We demonstrate that the effective optical band gap of bulk semiconductor CdZnTe:V crystals can be varied and controlled by dual application of light and electric field, both at moderate strengths. When the combined effect of light intensity and applied field exceeds a threshold, the effective band-gap shift persists even after the light is turned off, as long as the electric field is applied. The persistent effective band-gap shift is accompanied by persistent photocurrent and persistent change

more »

... in the (220) d spacing of crystalline lattice. However, all persistence effects can be reset to their original values when the applied field is turned off. Finding an efficient way to tune the electronic band gap in semiconductors could open new horizons in microelectronics and optoelectronics. However, varying the band gap requires large dc electric fields (the static Franz-Keldysh effect, in which the effective band gap is modified by field-assisted electron tunneling 1,2 ) or magnetic fields, 3 or exposing the material to high pressure or temperature. 4 These methods are greatly limited in their usefulness, since the effect on band gap is very small. For example, electric fields of several tens of kilovolts per centimeter 1,2 or magnetic fields of several Tesla 3 lead to a just tiny shift of a few meV in the optical band gaps of GaAs or InSb (1.43 and 0.17 eV, respectively). For many applications, it is highly desirable to tune the band gap by light, in real time and in a reversible way. Unfortunately, light-induced reversible effective band-gap shift based on the dynamic (ac) Franz-Keldysh effect 5 requires very high intensities (>5×10 9 W/cm 2 ). Here we demonstrate experimentally that the effective optical band gap of a bulk semiconductor with deep impurity levels can be varied by light and electric field, both at moderate strengths. The observed effective band-gap shift is the largest ever reported in any bulk material, and cannot be explained in the framework of any previously known effect. 1-5 We use moderate field (∼2 kV/cm) and moderate light intensity (∼1 W/cm 2 ) to modify the absorption edge 6 of a V-doped CdZnTe crystal (CZT:V) by as much as 60 meV, which is nearly 4% of its original value (1.56 eV). The band-gap change is induced by a laser beam at photon energies far below band gap. It is devoid of any thermal effects and is fully reversible: a few seconds after the field and the light are turned off, the band gap returns to its initial value. When the light intensity and electric field exceeds some threshold values, the band-gap shift becomes persistent, remaining unchanged as long as the electric field is applied, even after the light is turned off. Under the same conditions we also find persistent relative change (∼0.04%) of the (220) lattice d spacing. We believe that this method of controlling the band gap of semiconductors by light can be extended to other materials, and can be used to enhance nonlinear material properties for THz generation and control nonlinear ultrafast effects, such as two-photon absorption and self-phase modulation. Our bulk Cd 1−x Zn x Te:V single crystals were grown by the modified horizontal Bridgman technique, 7 with nominal Zn concentration x = 0.01, and nominal V doping of ∼10 17 cm −3 . In our experiments [ Fig. 1(a) ] the crystal is illuminated uniformly with a 980 nm wavelength laser beam, corresponding to photon energies much smaller than the band gap. We denote this beam as the "control beam." We apply a dc field E along the crystalline 001 direction, while the control beam is propagating along the 110 direction. The transmission of the sample is probed by a weak light source [see Fig. 1(a) ] in the wavelength range 800-1000 nm. First, we investigate the light-induced change in the absorption spectrum. At moderate fields (E < 2.2 kV/cm), increasing the control-beam intensity in the range 0.1-1.3 W/cm 2 results in a remarkable shift of the absorption edge toward longer wavelengths [ Fig. 1(b) ]. The effective band gap decreases linearly with light intensity [ Fig. 1(c) ], with a maximum bandgap shift of 64 meV, at I l = 1.3 W/cm 2 and E = 2.2 kV/cm. It is worth comparing the magnitude of the observed effect to the Franz-Keldysh and quantum-confined Stark effects in similar materials. The theory of the Franz-Keldysh effect 4 yields a band-gap shift of 10 meV at E = 50 kV/cm. In practice, the Franz-Keldysh effect in bulk crystals is even smaller, for example, in ZnS the band-gap shift due to the Franz-Keldysh effect is 3.8 meV at E = 50 kV/cm, 8 while in GaAs this shift is 200 μeV at E = 5 kV/cm. 9 Evidently the values observed by us are much higher than those ever reported for the Franz-Keldysh effect. In fact, the band-gap shifts observed in our bulk crystals are comparable to the quantum-confined Stark effect in quantum-well structures. For example, in Ge/SiGe quantum wells, an absorption peak shift by 48 nm (30 meV) was found at E = 80 kV/cm. 10 The band-gap shift we measure is twice as large at E = 2.2 kV/cm, that is, at 36 times smaller electric field, despite that we conduct experiments with bulk crystals. These facts lead us to conclude that our effect, being much stronger, has a different origin than the Franz-Keldysh effect. Having demonstrated huge light-induced shift in the effective band gap, we now show that this shift is fully reversible. We measure the absorption spectrum as a function of time, upon turning on simultaneously the control beam and E [ black hollow circles in Fig. 2(a) ], and upon turning both of them off blue solid circles in Fig. 2(a) ]. Figure 2 (a) displays that the band-gap shift is indeed fully reversible, and that the response time for the shift is ∼10 s.