What does it mean if the resistance of a semiconductor increases due to light? I have synthesized an $n$-type semiconductor material $\text{ZnO}$. Under light illumination, its resistance keeps increasing. What are the reasons for this?
 A: Paradoxical Drop in ZnO Semiconductor Current with Photonic Stimulation

Why does a ZnO semiconductor sample increase its resistance (constant voltage and current drop) with visible light photonic stimulation?

Experimental Setup, Materials, Observations, and Background:


*

*The sample is a thin film, n-type, ZnO semiconductor on a glass substrate.  The sample is connected in series with a picoammeter, and a 9 Volt power supply. 

*Upon application of power, the current through the ZnO sample gradually rises from 86  nanoamps at 2 seconds, to 92 nanoamps over the next 80 seconds.

*Illumination of the sample with various frequencies (red, orange, and green LED light), the current drops.


Why does the current decrease? This result is unexpected because more carrier pairs are formed with more photon absorption. Solar cells (which have a pn-junction) produce carriers by visible light absorption. 


*The band gap of ZnO is between 3.3-3.4 eV. The peak of solar radiation is 1.5 eV, and the band gap of the most efficient solar cell semiconductors is around that band gap energy, ($E_g$ Si 1.1 eV, CdTe 1.5 eV, GaAs 1.53 eV). Germanium's band gap is too low at .4 eV, so most of the photonic energy is converted into heat.  On a strictly theoretical basis, the ZnO $E_g$ is too large for visible light to produce any carrier pairs, as the $E_g$ of ZnO is much higher than the energy of red (1.65-2.0 eV), orange (2.0-2.15 eV), and green (2.17-2.5 eV) photons.

*If any visible light photons activate ZnO carriers into the conduction band, it may be due to the coincidence of 1) a photon and 2) statistically possible collisions which result in producing higher thermal energies, intersecting at the same time and location as the absorption of a photon. But, this "ZnO and conduction after visible light absorption" experiment demonstrates that such superimposition produces few carriers, and those numbers are overcome by another factor (see below). 

*Current is mediated by carriers (electrons and holes) in the conduction band (i.e., carriers in the valence band cannot move, and hence cannot mediate current).  ZnO carrier pairs must have at least 3.3-3.4 eV of energy to reach the conduction band and mediate current.  

*Even in the absence of other energy inputs - thermal collisions between lattice atoms still occasionally produce carrier pairs, as demonstrated by the small current produced in ZnO under voltage.  But, given the high band gap energy of ZnO, only rare thermal collisions between lattice atoms impart sufficient kinetic energy to free electrons from their lattice bonds and raise their energy to the conduction band.  

*After having acquired sufficient kinetic energy to reach the conduction band, the carrier pair (free electron and hole) are capable of drifting under the influence of an externally applied $E$ field (from the voltage source).  With or without an external field, the conduction band electrons move in random Brownian motion.  But, with a superimposed $E$ field, the carriers drift slowly under its influence; electrons drift toward the positive pole and holes toward the negative.   


Answer: Why Current in ZnO Drops upon Illumination with Visible Light


*

*The very low current flow (92 nanoamps) with a 9 V voltage source and slow buildup to steady-state current (80 seconds) indicate that the carriers are sparse and only slightly mobile.

*Solar cells are most commonly made of n and p doped Si (for economical reasons) or GaAs (for more efficient operation).  Photons (of sufficient energy) strike the semiconductor atoms and form carrier pairs. The freed electrons in the n-type material diffuse into the p-type zone and vice versa with holes, which creates a charge differential across the pn junction.  This charge differential creates an $E$ field across the junction, but this voltage differential extends to opposite sides of the solar cells.  This differential is discharged through a conductor, which creates a current through an external circuit and thereby delivers power to a load in the current-mediated equalization of the charge differential. More photons are required to recharge the solar cell charge-differential.


But, in a ZnO thin film semiconductor, when illuminated with red, orange, and green light, the current drops. This is the exact opposite effect of that seen in a solar cell.  Hence the question about the mechanism. 


*Without addition to a thermally activated collision, none of the chosen visible frequency photons form carrier pairs with sufficient energy to overcome the band gap (3.3-3.4 eV) between the valence band and conduction band.

*Carriers with insufficient energy to cross the band gap cannot enter the conduction band and participate in conduction.  

*Still, visible frequency photons are absorbed and transfer their energy to
momentarily create partially free carrier pairs.  But, since these electron and hole pairs do not have enough energy to reach the conduction band, they will rapidly recombine, collide with the lattice and heat thereby heat the lattice.  (Note, heating a solar cell reduces its efficiency, probably for the same reason that visible light reduces the conductivity of a ZnO semiconductor.) 

*But, some of these transient, low energy (sub $E_g$ energy) electron and hole pairs formed by visible photons will collide with the energized, conduction band electrons and holes and remove them as carriers.

*As a result, the carrier population of the conduction band declines, thereby reducing the current flow through the ZnO thin film semiconductor.

