We demonstrated the dual-detectable DNA-CTMA/n-GaN photodiode (DG-PD) for ultraviolet and visible lights. Halogen and UV lamps are employed to recognize the visible and UV wavelength, respectively. The DG-PD under dark condition has a negative-bias shift of current-voltage (I-V) curves by 0.78 V compared to reference diode without DNA. However, the I-V curves move towards positive bias side by 0.75 V and 1.02 V for the halogen- and UV-exposed photodiode, respectively. These cause electrically different polarity and amount for halogen- and UV-induced photocurrents, indicating that the DNA-CTMA on n-GaN is quite effective for recognizing visible and UV lights as a dual-detectable photodiode. The formation and charge transport mechanisms are also discussed.
© 2014 Optical Society of America
Photodiodes (PDs), which can recognize wider wavelengths covering from visible to ultraviolet (UV) regions, are conceptually ideal and technologically essential because they can provide a dual-detectable operation through just one photon-detection device instead of two devices consisting of each visible and UV detection parts. In addition, these PDs will be very attractive for the advanced applications areas such as medicare, display, general illumination, weather-monitoring, and military [1–3]. Particularly, a single PD with multiple detections of various wavelengths can eliminate the difficulties of fabrication and assembling several detectors with individual cooling assemblies. UV and visible dual-band detectors could be used in applications where the detection of both UV and visible regions is important. For example, display monitors exhibiting visible to UV rays have significant intensity variations in the emission spectra ranging from UV to visible regions. Moreover, the identification of multicolor with high responsivity at designated wavelengths can be useful applications like imaging, optical communication, monitoring, target identification, and remote controlling [4,5]. Until now many studies on high-performance PDs have been carried out but PDs that can satisfy the dual-detection characteristics have never been demonstrated due to inherent material limitations. For examples, wide band-gap GaN-based PDs are only suitable for detecting UV, and Si-based PDs show a good performance in a specified wavelength region ranging from visible to infra-red (IR) although GaN- and Si-based PDs have been already commercialized to several photo-sensing applications. Thus, how to achieve the dual detection of visible and UV for PDs is technologically essential to initiate advanced technology based on smart-sensing PDs.
In order to meet the dual detection for visible and UV lights in PDs, we propose deoxyribonucleic acid-cetyltrimetylammonium (DNA-CTMA) biopolymer-based GaN PDs, and describe the reasons of the proposed PD in brief. It has been known that DNA-CTMA has various material characteristics from insulating to semiconducting, and has also typical characteristics such as wider energy gap of ~4.7 eV , unidirectional charge transfer nature , and very sensitive to lights ranging from UV to visible . In particular, getting better understandings of charge transport in DNA-CTMA is quite important to develop high-performance PDs. Several models such as multistep hopping and superexchange (tunneling) have been proposed so far but still under debate . However, the distinct property, in other words, unidirectional charge transport in DNA-CTMA could play an important role in determining photo-conductive characteristics of DNA-CTMA-based PDs. In addition, when we take into account the fact that DNA-CTMA is a good interfacial layer for hole transport and electron blocking at the heterojunction interface , DNA-CTMA could provide a useful solution for finding an advanced PD configuration having higher conversion efficiency (photons to electron-hole pairs). GaN-based materials have also ultimate properties and thus have been widely employed to develop high-efficiency optoelectronic devices [10,11]. Especially, GaN grown on sapphire substrate contains high density of defects that act as a deep level center [12,13]. In spite of the fact that how to remove the crystalline defects in GaN-based materials is one of key issues, deep-level interaction (or excitation and transition) originating from the defects may be useful when DNA-CTMA is contacted with GaN for recognizing visible and UV lights. That is, if the DNA-CTMA acts as an optical light absorber on GaN (having band-to-band and deep-level interactions), the PDs showing the dual-detection of visible and UV lights could be attainable. The detailed descriptions for this will be explained and discussed later. In this present work, therefore, we demonstrate the high-quality DNA-CTMA-based GaN photodiode for showing the dual-detection. Reference diode without DNA-CTMA biopolymer was also fabricated and investigated to compare with the DNA-CTMA-based photodiode. Current-voltage (I-V) and transient photocurrent techniques were employed to investigate the optical and electrical characteristics. It should be notified that the movement of the I-V curves (toward forward bias side) and the electrically different-polarity photocurrents (depending on halogen and UV illumination conditions) are distinctly solid evidences for the dual detection.
In this work, 5 µm thick Si-doped GaN films were grown on c-plane sapphire substrate by using metalorganic chemical vapor deposition. From Hall-effect measurements, the carrier concentration was determined to be 4~5 × 1018 cm−3. The n-GaN wafer was ultrasonically degreased with acetone and isopropyl alcohol (IPA) for 5 min in each step, and then rinsed in deionized (DI) water. Using buffered-oxide etch (BOE) solution, the n-GaN was surface-treated to remove native oxides . For the fabrication of the photodiode, the Ti/Al/Ni/Au (12/200/40/100 nm) ohmic contact was formed on the surface-treated n-GaN by using metal shadow mask and electron-beam evaporation. For obtaining DNA biopolymer, we performed following protocols using a freeze-dried DNA (extracted from salmon fish sperm) powder purchased from Sigma Aldrich . Because the molecular weight of the DNA strands is millions of Daltons (Da) that can mean an unwanted ultrahigh electrical conductivity in DNA, we decreased the molecular weight down to 200 kDa by dissolving them in water inside a sonicator. For making device fabrication, the hydropholic nature of the DNA is undesirable and therefore cetyltrimetylammonium (CTMA) as a buffer was introduced to enhance an adhesion between DNA and GaN. The DNA was treated with the CTMA via an ion-exchange reaction. This reaction substitutes CTMA for Na ions associated with the DNA base pairs, resulting in DNA-CTMA and NaCl in aqueous solution. The DNA-CTMA precipitate was captured by a vacuum filtration unit. Generally speaking, DNA-CTMA precipitate is insoluble in water and soluble in organic solvents. Thus, butanol was chosen as a solvent for this process, due to its low vapor pressure and high viscosity essential to make the DNA-CTMA film uniform during the spin coating. The DNA-CTMA granules were mixed in butanol at a 0.25 wt% ratio for 12 hours in stirring. After the stirring process, the DNA-CTMA biopolymer (~150 nm) was spin-coated on another portion of the n-GaN surface. After the formation of the DNA-CTMA layer on the n-GaN, the Au (100 nm) Schottky electrodes with an area of 0.0025 cm2 were found using the e-beam evaporation. Finally, we fabricated the Au/DNA-CTMA/n-GaN photodiode completely. For simplicity, we name the Au/DNA-CTMA/n-GaN photodiode as the DNA-based GaN photodiode (DG-PD). The DG-PD active area is 0.16 cm2. Figure 1 shows the schematic description of DG-PD consisting of Au Schottky and Ti/Al/Ni/Au ohmic electrode. The current-voltage (I-V) characteristics were measured using a semiconductor parameter analyzer (Keithley source measuring unit 4200), and spectral photoresponsivity characteristics were measured using a xenon arc lamp (150 W) with monochromator system (Oriel 74000) and the Newport Low-Power detector. The photocurrent measurements were carried out using a 100 W halogen lamp with a power density of 10 mW/cm2 and a 4 W UV lamp (320-365 nm) having a power density of 7 mW/cm2.
3. Results and discussion
The logarithmic I-V characteristics of the DG-PD (measured at room temperature) were investigated under dark condition, halogen and UV illuminations as shown in Fig. 2.The inset in Fig. 2 shows the logarithmic I-V characteristics of the DG-PD and Au/n-GaN reference diode measured under the dark condition. In comparison with I-V behavior of the reference diode, that of the DG-PD is shifted to negative bias position of 0.78 V. The remarkable shift in the inset of Fig. 2 may be attributed to the formation of reverse molecule charge interaction between DNA-CTMA and n-GaN. Previous investigations of the I-V characteristics in DNA-CTMA/sapphire and DNA-CTMA/glass (not shown here) showed that there is no shift of the I-V behaviors, which is the same for reported DNA-CTMA/Si photodiodes . These differences indicate that the electrically reverse internal field (originating from the formation of reverse molecular charge-induced interlayer between DNA-CTMA and n-GaN) may happen and consequently the negative movement of the I-V curves for DG-PD occurs under dark condition. In Fig. 2, we found that the I-V curves movetoward positive bias side under halogen and UV illumination. Very interestingly, the zero position of the I-V curve for the halogen-illuminated DG-PD is still in negative bias side although the zero position moves by 0.75 V positively. On the other hand, the UV illumination gives rise to further movement toward positive voltage side by 1.02 V and the zero position is in positive voltage region. These behaviors may be associated with unidirectional charge transport mechanisms (depending on the excitation-light wavelength) and different charge interactions between DNA-CTMA and n-GaN, and thus will be described later.
In order to investigate the feasibility for dual-band detection of lights ranging from UV to visible spectra, the measurement for the bias-dependent responsivity characteristics was conducted. Figure 3 shows the relations between responsivity and wavelength according to electrically different biases. Applying a negative bias of −0.5 V into the DG-PD causes negative responsivity, while forward biases (including zero bias) lead to positive values of the responsivity across whole wavelength range as shown in Fig. 3(a)-3(d). In addition, the negative response characteristics are much more dominant in the visible spectral region than the UV region shown in Fig. 3(a). On the other hand, the positive response characteristics are considerably dominant in UV wavelength ranges. These behaviors indicate that the DG-PD can detect the visible lights at a negative bias effectively while the detection of the UV lights can be available at positive biases of ≥ 0 V. From the comparison with Fig. 2, the negatively visible-sensitive responsivity (at below zero bias) and the positively UV-sensitive responsivity (at ≥ 0 V) are virtually consistent with the zero position of halogen-illuminated and UV-illuminated photocurrent in Fig. 2, respectively. These results clearly reveal that the DG-PD can provide the dual detection of lights between visible (in particular, below 620 nm) and UV region.
Figure 4 shows the transient photoresponse characteristics and operation modes of the DG-PDs under switching conditions of halogen and UV irradiation. The insets of Fig. 4 show photoreponse behaviors in milliseconds. Comparisons present that the halogen-illuminated DG-PD has a negative photocurrent of −6.9 × 10−9 A with the corresponding photoresponse of −4.3 × 10−6 A/W shown in Fig. 4(a), while the UV-irradiated DG-PD shows a positive photocurrent of 2 × 10−6 A with the corresponding photoresponse of 1.8 × 10−3 A/W shown in Fig. 4(b). It is noteworthy that the positive photocurrent and photoresponse of the UV-illuminated DG-PD are much larger than those of the halogen-illuminated DG-PD by three orders of magnitude. From the comparisons in Figs. 2-4, it can be reasonable that the zero position of the I-V curves determine a polarity of the photocurrent. The electrically different polarity of photocurrent is directly indicative of dual detection of visible and UV lights. The calculated detectivity of the DG-PDs under the halogen-illumination is 5.27 × 107 cmHz1/2/W (Jones) and the UV-illumination is 2.46 × 108 Jones, respectively.
Based on the I-V data and transient response characteristics, we propose the feasible charge transport mechanisms in DNA-CTMA and then describe the operation modes in the DG-PD. Basically, DNA is a molecular wire that can conduct charged carriers without virtually any resistance, and it has two possible charge transport mechanisms for unidirectional charge migration: one is charge hopping between discrete base orbitals, and the other is tunneling (superexchange) through the DNA-CTMA [9,17–20]. Owing to these characteristics, a virtual bridge between donor and acceptor is believed to be reasonable for interpretation of DNA-based devices [9,17–20]. According to Jortner et al.  in hopping mechanism, the donor and acceptor orbitals are close in energy to the bridge. Thus, in hopping occurs an induced hopping of charges that transiently occupy the bridge like one energy site to the next energy site with shallower distance. Whereas, in domain hopping, the charge transport through longer molecular distances in comparison with the molecular distances is experience in hopping. Finally in tunneling case, the charge could tunnel through the bridge without any hopping phenomenon because the DNA orbitals are energetically higher than the donor and acceptor.
From the standpoint of the charge transport mechanisms in DNA-CTMA from Jortner et al. , the overall operation mechanisms and possible energy band diagrams of the DG-PD under dark, halogen, and UV illumination cases could be explained using Fig. 5.Firstly, in comparison with dark current [Fig. 5(a))], very small amount of negative photocurrent for the halogen-illuminated DG-PD may be related with a charge interaction between charges in the DNA-CTMA and deep-level states on the GaN surface. This charge interaction may cause a slight tilt of energy band in DNA-CTMA [Fig. 5(b)] and hence result in both the generation of photocurrent through the deep-level excitation in the n-GaN [Fig. 5(b)] and the positive movement of the I-V curve as shown in Fig. 2. Eventually, positive charges move toward the Au side through the DNA-CTMA (acting as a hole transport layer), and electrons are very slightly accumulated at the GaN. In other words, the halogen irradiation may result in the transiently physical variations in length and sequence of DNA-CTMA and thus the charge charges are transiently delocalized over the sequence-dependent domains [Fig. 5(b)]. In contrast to hopping mechanisms effective in dark and halogen-incidence condition, the charge transport could be explained as a tunneling of charges (as a superexchange) for the UV illumination. Further positive movement of the I-V curve (Fig. 2) and high amount of the photocurrent for the UV incidence on the DG-PD could be due to the effects combined with the enhancement of band-to-band excitation in the n-GaN and the charge interaction between DNA-CTMA and n-GaN that could give rise to highly tilted energy band-gap of DNA-CTMA and the tunneling of charges through the DNA-CTMA as illustrated from Fig. 5(c). Verifyingthe hopping or the tunneling of charges in DNA-CTMA is very critical to support the charge transport mechanism. To meet this, we fabricated DNA-CTMA on glass and investigated the series resistance under dark, halogen, and UV illumination (not shown here). The measurement showed that the series resistance is determined to be 1.98 × 105, 190, and 28.7 Ω for dark, halogen-, and UV-incident cases, respectively. The significantly decreased series resistance for the UV irradiation case indicates that the tunneling through DNA-CTMA is dominant unlike others. Thus, we can conclude that at lower energy like halogen illumination, the domain hopping occurs, and for higher energy like UV illumination, the tunneling occurs. Finally, due to these reasons explained above, the DG-PD can have a dual-detectable capability for the visible and UV lights.
In summary, we successfully demonstrated how the proposed a DNA-CTMA/n-GaN photodiode can provide a good solution for the dual detection in recognizing visible and ultraviolet lights by means of electrically different polarity and amount of photocurrents. The distinct characteristics could be due to the combined effects of different charge transport mechanisms in the DNA-CTMA (depending on light wavelengths) and charge interactions between DNA-CTMA and n-GaN.
This study was supported by the MOTIE of Korea through the industrial infrastructure program under Grant No: 10033630, and the support from National Research Foundation (NRF) grant through Human Resource Training Project for Regional Innovation.
References and links
1. E. Muñoz, “(Al,In,Ga)N-based photodetectors. Some materials issues,” Phys. Status Solidi, B Basic Res. 244(8), 2859–2877 (2007). [CrossRef]
2. T. Mueller, F. Xia, and P. Avouris, “Graphene photodetectors for high-speed optical communications,” Nat. Photonics 4(5), 297–301 (2010). [CrossRef]
3. X. Wang, Z. Cheng, K. Xu, H. K. Tsang, and J.-B. Xu, “High-responsivity graphene/silicon-heterostructure waveguide photodetectors,” Nat. Photonics 7(11), 888–891 (2013). [CrossRef]
4. J. Wu, Z. Li, D. Shao, M. O. Manasreh, V. P. Kunets, Z. M. Wang, G. J. Salamo, and B. D. Weaver, “Multicolor photodetector based on GaAs quantum rings grown by droplet epitaxy,” Appl. Phys. Lett. 94(17), 171102 (2009). [CrossRef]
5. E. H. Steenbergen, M. J. DiNezza, W. H. G. Dettlaff, S. H. Lim, and Y.-H. Zhang, “Optically-addressed two-terminal multicolor photodetector,” Appl. Phys. Lett. 97(16), 161111 (2010). [CrossRef]
6. Q. Sun, G. Subramanyam, L. Dai, M. Check, A. Campbell, R. Naik, J. Grote, and Y. Wang, “Highly efficient quantum-dot light-emitting diodes with DNA-CTMA as a combined hole-transporting and electron-blocking layer,” ACS Nano 3(3), 737–743 (2009). [CrossRef] [PubMed]
10. J.-S. Jang, “High output power GaN-based light-emitting diodes using an electrically reverse-connected p-Schottky diode and p-InGaN-GaN superlattice,” Appl. Phys. Lett. 93(8), 081118 (2008). [CrossRef]
11. M.-L. Lee, T. S. Mue, F. W. Huang, J. H. Yang, and J. K. Sheu, “High-performance GaN metal-insulator-semiconductor ultraviolet photodetectors using gallium oxide as gate layer,” Opt. Express 19(13), 12658–12663 (2011). [CrossRef] [PubMed]
12. J. Wu, “When group-III nitrides go infrared: New properties and perspectives,” J. Appl. Phys. 106(1), 011101 (2009). [CrossRef]
13. S.-M. Kim, Y.-M. Yu, J.-H. Baek, S.-R. Jeon, H.-J. Ahn, and J.-S. Jang, “Plasma-induced damage influence on the n-contact properties and device performance of ultraviolet InGaN/AlGaN light-emitting diodes,” J. Electrochem. Soc. 154(5), H384–H388 (2007). [CrossRef]
14. J.-S. Jang and T.-Y. Seong, “Mechanisms for the reduction of the Schottky barrier heights of high-quality nonalloyed Pt contacts on surface-treated p-GaN,” J. Appl. Phys. 88(5), 3064–3066 (2000). [CrossRef]
16. R. K. Gupta, A. A. Al-Ghamdi, O. Al-Hartomy, H. Hasar, F. El-Tantawy, and F. Yakuphanoglu, “Series resistance controlling photosensor of Ag/DNA/p-Si/Al diode,” Synth. Met. 162(11–12), 981–987 (2012). [CrossRef]
19. M. Bixon and J. Jortner, “Long-range and very long-range charge transport in DNA,” Chem. Phys. 281(2–3), 393–408 (2002). [CrossRef]
20. J. Jortner, M. Bixton, A. A. Voityuk, and N. Rosch, “Superexchange mediated charge hopping in DNA,” J. Phys. Chem. 106(33), 7599–7606 (2002). [CrossRef]