Access Official Technical PDF

Share Summary:LinkedInEmail

A TiO2 nanorod UV photodetector can combine a nanostructured light-absorbing layer with a semiconductor junction to convert ultraviolet illumination into measurable current. One demonstrated design uses rutile titanium dioxide (TiO2) nanorods grown on p-type silicon, forming a heterojunction between n-type TiO2 and p-type Si. The reported device offers a useful case study in how fabrication choices, material structure and electrical bias relate to UV sensor behavior.

How the TiO2–silicon heterojunction is built

The device uses a TiO2 seed layer on a p-type, (111)-oriented silicon substrate. The seed layer was prepared by radio-frequency reactive magnetron sputtering. Rutile TiO2 nanorods were then grown on the seeded substrate using chemical bath deposition (CBD), a solution-based process that does not require highly specialized growth equipment.

For the reported process, titanium(III) chloride was mixed with deionized water, and urea was added to adjust the solution to approximately pH 0.7. After stirring, the silicon substrate was immersed vertically in the bath, which was heated to 55°C. Following two hours of growth, the sample was rinsed, dried and heat-treated in air at 550°C for two hours to improve crystallinity.

Aluminum was deposited as a grid front contact on the nanorod layer, while indium was applied to the back of the silicon substrate. The resulting stack was Al/TiO2 nanorods/p-Si(111)/In, with a reported active area of about 0.39 cm². In this arrangement, the interface between n-type TiO2 and p-type silicon forms the junction central to the sensor’s electrical response.

What material characterization revealed

X-ray diffraction identified the deposited TiO2 as polycrystalline tetragonal rutile. A prominent diffraction peak at 2θ = 27.4° was attributed to the (101) plane, alongside peaks associated with other rutile planes. Field-emission scanning electron microscopy showed a dense distribution of nanorods, with some rods grouped into flower-like formations.

Most nanorods measured approximately 18–22 nm in diameter and averaged about 85 nm in length. The flower-like structures were reported as roughly 350–400 nm across, each composed of more than 40 rods. These observations describe the morphology of the tested sample; they are not, by themselves, a guarantee of equivalent structure in other growth processes or at different scales.

Photoluminescence measurements showed a strong ultraviolet emission centered near 366 nm and weaker visible emissions around 436 and 716 nm. The researchers associated the UV peak with near-band-edge emission and the visible features with transitions involving surface-related states. The stronger UV emission relative to the visible peaks was interpreted as an indication of comparatively low structural and surface defect levels in the prepared nanorods.

UV sensor performance and junction behavior

Current–voltage measurements were taken in darkness and under illumination at 365 nm, with an incident optical intensity of 2.3 mW/cm². The device displayed rectifying behavior, consistent with formation of a p–n junction. At an applied bias of 5 V, the reported current was 1.05 × 10-5 A in darkness and 6.7 × 10-4 A under UV illumination. The paper also reports a sensitivity of 62.8 × 10² under the stated illumination and bias conditions.

The proposed operating mechanism helps explain why bias matters. UV absorption generates electron–hole pairs in or near the TiO2 nanorods and junction region. The electric field across the depletion region separates the carriers, with electrons moving toward the n-type side and holes toward the p-type side. Their collection at the contacts contributes to photocurrent. At small applied bias, the depletion region limits current and the response is less sensitive to illumination; at higher bias, the measured current rises.

The reported photoresponse spectrum covered wavelengths from 320 to 500 nm at 5 V. Responsivity peaked at 467 mA/W under 365 nm illumination, and the response declined sharply near 380 nm, close to the reported absorption edge of rutile TiO2. A photocurrent of 6.7 × 10-4 A was also reported at 325 nm. These values refer to different measurement wavelengths and should be read in that context. The study further reports a current gain of 63.8 at 365 nm under 5 V bias.

Switching response and repeatability

To evaluate dynamic behavior, the sensor was exposed to pulsed 365 nm UV light while the illumination was repeatedly switched on and off. The current traces showed recurring rectangular-shaped profiles, with similar photocurrent levels across the displayed cycles. At 5 V, the reported response time was 48 ms and the recovery time was 40 ms. Response time was defined as the interval for current to rise from 10% to 90% of its saturation value; recovery time measured the decline from 90% back to 10%.

The authors relate the device’s response to the nanorod structure, the TiO2–silicon junction and carrier collection. They also discuss oxygen-related surface trapping as a factor that may affect carrier recombination and lifetime. These are proposed explanations for the observed behavior, rather than a complete account of how every device will perform in a different operating environment.

Practical implications for UV sensing

For engineers assessing semiconductor UV sensors, this design illustrates a fabrication route that combines sputtered seed-layer preparation, solution-based nanorod growth and conventional front and back contacts. The reported measurements establish proof of device operation under specified laboratory conditions, including the illumination wavelength, intensity and applied bias. They do not, on their own, establish performance under industrial environmental conditions or define system-level requirements such as packaging, calibration or long-term service life.

The practical takeaway is that the TiO2 nanorod/p-silicon heterojunction is a promising device architecture for further evaluation as a UV photodetector. Its reported rectification, responsivity and millisecond-scale switching provide useful reference points, while the CBD process offers a comparatively straightforward route for forming the active nanostructured layer. Application decisions should account for the test conditions and validate the complete sensor assembly against the intended measurement and operating requirements.

Access Official Technical PDF

About The Author