. J Phys D: Appl Phys 40(6): 1659.

Yin CM, Tang N, Zhang S, Duan JX, Xu FJ, Song J, Mei FH, Wang XQ, Shen B, Chen YH, Yu JL, Ma H (2011) Observation of the photoinduced anomalous Hall effect in GaN-based heterostructures. Appl Phys Lett 98(12): 122104.

Yu JL, Chen YH, Jiang CY, Liu Y, Ma H, Zhu LP (2012) Observation of the photoinduced anomalous Hall effect spectra in insulating InGaAs/AlGaAs quantum wells at room temperature. Appl Phys Lett 100(14): 142109.

Zhu LP, Liu Y, Jiang CY, Yu JL, Gao HS, Ma H, Qin XD, Li Y, Wu Q, Chen YH (2014) Spin depolarization under low electric fields at low temperatures in undoped InGaAs/AlGaAs multiple quantum well. Appl Phys Lett 105(15): 152103.

Mei FH, Zhang S, Tang N, Duan JX, Xu FJ, Chen YH, Ge WK, Shen B (2014) Spin transport study in a Rashba spin-orbit coupling system. Sci Rep 4: 4030.

Yu JL, Chen YH, Jiang CY, Liu Y, Ma H (2011) Room-temperature spin photocurrent spectra at interband excitation and comparison with reflectance-difference spectroscopy in InGaAs/AlGaAs quantum wells. J Appl Phys 109(5): 053519.

Zhu LP, Liu Y, Jiang CY, Qin XD, Li Y, Gao HS, Chen YH (2014) Excitation wavelength dependence of the anomalous circular photogalvanic effect in undoped InGaAs/AlGaAs quantum wells. J Appl Phys 115(8): 083509.

Shen SQ (2005) Spin transverse force on spin current in an electric field. Phys Rev Lett 95: 187203.

Ando K, Takahashi S, Ieda J, Kurebayashi H, Trypiniotis T, Barnes CHW, Maekawa S, Saitoh E (2011) Electrically tunable spin injector free from the impedance mismatch problem. Nat Mater 10(9): 655.

Gvozdic DM, Ekenberg U (2006) Superefficient electric-field-induced spin-orbit splitting in strained p-type quantum wells. Europhys Lett 73(6): 927.

Dai X, Zhang FC (2007) Light-induced Hall effect in semiconductors with spin-orbit coupling. Phys Rev B 76(8): 085343.

Winkler R (2003) Spin-orbit coupling effects in two-dimensional electron and hole systems, Vol. 191. Springer, Germany.

Malinowski A, Britton RS, Grevatt T, Harley RT, Ritchie DA, Simmons MY (2000) Spin relaxation in GaAs/Al

_{ x }*Ga*_{1−x }As quantum wells. Phys Rev B 62: 13034.Lü C, Cheng JL, Wu MW (2006) Hole spin dephasing in p-type semiconductor quantum wells. Phys Rev B 73(12): 125314.

Mendez EE, Price PJ, Heiblum M (1984) Temperature dependence of the electron mobility in GaAs-GaAlAs heterostructures. Appl Phys Lett 45(3): 294.

Störmer HL, Gossard AC, Wiegmann W, Blondel R, Baldwin K (1984) Temperature dependence of the mobility of two-dimensional hole systems in modulation-doped GaAs-(AlGa) As. Appl Phys Lett 44(1): 139.

## Acknowledgements

The work was supported by the 973 program (2012CB921304, 2013CB632805, and 2012CB619306) and the National Natural Science Foundation of China (61474114, 60990313, and 11574302).

## Author information

### Authors and Affiliations

### Corresponding authors

## Additional information

### Competing interests

The authors declare that they have no competing interests.

### Authors’ contributions

LZ conducted the experiments and wrote the paper. LZ and YC designed the experiments and performed the sample fabrications. All authors contributed through scientific discussions and read and approved the final manuscript.

## Rights and permissions

**Open Access** This article is distributed under the terms of the Creative Commons Attribution 4.0 International License(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

## About this article

### Cite this article

Zhu, L., Liu, Y., Huang, W. *et al.* Spin transport in undoped InGaAs/AlGaAs multiple quantum well studied via spin photocurrent excited by circularly polarized light.
*Nanoscale Res Lett* **11**, 8 (2016). https://doi.org/10.1186/s11671-015-1218-3

Received:

Accepted:

Published:

DOI: https://doi.org/10.1186/s11671-015-1218-3