Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Ag nanoparticle-decorated single wall carbon nanotube films for photovoltaic applications

Ag nanoparticle-decorated single wall carbon nanotube films for photovoltaic applications Mater Renew Sustain Energy (2016) 5:1 DOI 10.1007/s40243-015-0065-6 ORIGINAL PAPER Ag nanoparticle-decorated single wall carbon nanotube films for photovoltaic applications 1 1 1 • • • Mokhtar Anouar Rhanem Jbilat Vincent Le Borgne 1 1 Dongling Ma My Ali El Khakani Received: 16 July 2015 / Accepted: 19 November 2015 / Published online: 9 January 2016 The Author(s) 2016. This article is published with open access at Springerlink.com Abstract We report on the use of pulsed laser deposition Keywords Silver nanoparticles  Single wall carbon to decorate single wall carbon nanotube (SWCNTs) films nanotubes  Photovoltaic  Solar cells  Pulsed laser with Ag nanoparticles (NPs) in hybrid SWCNTs/n-Si deposition photovoltaic (PV) devices. PV devices are built by coating n-Si surfaces with a controlled density of a SWCNTs’ suspension via an air-brush method. By adjusting the Introduction number of laser pulses (N ) from the KrF laser, we are Lp able to control the size and density of the Ag NPs covering Nanomaterial-based photovoltaic (PV) devices are attract- the SWCNTs films. Through adjustment of N , we are ing much attention as promising candidates for the design Lp able to improve the power conversion efficiency of the of third-generation solar cells. Among nanomaterials, sin- SWCNTs/n-Si devices from 3.5 to over 6 % at N = 1250 gle wall carbon nanotubes (SWCNTs) have shown to be a Lp and the corresponding fill factor (FF) from 35 to 60 %. prime candidate for third-generation solar cells and an This increase is shown to be correlated with Raman, important topic of research. The unique qualities of electrical and optical properties of the Ag NPs-coated SWNCTs, such as one-dimensional nanoscale structure, SWCNTs films. UV–Vis spectroscopy measurements show high aspect ratios, large specific surface area, high charge the presence of optical scattering that is directly attributed mobility [1], and excellent optical and electronic properties to the presence of plasmon in the range of 450–600 nm and [2] have led to sustained and ever increasing research internal quantum efficiency measurement shows significant efforts over the past few years. SWCNTs have been inte- improvement over this range. In addition to this direct grated in a variety of photovoltaic devices including sili- increase of the generate photocurrent, the overall Ag NPs con-based [3], polymer-based [4] and dye sensitized solar film resistance is sufficiently lowered to ensure higher FF cells (DSSC) [5], either as a photogenerating medium, as and thus a higher PCE. Beyond the optimal value of transparent electrodes or as scaffolding for other photoac- N [ 1250, we show that the decreasing PCE is caused by tive materials such as TiO . Recently, SWCNTs/n-Si PV Lp 2 low optical transmission of the Ag NPs films and poorer devices have evolved into their own field of research at the rectification. intersection of well-established silicon-based technology and cutting-edge nanomaterial science. These devices typically rely on nanoheterojunctions of n-type silicon and p-type SWCNTs films [6–11]. It has been reported that the transparency and electrical conductivity of the SWCNTs & My Ali El Khakani films play a direct role in the performance of such PV elkhakani@emt.inrs.ca devices [12, 13]. A balance has to be struck between the Vincent Le Borgne transparency of the SWCNTs film, to take full advantage of leborgne@emt.inrs.ca the silicon substrate, and conductivity of the SWCNTs to facilitate charge extraction. Therefore, most of the current Institut National de la Recherche Scientifique, 1650 Lionel research efforts on such SWCNTs/n-Si PV devices focus Boulet, Varennes, Quebec J3X 1S2, Canada 123 1 Page 2 of 9 Mater Renew Sustain Energy (2016) 5:1 on the optimization of the transparency and the conduc- PLD allows us to exert direct control over the size dis- tivity of the SWCNTs films to maximize the performance tribution and density of the Ag NPs within the NHs. We of these PV devices. We have shown in previous work that thus show that the PCE can be optimized by varying the these two parameters can be directly altered through film N . The maximum PCE of NHs PV devices is found to Lp fabrication procedures and combined into a figure of merit be almost two times that of devices built from pristine that correlates with PV properties [3]. In addition to the SWCNTs. This increase arises from improvements in the optimization of the SWCNTs’ films via fabrication proce- fill factor (FF) as well as from plasmon-induced effects. dures, chemical doping, often through the addition of nitric We also show that there appears to be a critical density of acid [7, 9, 14], has been widely used as a method to NPs that leads to a very strong increase of the PCE increase conductivity without significantly altering trans- properties. This is attributed to the presence of a perco- parency. Doping can achieve other desirable results: lated continuous network of large Ag particles that inhi- though SWCNTs are naturally p-doped [15], the addition bits the performance-enhancing quantum properties of Ag of solutions such as SOCl [16]orH O [17] to the NPs. 2 2 2 SWCNTs leads to increased p-doping, which in turn improves the efficiency of the SWCNTs/n-Si junction. Improvement of the performances of the SWCNTs/n-Si Experimental devices has also been achieved by methods that rely on alterations achieved through physical means such as light Preparation of the SWCNTs devices trapping and plasmon enhancement. Indeed, recent work has shown that TiO films deposited by spin-coating and The as-received SWCNTs (Carbon Solution) were sus- used an anti-reflection coating significantly increased the pended in di-methyl-formamide (DMF) at a concentration light absorption of the SWCNTs/n-Si devices [14]. The of 0.1 mg/ml with 30 min sonication in an ultrasonic bath. addition of silver nanowires has also been proposed to SWCNTs films were obtained by spray-coating 1 9 1cm increase the effective conductivity of the SWCNTs films n-Si pieces. To ensure direct contact between SWCNTs without significantly reducing optical transparency [18]. and n-Si, the substrate is treated with dilute HF to remove Similarly, recent work has introduced the idea of adding native oxide immediately before SWCNTs deposition. metal nanoparticles (NPs) to SWCNTs/n-Si. For instance, SWCNTs films were optimized following a previously Ag NPs have been proposed as additional light scattering published procedure that relies on the optimization of a sites, thus increasing the light effectively available for figure of merit (FoM = T /R )[12] where T is the optical photogeneration. Decoration of a solar cell with Ag NPs transmission of the films at 550 nm and R is the sheet -3 -1 has been successfully used in thin film silicon solar cells resistance. Typical values of the FoM were of 10 X . [19, 20], showing a proof of concept in a well-understood SWCNTs films were exposed for 2 s to nitric acid type of solar cell and that this is a viable route to perfor- vapors to dope them and increase the PV response. After mance enhancement. The use of Ag NPs in SWCNTs/PV drying the deposited nitric acid, the samples are inserted in devices has also been investigated: while improvement has a PLD chamber to be decorated with Ag NPs. In this been reported, the maximal reported PCE value of only process, a KrF laser beam (248 nm, 20 ns, 165 mJ/pulse) is 1.32 % [21] is low and likely could be improved upon. focalized onto an Ag target placed in a chamber with a In this work, we demonstrate the potential of pulsed 300 mTorr He background pressure. Ag NPs are deposited laser deposition (PLD) to decorate SWCNTs films with directly onto a variety of substrates including the Ag NPs. PLD is a synthesis method known to produce SWCNTs/n-Si structures as well as bare Si and bare quartz clean and uniform NPs over a surface, without the risk of substrates for parallel characterizations. The SWCNTs/n-Si chemical contamination that arises from chemical meth- structures were systematically decorated with Ag-NPs by ods. It has been employed to decorate CNTs in several varying the number of laser pulses (N ). Lp applications, including field effect emission [22] and optoelectronics [23, 24]. We show that the PLD method Materials’ characterizations yields clean nanohybrids (NHs) and that the SWCNTs films are left structurally intact after the deposition pro- The optical transmission and reflection spectra of Ag NPs, cess. The presence of the Ag surface plasmon in Raman SWCNTs films and Ag NPs-decorated SWCNTs films spectroscopy and UV–Vis optical transmission properties were obtained using a Varian Cary 5000. NPs sizes and is shown directly through Raman and optical transmission SWNCTs films uniformity were characterized through spectroscopy. Finally, the Ag NPs/SWCNTs NHs syn- SEM images (JEOL-6900). Finally, the Raman spectra thesized directly on n-Si substrates were used to fabricate were measured with a 514 Ar laser in a RM3300 Ren- PV devices. The number of laser pulses (N ) used during ishaw system. Lp 123 Mater Renew Sustain Energy (2016) 5:1 Page 3 of 9 1 Device characterization The external quantum efficiency (EQE) spectra of SWCNTs/n-Si devices were obtained using a lock-in amplifier (Ametek). Devices were illuminated by monochromatic light from 150 W Xe lamp chopped at 6 Hz. Lamp power was systematically measured with a calibrated Si-diode (Newport). The EQE was obtained as follows: EQE = 100 9 hc 9 I /kP where P is the inci- ph dent light power at a given wavelength (k), c is the speed of the light, h is Planck constant and I is the photocurrent ph generated by the device. Internal quantum efficiency was obtained by the following equation: IQE = EQE/ (1 - T) where T is the optical transmission of either the Ag NPs/SWCNTs NHs or the pristine SWCNTs films. J–V measurements under solar radiation (Oriel AM 1.5 solar simulator) were performed with an Agilent 6900. All J–V measurements were performed through a circular aperture of 2 mm diameter. Results Materials’ characterization Ag NPs are commonly used as an efficient material for surface-enhanced Raman spectroscopy (SERS). [25] Raman spectroscopy is thus a powerful first-line charac- terization tool for Ag-decorated SWCNTs and can easily detect any Ag-induced alterations. Figure 1 shows Raman spectra for selected values of N , including pristine Lp Fig. 1 a Comparison of the Raman spectra of SWCNTs films SWCNTs (bottom curve) up to N = 2000 (top curve). Lp decorated with Ag NPs for selected values of N . b Positions of the Lp All spectra are measured under identical conditions, which 2D and G peaks as a function of N Lp allow for meaningful comparison of absolute Raman intensity. At N = 0 (i.e., pristine SWCNTs), the Raman Lp spectrum is shown to have an intense G band with respect N = 250 with respect to pristine SWCNTs. This increase Lp to the D band, indicating the high quality of the SWCNTs is attributed to SERS induced by the presence of Ag NPs [26]. The G-band line shape indicates that semiconductor [30]. Starting at N = 500, the absolute Raman intensity Lp SWCNTs are preferentially excited by the Ar laser [27]. decreases, and all signal enhancement has disappeared at -1 The RBM peak position (around 170 cm ) arises from N = 1500. Higher values of N would presumably yield Lp Lp SWCNTs of 1.5 nm diameter [28]. There is a slight larger NPs [24], which would have a gradually less pro- broadening of the RBM peak as the value of N is nounced SERS as a function of N until the Ag film fully Lp Lp increased that is attributed to pressure the large amount of percolates. Thus, when N goes beyond a critical thresh- Lp silver exerts on the SWCNTs wall. The RBM peak disap- old, the Ag NPs film behaves more as continuous films pears entirely at N [ 1250, indicating that beyond this rather than as a collection of NPs, thus negating the effects Lp threshold, the SWCNTs films’ electronic properties could of the NPs’ SERS. be drastically modified [29]. Additionally, the G band On top of SERS altering the Raman intensity, doping- broadens and its two components overlap more. Despite like behavior induced by the presence of Ag NPs can be these electronic changes, the D band, which is proportional observed directly through Raman spectroscopy. Indeed, to the density of defects, varies very little, indicating that starting at N = 1500, a significant shift of the 2D and G Lp the decoration of SWCNTs with Ag NPs does not induce bands was observed [31–34]. The 2D band is shifted to -1 defects. As the value of N is varied between 0 and 2000, lower frequency by 6 cm (Fig. 1b and the G band is Lp -1 intensity of the Raman spectra increases by a factor of 5 at shifted to higher frequency by about 7 cm , both typically 123 1 Page 4 of 9 Mater Renew Sustain Energy (2016) 5:1 signs of n-type doping [35]. Though there are no chemical that the SWCNTs are clean and devoid of contamination bonds between C and Ag, the physical proximity of the Ag from amorphous carbon or other undesirable carbon spe- NPs can induce local effects similar to charge transfer [34] cies, which is in line with the low intensity D band of the and, in turn, n-type doping by making the Ag electrons Raman spectra. In Fig. 2b, the SEM image for Ag NPs available to the SWCNTs. To summarize, Raman spec- deposited with N = 500 shows sparse and rather small Lp troscopy shows that the SWCNTs/Ag NPs NHs are suc- diameters of NPs on the SWCNTs. As N is increased, the Lp cessfully synthesized through the increased intensity of the NPs grow in size and density until the SWCNTs are fully Raman signal, and that they significantly alter the elec- covered (Fig. 2d), corroborating the Raman spectroscopy tronic properties of the SWCNTs at high values of N . results shown in Fig. 1.At N [ 1250, the Ag NPs appear Lp Lp Ag NHs were further investigated through SEM to to form a continuous film over the SWCNTs bundles, assess the deposition efficiency of the PLD method as well mitigating any SERS and effectively acting as a single Ag and NPs size distribution. Figure 2 compares SEM images layer. NPs size and density have been estimated through of NHs for several values of N .At N = 0 (Fig. 2a), SEM images. Indeed, in Fig. 2e, the typical NPs size dis- Lp Lp pristine SWCNTs are seen to form a continuous film of tribution is shown for three values of N (N = 500, Lp Lp interwoven bundles that offer a very large effective area 1250 and 2000) (e). The median size for N = 500 is Lp onto which the NPs can be deposited. It can also be seen found to be about 20 nm and it is pushed up to 40 nm at Fig. 2 SEM images of SWCNTs films decorated with Ag NPs for selected values of N a N = 0, b N = 500, Lp Lp Lp c N = 1250, d N = 2000, Lp Lp e diameter distributions of the Ag NPs 123 Mater Renew Sustain Energy (2016) 5:1 Page 5 of 9 1 N = 2000. This shift is accompanied by an increase in significantly, going from as low as 400 nm for N = 50 to Lp Lp the spread in the NPs due to NP coalescence at high values almost 600 nm for N = 1250. The same measurements Lp of N (i.e., two initially distinct NPs that grow sufficiently were also performed on SWCNTs/Ag-NPs NHs deposited Lp to be conjoined), leading to particles with sizes ranging on quartz. Figure 3b shows the optical transmission spectra between 10 and 90 nm at N = 2000. Overall, SEM for NHs for selected values of N . The absorption bands at Lp Lp images show that both the median particle size and density 1050 and 750 nm are attributed to the S and M transi- 22 11 are correlated with N and that the Ag NPs form contin- tions of the 1.5 nm diameter SWCNTs [36]. The Ag NP Lp uous films at high values of N . plasmon peak is present for most conditions, with positions Lp UV–Vis spectroscopy measurements can provide addi- ranging between 450 and 510 nm. The blue shifted plas- tional information on the presence of plasmon resonance in mon peaks of Ag NPs on SWCNTs seem to indicate that Ag NPs. Figure 3a directly compares the UV–Vis trans- these NPs are significantly smaller than those on deposited mission spectra of Ag NPs deposited on quartz under the on quartz. The high surface area of the SWCNTs film, as same conditions as that of NPs deposited on SWCNTs shown through SEM, provides a much higher effective films. N as low as 50 generates NPs that have a distinct density of nucleation sites onto which Ag can deposit and Lp plasmon peak. At low values of N , the plasmon peak is form NPs. This leads to a higher density of Ag NPs and Lp well defined with little to no loss in the infrared range. thus decreases the number of Ag atoms available for each Conversely, as the number of laser pulses is increased, the NP. In addition to the shift in position, the transmission plasmon peak broadens and red-shifts until the plasmon spectra show that the Ag plasmon peaks are noticeably peak entirely disappears for N = 2000. Over this range weaker in NHs than on quartz despite the higher NP den- Lp of N , the position of the dispersion peak shifts sity. Proximity-induced charge transfer between SWCNTs Lp and Ag could deplete the number of electrons available for surface plasmon generation in the Ag NPs [34, 37]. How- ever, there is no evidence of charge transfer in the positions of the 2D band in Raman spectra, where no trace of doping of the SWCNTs was found for N \ 1500. It seems, Lp therefore, that this quenching of the plasmon peak could be caused by the different medium refractive index. Indeed, the effective plasmon scattering cross section is strongly affected by the medium’s refractive index and the plasmon peak intensity decreases for higher medium refractive index [38]. SWCNTs’ refractive index can be approxi- mated to be similar to that of graphite (n = 2.6) [39] for visible light, which is much higher than that of the quartz used (n = 1.46). This large difference in the medium refractive index could partially account for the observed decrease in intensity. The medium also has a direct effect on the plasmon peak position; a higher refractive index also red-shifts plasmon peaks. At first glance, this appears to be in contradiction with the observation that the NHs plasmon peak is blue shifted when Ag NPs are deposited on SWCNTs with respect to when they are deposited on quartz. However, a blue shift induced by sufficiently small particles can balance the red shift induced by the high refractive index of SWCNTs. Similar to what was observed for on-quartz Ag NPs, the plasmon peak disappears entirely at N [ 1500 with percolation of the Ag NP coating. Lp Finally, no plasmon peak could be measured at low N . Lp This is attributed to a combination of a smaller density of NPs and of plasmon peak quenching. Through UV–Vis spectroscopy, we are able to show that there is a plasmon peak in the visible light range, which is Fig. 3 a UV–Vis transmission spectra of Ag NPs deposited on desirable for improvement of solar PV devices. The effect quartz. b UV–Vis transmission spectra of SWCNTs films decorated of the Ag plasmon can also be directly assessed through with Ag NPs 123 1 Page 6 of 9 Mater Renew Sustain Energy (2016) 5:1 Fig. 4 Comparison of IQE spectra of SWCNTs/n-Si PV devices with and without Ag NPs IQE measurements, in which a localized increase at the plasmon peak’s wavelength would occur if Ag NPs con- tribute directly to photogeneration. On the other hand, if the Ag NPs strictly acted as an anti-reflection coating, the shape of IQE spectra of the SWCNTs/n-Si and AgNPs- SWCNTs/n-Si devices would be very similar but shifted to higher efficiency values after decoration. Figure 4 com- pares the IQE spectra for the SWCNTs-Ag NPs n-Si devices with pristine SWCNTs devices. The IQE of the pristine device is about 20 % in the UV range where it Fig. 5 a J–V curves of PV devices at selected values of N . Lp picks up and increases until the near infrared, where it b Comparison of the evolution of J , FF, V and the PCE as sc oc declines towards 1100 nm, close to the value of the functions of N Lp bandgap of silicon (1.11 eV). Overall, the IQE spectrum shape is similar to that of typical silicon-based devices, but current (J ) value (23 mA/cm ) and low fill factor values sc with a somewhat lower IQE in the near UV and visible (FF = 35 %), which lead to a PCE of 3.5 %. Upon the spectra [12]. This local decrease is attributed to the pres- addition of Ag NPs, PCE increases from a base value of ence of the SWCNTs film that has a strong absorption peak *3.5 % at N = 0, to 6 % at N = 1250 as shown in Lp Lp in the UV (as shown in Fig. 3b). In comparison, the IQE Fig. 5b. Over the same range, J also increases somewhat sc spectrum of the Ag NPs-SWCNTs device starts at about in value but stabilizes at 26 mA/cm . This variation in J sc 30 % and is overall 10 % higher until the IR range where it is too small to be the only driving force behind the increase becomes equivalent to that of the pristine devices. The in PCE. Open circuit voltage (V ) has a somewhat low oc absence of a clearly defined plasmon-induced peak in the value (0.37 V) at N = 0 that is improved upon with the Lp IQE spectra shows that the AgNPs act mostly as an anti- addition of Ag NPs up to 0.41 V, in line with previously reflection layer, trapping a larger number of high energy reported values [11]. FF follows a very similar pattern to photons close to the SWCNTs/n-Si interface. that of PCE, increasing from 35 % to 60 % between N = 0 and N = 1250 (Fig. 5b). Both the PCE and FF Lp Lp PCE optimization decrease rapidly after the optimal point and the PCE plummets down to 2.5 % for higher values of N . Addi- Lp Preliminary IQE measurements have demonstrated that tionally, it can be seen that while J decreases severely, sc there is a net positive effect of the addition of Ag NPs to V remains in the same range. oc the SWCNTs films. Therefore, PCE was optimized as a These measurements clearly show that N = 1250 Lp function of N . Figure 5a compares the J–V curves for Lp leads to an optimal value of PCE. There are several phe- several values of N of NHs. The pristine SWCNTs films Lp nomena that can contribute to PCE optimization: doping devices (N = 0, red circles) show a high short-circuit Lp and plasmonic effects, hot electron injection and electric 123 Mater Renew Sustain Energy (2016) 5:1 Page 7 of 9 1 effects through lowering of the overall film resistance. seen that the resistance decreases rapidly from 1200 to Doping as a means of increasing can effectively be ruled 200 X over the whole range of N , leading to the increase Lp out. Indeed, as shown in the Raman spectra of Fig. 1, little of the FF and thus of the PCE. However, partially coun- to no doping appears until N = 1500 where undesirable teracting this effect, optical transparency of the films Lp n-type doping happens. While this could account for the decreases significantly from 60 to 30 % over the same decrease in PCE at high N , it cannot be the underlying range of N as shown in Fig. 6. In addition to the decrease Lp Lp mechanism of the increase in PCE. Moreover, in addition of the NHs resistance, there is a degradation in the ability to doping effects, plasmon contribution effects should be of the NHs/n-Si junctions to act as a properly rectifying carefully considered. The IQE spectra of Fig. 4 show an diode, thus contributing to the decrease key parameters increased intensity in the near UV that carries over into the such as FF and PCE. Indeed, the reverse bias current (i.e. visible range. This localized increase is attributed to the the I-V curve’s slopes at negative voltages) is more than presence of Ag NPs and their plasmonic dispersion of light, five times smaller than the slopes of devices with low N Lp which increases the chance of photons interacting with the values. This also coincides with the switch from a discrete SWCNTs/n-Si heterojunction. In effect, the integrated IQE NPs Ag film to a continuous Ag NPs film as was evidenced (i.e., the area under the IQE curve) increases by about through UV–Vis Spectroscopy and SEM. Also, Raman 12 %. Given that IQE and J are directly correlated, it spectroscopy showed an n-type doping of the SWCNTs at sc should be expected that this increase in EQE will similarly N [ 1250, through the shifts of the G and 2D bands [38], Lp affect J . Indeed, J is about 13 % higher in SWCNTs/ which would also have in effect to degrade the rectification sc sc Ag-NPs PV devices with respect to pristine SWCNTs PV characteristics of the heterojunction with n-Si. To sum- devices. Finally, plasmon-enhanced hot electron injection marize, the PCE SWCNTs/Ag NPs PV devices appear to possibly contributes to the increase in IQE and J . In this be improved through a combination of NH film plasmonic sc model, the Ag NPs are considered to effectively act as a effects and resistance lowering. These net positive effects semiconductor from which sufficiently excited electrons are, however, eventually counterbalanced by Ag-induced could be injected directly in the SWCNTs films, rendering doping of the SWCNTs that leads to poorer rectification more charges available for collection. The efficiency of this and a decrease of the optical transparency of the Ag NP/ process is known to be low and potentially counts for only SWCNTs film. a small fraction of the J and IQE increase [40]. sc However, the increases in J and IQE can only partially sc explain the more large increase in PCE. As shown in Conclusion Fig. 5b, the main cause of the PCE improvement is the variation of the FF. The latter is directly related to the In conclusion, we report in this paper a straightforward and device properties (e.g., film resistance, contact resistances, clean way to decorate SWCNTs films with Ag NPs to form etc.). The device’s electric properties thus affect PCE more NHs that are directly integrated into PV devices. Using directly than it affects IQE or V . Figure 6 shows the Raman spectroscopy, we can pinpoint the conditions at oc electrical resistance of AgNPs/SWCNTs films. It can be which the Ag NP film percolates and act more as a bulk silver film. We were also able to concomitantly identify this point in optical spectroscopy measurements and cor- relate this with PCE measurements. PCE of SWCNTs/n-Si devices is indeed increased from 3.5 to 6 % through dec- oration of the SWCNTs films with Ag NPs. We show that the PV properties are directly related to the Ag NPs films’ properties and to the various interactions between SWCNTs and Ag. Indeed, there is a combination of factors that contribute to the improvement of PCE. Raman spec- troscopy measurements show that little to know the charge transfer occurs at low values of N but that major charge Lp transfer-like phenomena hamper photogeneration at high N values. However, through optical and electrical char- Lp acterizations, we show that the PCE improvement appears to arise from a synergy between improved optical absorp- tion and electrical conductivity. Future work should focus on numerical modeling of this device to better understand Fig. 6 Film resistance and optical transmission at 550 nm of the importance of each factor. SWCNTs/Ag-NPs hybrids as a function of N Lp 123 1 Page 8 of 9 Mater Renew Sustain Energy (2016) 5:1 Open Access This article is distributed under the terms of the Crea- nanotube-based field effect transistors. Nanotechnology 20(17), tive Commons Attribution 4.0 International License (http:// 175203 (2009) creativecommons.org/licenses/by/4.0/), which permits unrestricted 16. Li, Z., Kunets, V.P., Saini, V., Xu, Y., Dervishi, E., Salamo, G.J., use, distribution, and reproduction in any medium, provided you give Biris, A.R., Biris, A.S.: SOCl enhanced photovoltaic conversion appropriate credit to the original author(s) and the source, provide a link of single wall carbon nanotube/n-silicon heterojunctions. Appl. to the Creative Commons license, and indicate if changes were made. Phys. Lett. 93(24), 243117 (2008) 17. Bai, X., Wang, H.G., Wei, J.Q., Jia, Y., Zhu, H.W., Wang, K.L., Wu, D.H.: Carbon nanotube-silicon hybrid solar cells with hydrogen peroxide doping. Chem. Phys. Lett. 533, 70–73 (2012) References 18. Li, X., Jung, Y., Huang, J.-S., Goh, T., Taylor, A.D.: Device area scale-up and improvement of SWNT/Si solar cells using silver 1. Du¨rkop, T., Getty, S.A., Cobas, E., Fuhrer, M.S.: Extraordinary nanowires. Advanced Energy Materials 4(12), (2014) mobility in semiconducting carbon nanotubes. Nano Lett. 4(1), 19. Schaadt, D.M., Feng, B., Yu, E.T.: Enhanced semiconductor 35–39 (2004) optical absorption via surface plasmon excitation in metal 2. Hecht, D.S., Hu, L., Irvin, G.: Emerging transparent electrodes nanoparticles. Appl. Phys. Lett. 86, 063106 (2005) based on thin films of carbon nanotubes, graphene, and metallic 20. Chen, X., Jia, B., Saha, J.K., Cai, B., Stokes, N., Qiao, Q., Wang, nanostructures. Adv. Mater. 23(13), 1482–1513 (2011) Y., Shi, Z., Gu, M.: Broadband enhancement in thin-film amor- 3. Le Borgne, V., Gautier, L.A., El Khakani, M.A.: Figure of merit phous silicon solar cells enabled by nucleated silver nanoparti- based maximization of the quantum efficiency of SWCNTs/n-Si cles. Nano Lett. 12(5), 2187–2192 (2012) photovoltaic devices. Appl. Phys. Lett. 103, 073103 (2013) 21. Chen, L., Zhang, S., Chang, L., Zeng, L., Yu, X., Zhao, J., Zhao, 4. Ham, M.-H., Paulus, G.L.C., Lee, C.Y., Song, C., Kalantar- S., Xu, C.: Photovoltaic conversion enhancement of single wall Zadeh, K., Choi, W., Han, J.-H., Strano, M.S.: Evidence for high- carbon-Si heterojunction solar cell decorated with Ag nanopar- efficiency exciton dissociation at polymer/single-walled carbon ticles. Electrochim. Acta 93, 293–300 (2013) nanotube interfaces in planar nano-heterojunction photovoltaics. 22. Gautier, L.-A., Le Borgne, V., Delegan, N., Pandiyan, R., El ACS Nano 4(10), 6251–6259 (2010) Khakani, M.A.: Field electron emission enhancement of 5. Park, J.-G., Akhtar, M.S., Li, Z.Y., Cho, D.-S., Lee, W., Yang, graphenated MWCNTs emitters following their decoration with O.-B.: Application of single walled carbon nanotubes as counter Au nanoparticles by a pulsed laser ablation process. Nanotech- electrode for dye sensitized solar cells. Electrochim. Acta 85, nology 26(4), 045706 (2015) 600–604 (2012) 23. Ka, I., Le Borgne, V., Ma, D., El Khakani, M.A.: Pulsed laser 6. Ong, P.-L., Euler, W.B., Levitsky, I.A.: Hybrid solar cells based ablation based direct synthesis of single-wall carbon nanotubes/ on single-walled carbon nanotubes/Si heterojunctions. Nan- PbS-quantum-dots nanohybrids exhibiting strong, spectrally wide otechnology 21(10), 105203 (2010) and fast photoresponse. Adv. Mater. 24, 6289–6294 (2012) 7. Jung, Y., Li, X., Rajan, N.K., Taylor, A.D., Reed, M.A.: Record 24. Ka, I., Ma, D., El Khakani, M.A.: Tailoring the photoluminescence high efficiency single-walled carbon nanotube/silicon p–n junc- of PbS-nanoparticles layers deposited by means of the pulsed laser tion solar cells. Nano Lett. 13(1), 95–99 (2013) ablation technique. J. Nanopart. Res. 13(6), 2269–2274 (2011) 8. Kozawa, D., Hiraoka, K., Miyauchi, Y., Mouri, S., Matsuda, K.: 25. Moskovits, M.: Surface-enhanced Raman spectroscopy: a brief Analysis of the photovoltaic properties of single-walled carbon retrospective. J. Raman Spectrosc. 36(6–7), 485–496 (2005) nanotube/silicon heterojunction solar cells. Appl. Phys. Express 26. Rinzler, A.G., Liu, J., Dai, H., Nikolaev, P., Huffman, C.B., 5(4), 42304 (2012) Rodriguez-Macı´as, F.J., Boul, P.J., Lu, A.H., Heymann, D., 9. Jia, Y., Li, P., Gui, X., Wei, J., Wang, K., Zhu, H., Wu, D., Colbert, D.T., Lee, R.S., Fischer, J.E., Rao, A.M., Eklund, P.C., Zhang, L., Cao, A., Xu, Y.: Encapsulated carbon nanotube-oxide- Smalley, R.E.: Large-scale purification of single-wall carbon silicon solar cells with stable 10 % efficiency. Appl. Phys. Lett. nanotubes: process, product, and characterization. Appl. Phys. 98(13), 133115 (2011) A Mater. Sci. Process. 67(1), 29–37 (1998) 10. Tune, D.D., Flavel, B.S., Krupke, R., Shapter, J.G.: Carbon 27. Kataura, H., Kumazawa, Y., Maniwa, Y., Umezu, I., Suzuki, S., nanotube-silicon solar cells. Adv. Energy Mater. 2(9), 1043–1055 Ohtsuka, Y., Achiba, Y.: Optical properties of single-wall carbon (2012) nanotubes. Synth. Met. 103(1–3), 2555–2558 (1999) 11. Tune, D.D., Blanch, A.J., Krupke, R., Flavel, B.S., Shapter, J.G.: 28. Bachilo, S.M., Strano, M.S., Kittrell, C., Hauge, R.H., Smalley, Nanotube film metallicity and its effect on the performance of R.E., Weisman, R.B.: Structure-assigned optical spectra of sin- carbon nanotube-silicon solar cells. Phys. Status Solidi A 211(7), gle-walled carbon nanotubes. Science 298(5602), 2361–2366 1479–1487 (2014) (2002) 12. Le Borgne, V., Gautier, L.A., Castrucci, P., Del Gobbo, S., De 29. Valmalette, J.-C., Tan, Z., Abe, H., Ohara, S.: Raman scattering Crescenzi, M., El Khakani, M.A.: Enhanced UV photoresponse of linear chains of strongly coupled Ag nanoparticles on of KrF-laser-synthesized single-wall carbon nanotubes/n-silicon SWCNTs. Sci. Rep. 4, 5238 (2014) hybrid photovoltaic devices. Nanotechnology 23(21), 215206 30. Kneipp, K., Kneipp, H., Itzkan, I., Dasari, R.R., Feld, M.S.: (2012) Ultrasensitive chemical analysis by Raman spectroscopy. Chem. 13. Li, Z., Kunets, V.P., Saini, V., Xu, Y., Dervishi, E., Salamo, G.J., Rev. 99, 2957–2976 (1999) Biris, A.R., Biris, A.S.: Light-harvesting using high density 31. Kharlamova, M.V., Niu, J.J.: Donor doping of single-walled p-type single wall carbon nanotube/n-type silicon heterojunc- carbon nanotubes by filling of channels with silver. J. Exp. Theor. tions. ACS Nano 3(6), 1407–1414 (2009) Phys. 115(3), 485–491 (2012) 14. Shi, E., Zhang, L., Li, Z., Li, P., Shang, Y., Jia, Y., Wei, J., Wang, 32. Niu, Z.Q., Fang, Y.: Surface-enhanced Raman scattering of sin- K., Zhu, H., Wu, D., Zhang, S., Cao, A.: TiO -coated carbon gle-walled carbon nanotubes on silver-coated and gold-coated nanotube-silicon solar cells with efficiency of 15%. Sci. Rep. 2, filter paper. J. Colloid Interface Sci. 303(1), 224–228 (2006) 884 (2012) 33. Tsang, J.C., Freitag, M., Perebeinos, V., Liu, J., Avouris, Ph: 15. Aı¨ssa, B., El Khakani, M.A.: The channel length effect on the Doping and phonon renormalization in carbon nanotubes. Nat. electrical performance of suspended-single-wall-carbon- Nanotechnol. 2(11), 725–730 (2007) 123 Mater Renew Sustain Energy (2016) 5:1 Page 9 of 9 1 34. Lin, Y., Watson, K.A., Fallbach, M.J., Ghose, S., Smith, J.G., 37. Stapleton, A.J., Afre, R.A., Ellis, A.V., Shapter, J.G., Andersson, Delozier, D.M., Cao, W., Crooks, R.E., Connell, J.W.: Rapid, G.G., Quinton, J.S., Lewis, D.A.: Highly conductive interwoven solventless, bulk preparation of metal nanoparticle-decorated carbon nanotube and silver nanowire transparent electrodes. Sci. carbon nanotubes. ACS Nano 3(4), 871–884 (2009) Technol. Adv. Mater. 14(3), 035004 (2013) 35. Maciel, I.O., Campos-Delgado, J., Cruz-Silva, E., Pimenta, M.A., 38. Kelly, K.L., Coronado, E., Zhao, L.L., Schatz, G.C.: The optical Sumpter, B.G., Meunier, V., Lo´pez-Urı´as, F., Mun˜oz-Sandoval, properties of metal nanoparticles: the influence of size, shape, and E., Terrones, H., Terrones, M., Jorio, A.: Synthesis, electronic dielectric environment. Journal of Physical Chemistry B 107(3), structure, and Raman scattering of phosphorus-doped single-wall 668–677 (2003) carbon nanotubes. Nano Lett. 9(6), 2267–2272 (2009) 39. Djurisic, A.B., Li, H.: Optical properties of graphite. J. Appl. 36. Itkis, M.E., Perea, D.E., Niyogi, S., Rickard, S.M., Hamon, Phys. 85(10), 7404–7410 (1999) M.A., Hu, H., Zhao, B., Haddon, R.C.: Purity evaluation of as- 40. Clavero, C.: Plasmon-induced hot-electron generation at prepared single-walled carbon nanotube soot by use of solu- nanoparticle/metal-oxide interfaces for photovoltaic and photo- tion-phase near-IR spectroscopy. Nano Lett. 3(3), 309–314 catalytic devices. Nat. Photonics 8(2), 95–103 (2014) (2003) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Materials for Renewable and Sustainable Energy Springer Journals

Ag nanoparticle-decorated single wall carbon nanotube films for photovoltaic applications

Loading next page...
 
/lp/springer-journals/ag-nanoparticle-decorated-single-wall-carbon-nanotube-films-for-GVB2rCDDwz
Publisher
Springer Journals
Copyright
Copyright © 2016 by The Author(s)
Subject
Material Science; Materials Science, general; Renewable and Green Energy; Renewable and Green Energy
ISSN
2194-1459
eISSN
2194-1467
DOI
10.1007/s40243-015-0065-6
Publisher site
See Article on Publisher Site

Abstract

Mater Renew Sustain Energy (2016) 5:1 DOI 10.1007/s40243-015-0065-6 ORIGINAL PAPER Ag nanoparticle-decorated single wall carbon nanotube films for photovoltaic applications 1 1 1 • • • Mokhtar Anouar Rhanem Jbilat Vincent Le Borgne 1 1 Dongling Ma My Ali El Khakani Received: 16 July 2015 / Accepted: 19 November 2015 / Published online: 9 January 2016 The Author(s) 2016. This article is published with open access at Springerlink.com Abstract We report on the use of pulsed laser deposition Keywords Silver nanoparticles  Single wall carbon to decorate single wall carbon nanotube (SWCNTs) films nanotubes  Photovoltaic  Solar cells  Pulsed laser with Ag nanoparticles (NPs) in hybrid SWCNTs/n-Si deposition photovoltaic (PV) devices. PV devices are built by coating n-Si surfaces with a controlled density of a SWCNTs’ suspension via an air-brush method. By adjusting the Introduction number of laser pulses (N ) from the KrF laser, we are Lp able to control the size and density of the Ag NPs covering Nanomaterial-based photovoltaic (PV) devices are attract- the SWCNTs films. Through adjustment of N , we are ing much attention as promising candidates for the design Lp able to improve the power conversion efficiency of the of third-generation solar cells. Among nanomaterials, sin- SWCNTs/n-Si devices from 3.5 to over 6 % at N = 1250 gle wall carbon nanotubes (SWCNTs) have shown to be a Lp and the corresponding fill factor (FF) from 35 to 60 %. prime candidate for third-generation solar cells and an This increase is shown to be correlated with Raman, important topic of research. The unique qualities of electrical and optical properties of the Ag NPs-coated SWNCTs, such as one-dimensional nanoscale structure, SWCNTs films. UV–Vis spectroscopy measurements show high aspect ratios, large specific surface area, high charge the presence of optical scattering that is directly attributed mobility [1], and excellent optical and electronic properties to the presence of plasmon in the range of 450–600 nm and [2] have led to sustained and ever increasing research internal quantum efficiency measurement shows significant efforts over the past few years. SWCNTs have been inte- improvement over this range. In addition to this direct grated in a variety of photovoltaic devices including sili- increase of the generate photocurrent, the overall Ag NPs con-based [3], polymer-based [4] and dye sensitized solar film resistance is sufficiently lowered to ensure higher FF cells (DSSC) [5], either as a photogenerating medium, as and thus a higher PCE. Beyond the optimal value of transparent electrodes or as scaffolding for other photoac- N [ 1250, we show that the decreasing PCE is caused by tive materials such as TiO . Recently, SWCNTs/n-Si PV Lp 2 low optical transmission of the Ag NPs films and poorer devices have evolved into their own field of research at the rectification. intersection of well-established silicon-based technology and cutting-edge nanomaterial science. These devices typically rely on nanoheterojunctions of n-type silicon and p-type SWCNTs films [6–11]. It has been reported that the transparency and electrical conductivity of the SWCNTs & My Ali El Khakani films play a direct role in the performance of such PV elkhakani@emt.inrs.ca devices [12, 13]. A balance has to be struck between the Vincent Le Borgne transparency of the SWCNTs film, to take full advantage of leborgne@emt.inrs.ca the silicon substrate, and conductivity of the SWCNTs to facilitate charge extraction. Therefore, most of the current Institut National de la Recherche Scientifique, 1650 Lionel research efforts on such SWCNTs/n-Si PV devices focus Boulet, Varennes, Quebec J3X 1S2, Canada 123 1 Page 2 of 9 Mater Renew Sustain Energy (2016) 5:1 on the optimization of the transparency and the conduc- PLD allows us to exert direct control over the size dis- tivity of the SWCNTs films to maximize the performance tribution and density of the Ag NPs within the NHs. We of these PV devices. We have shown in previous work that thus show that the PCE can be optimized by varying the these two parameters can be directly altered through film N . The maximum PCE of NHs PV devices is found to Lp fabrication procedures and combined into a figure of merit be almost two times that of devices built from pristine that correlates with PV properties [3]. In addition to the SWCNTs. This increase arises from improvements in the optimization of the SWCNTs’ films via fabrication proce- fill factor (FF) as well as from plasmon-induced effects. dures, chemical doping, often through the addition of nitric We also show that there appears to be a critical density of acid [7, 9, 14], has been widely used as a method to NPs that leads to a very strong increase of the PCE increase conductivity without significantly altering trans- properties. This is attributed to the presence of a perco- parency. Doping can achieve other desirable results: lated continuous network of large Ag particles that inhi- though SWCNTs are naturally p-doped [15], the addition bits the performance-enhancing quantum properties of Ag of solutions such as SOCl [16]orH O [17] to the NPs. 2 2 2 SWCNTs leads to increased p-doping, which in turn improves the efficiency of the SWCNTs/n-Si junction. Improvement of the performances of the SWCNTs/n-Si Experimental devices has also been achieved by methods that rely on alterations achieved through physical means such as light Preparation of the SWCNTs devices trapping and plasmon enhancement. Indeed, recent work has shown that TiO films deposited by spin-coating and The as-received SWCNTs (Carbon Solution) were sus- used an anti-reflection coating significantly increased the pended in di-methyl-formamide (DMF) at a concentration light absorption of the SWCNTs/n-Si devices [14]. The of 0.1 mg/ml with 30 min sonication in an ultrasonic bath. addition of silver nanowires has also been proposed to SWCNTs films were obtained by spray-coating 1 9 1cm increase the effective conductivity of the SWCNTs films n-Si pieces. To ensure direct contact between SWCNTs without significantly reducing optical transparency [18]. and n-Si, the substrate is treated with dilute HF to remove Similarly, recent work has introduced the idea of adding native oxide immediately before SWCNTs deposition. metal nanoparticles (NPs) to SWCNTs/n-Si. For instance, SWCNTs films were optimized following a previously Ag NPs have been proposed as additional light scattering published procedure that relies on the optimization of a sites, thus increasing the light effectively available for figure of merit (FoM = T /R )[12] where T is the optical photogeneration. Decoration of a solar cell with Ag NPs transmission of the films at 550 nm and R is the sheet -3 -1 has been successfully used in thin film silicon solar cells resistance. Typical values of the FoM were of 10 X . [19, 20], showing a proof of concept in a well-understood SWCNTs films were exposed for 2 s to nitric acid type of solar cell and that this is a viable route to perfor- vapors to dope them and increase the PV response. After mance enhancement. The use of Ag NPs in SWCNTs/PV drying the deposited nitric acid, the samples are inserted in devices has also been investigated: while improvement has a PLD chamber to be decorated with Ag NPs. In this been reported, the maximal reported PCE value of only process, a KrF laser beam (248 nm, 20 ns, 165 mJ/pulse) is 1.32 % [21] is low and likely could be improved upon. focalized onto an Ag target placed in a chamber with a In this work, we demonstrate the potential of pulsed 300 mTorr He background pressure. Ag NPs are deposited laser deposition (PLD) to decorate SWCNTs films with directly onto a variety of substrates including the Ag NPs. PLD is a synthesis method known to produce SWCNTs/n-Si structures as well as bare Si and bare quartz clean and uniform NPs over a surface, without the risk of substrates for parallel characterizations. The SWCNTs/n-Si chemical contamination that arises from chemical meth- structures were systematically decorated with Ag-NPs by ods. It has been employed to decorate CNTs in several varying the number of laser pulses (N ). Lp applications, including field effect emission [22] and optoelectronics [23, 24]. We show that the PLD method Materials’ characterizations yields clean nanohybrids (NHs) and that the SWCNTs films are left structurally intact after the deposition pro- The optical transmission and reflection spectra of Ag NPs, cess. The presence of the Ag surface plasmon in Raman SWCNTs films and Ag NPs-decorated SWCNTs films spectroscopy and UV–Vis optical transmission properties were obtained using a Varian Cary 5000. NPs sizes and is shown directly through Raman and optical transmission SWNCTs films uniformity were characterized through spectroscopy. Finally, the Ag NPs/SWCNTs NHs syn- SEM images (JEOL-6900). Finally, the Raman spectra thesized directly on n-Si substrates were used to fabricate were measured with a 514 Ar laser in a RM3300 Ren- PV devices. The number of laser pulses (N ) used during ishaw system. Lp 123 Mater Renew Sustain Energy (2016) 5:1 Page 3 of 9 1 Device characterization The external quantum efficiency (EQE) spectra of SWCNTs/n-Si devices were obtained using a lock-in amplifier (Ametek). Devices were illuminated by monochromatic light from 150 W Xe lamp chopped at 6 Hz. Lamp power was systematically measured with a calibrated Si-diode (Newport). The EQE was obtained as follows: EQE = 100 9 hc 9 I /kP where P is the inci- ph dent light power at a given wavelength (k), c is the speed of the light, h is Planck constant and I is the photocurrent ph generated by the device. Internal quantum efficiency was obtained by the following equation: IQE = EQE/ (1 - T) where T is the optical transmission of either the Ag NPs/SWCNTs NHs or the pristine SWCNTs films. J–V measurements under solar radiation (Oriel AM 1.5 solar simulator) were performed with an Agilent 6900. All J–V measurements were performed through a circular aperture of 2 mm diameter. Results Materials’ characterization Ag NPs are commonly used as an efficient material for surface-enhanced Raman spectroscopy (SERS). [25] Raman spectroscopy is thus a powerful first-line charac- terization tool for Ag-decorated SWCNTs and can easily detect any Ag-induced alterations. Figure 1 shows Raman spectra for selected values of N , including pristine Lp Fig. 1 a Comparison of the Raman spectra of SWCNTs films SWCNTs (bottom curve) up to N = 2000 (top curve). Lp decorated with Ag NPs for selected values of N . b Positions of the Lp All spectra are measured under identical conditions, which 2D and G peaks as a function of N Lp allow for meaningful comparison of absolute Raman intensity. At N = 0 (i.e., pristine SWCNTs), the Raman Lp spectrum is shown to have an intense G band with respect N = 250 with respect to pristine SWCNTs. This increase Lp to the D band, indicating the high quality of the SWCNTs is attributed to SERS induced by the presence of Ag NPs [26]. The G-band line shape indicates that semiconductor [30]. Starting at N = 500, the absolute Raman intensity Lp SWCNTs are preferentially excited by the Ar laser [27]. decreases, and all signal enhancement has disappeared at -1 The RBM peak position (around 170 cm ) arises from N = 1500. Higher values of N would presumably yield Lp Lp SWCNTs of 1.5 nm diameter [28]. There is a slight larger NPs [24], which would have a gradually less pro- broadening of the RBM peak as the value of N is nounced SERS as a function of N until the Ag film fully Lp Lp increased that is attributed to pressure the large amount of percolates. Thus, when N goes beyond a critical thresh- Lp silver exerts on the SWCNTs wall. The RBM peak disap- old, the Ag NPs film behaves more as continuous films pears entirely at N [ 1250, indicating that beyond this rather than as a collection of NPs, thus negating the effects Lp threshold, the SWCNTs films’ electronic properties could of the NPs’ SERS. be drastically modified [29]. Additionally, the G band On top of SERS altering the Raman intensity, doping- broadens and its two components overlap more. Despite like behavior induced by the presence of Ag NPs can be these electronic changes, the D band, which is proportional observed directly through Raman spectroscopy. Indeed, to the density of defects, varies very little, indicating that starting at N = 1500, a significant shift of the 2D and G Lp the decoration of SWCNTs with Ag NPs does not induce bands was observed [31–34]. The 2D band is shifted to -1 defects. As the value of N is varied between 0 and 2000, lower frequency by 6 cm (Fig. 1b and the G band is Lp -1 intensity of the Raman spectra increases by a factor of 5 at shifted to higher frequency by about 7 cm , both typically 123 1 Page 4 of 9 Mater Renew Sustain Energy (2016) 5:1 signs of n-type doping [35]. Though there are no chemical that the SWCNTs are clean and devoid of contamination bonds between C and Ag, the physical proximity of the Ag from amorphous carbon or other undesirable carbon spe- NPs can induce local effects similar to charge transfer [34] cies, which is in line with the low intensity D band of the and, in turn, n-type doping by making the Ag electrons Raman spectra. In Fig. 2b, the SEM image for Ag NPs available to the SWCNTs. To summarize, Raman spec- deposited with N = 500 shows sparse and rather small Lp troscopy shows that the SWCNTs/Ag NPs NHs are suc- diameters of NPs on the SWCNTs. As N is increased, the Lp cessfully synthesized through the increased intensity of the NPs grow in size and density until the SWCNTs are fully Raman signal, and that they significantly alter the elec- covered (Fig. 2d), corroborating the Raman spectroscopy tronic properties of the SWCNTs at high values of N . results shown in Fig. 1.At N [ 1250, the Ag NPs appear Lp Lp Ag NHs were further investigated through SEM to to form a continuous film over the SWCNTs bundles, assess the deposition efficiency of the PLD method as well mitigating any SERS and effectively acting as a single Ag and NPs size distribution. Figure 2 compares SEM images layer. NPs size and density have been estimated through of NHs for several values of N .At N = 0 (Fig. 2a), SEM images. Indeed, in Fig. 2e, the typical NPs size dis- Lp Lp pristine SWCNTs are seen to form a continuous film of tribution is shown for three values of N (N = 500, Lp Lp interwoven bundles that offer a very large effective area 1250 and 2000) (e). The median size for N = 500 is Lp onto which the NPs can be deposited. It can also be seen found to be about 20 nm and it is pushed up to 40 nm at Fig. 2 SEM images of SWCNTs films decorated with Ag NPs for selected values of N a N = 0, b N = 500, Lp Lp Lp c N = 1250, d N = 2000, Lp Lp e diameter distributions of the Ag NPs 123 Mater Renew Sustain Energy (2016) 5:1 Page 5 of 9 1 N = 2000. This shift is accompanied by an increase in significantly, going from as low as 400 nm for N = 50 to Lp Lp the spread in the NPs due to NP coalescence at high values almost 600 nm for N = 1250. The same measurements Lp of N (i.e., two initially distinct NPs that grow sufficiently were also performed on SWCNTs/Ag-NPs NHs deposited Lp to be conjoined), leading to particles with sizes ranging on quartz. Figure 3b shows the optical transmission spectra between 10 and 90 nm at N = 2000. Overall, SEM for NHs for selected values of N . The absorption bands at Lp Lp images show that both the median particle size and density 1050 and 750 nm are attributed to the S and M transi- 22 11 are correlated with N and that the Ag NPs form contin- tions of the 1.5 nm diameter SWCNTs [36]. The Ag NP Lp uous films at high values of N . plasmon peak is present for most conditions, with positions Lp UV–Vis spectroscopy measurements can provide addi- ranging between 450 and 510 nm. The blue shifted plas- tional information on the presence of plasmon resonance in mon peaks of Ag NPs on SWCNTs seem to indicate that Ag NPs. Figure 3a directly compares the UV–Vis trans- these NPs are significantly smaller than those on deposited mission spectra of Ag NPs deposited on quartz under the on quartz. The high surface area of the SWCNTs film, as same conditions as that of NPs deposited on SWCNTs shown through SEM, provides a much higher effective films. N as low as 50 generates NPs that have a distinct density of nucleation sites onto which Ag can deposit and Lp plasmon peak. At low values of N , the plasmon peak is form NPs. This leads to a higher density of Ag NPs and Lp well defined with little to no loss in the infrared range. thus decreases the number of Ag atoms available for each Conversely, as the number of laser pulses is increased, the NP. In addition to the shift in position, the transmission plasmon peak broadens and red-shifts until the plasmon spectra show that the Ag plasmon peaks are noticeably peak entirely disappears for N = 2000. Over this range weaker in NHs than on quartz despite the higher NP den- Lp of N , the position of the dispersion peak shifts sity. Proximity-induced charge transfer between SWCNTs Lp and Ag could deplete the number of electrons available for surface plasmon generation in the Ag NPs [34, 37]. How- ever, there is no evidence of charge transfer in the positions of the 2D band in Raman spectra, where no trace of doping of the SWCNTs was found for N \ 1500. It seems, Lp therefore, that this quenching of the plasmon peak could be caused by the different medium refractive index. Indeed, the effective plasmon scattering cross section is strongly affected by the medium’s refractive index and the plasmon peak intensity decreases for higher medium refractive index [38]. SWCNTs’ refractive index can be approxi- mated to be similar to that of graphite (n = 2.6) [39] for visible light, which is much higher than that of the quartz used (n = 1.46). This large difference in the medium refractive index could partially account for the observed decrease in intensity. The medium also has a direct effect on the plasmon peak position; a higher refractive index also red-shifts plasmon peaks. At first glance, this appears to be in contradiction with the observation that the NHs plasmon peak is blue shifted when Ag NPs are deposited on SWCNTs with respect to when they are deposited on quartz. However, a blue shift induced by sufficiently small particles can balance the red shift induced by the high refractive index of SWCNTs. Similar to what was observed for on-quartz Ag NPs, the plasmon peak disappears entirely at N [ 1500 with percolation of the Ag NP coating. Lp Finally, no plasmon peak could be measured at low N . Lp This is attributed to a combination of a smaller density of NPs and of plasmon peak quenching. Through UV–Vis spectroscopy, we are able to show that there is a plasmon peak in the visible light range, which is Fig. 3 a UV–Vis transmission spectra of Ag NPs deposited on desirable for improvement of solar PV devices. The effect quartz. b UV–Vis transmission spectra of SWCNTs films decorated of the Ag plasmon can also be directly assessed through with Ag NPs 123 1 Page 6 of 9 Mater Renew Sustain Energy (2016) 5:1 Fig. 4 Comparison of IQE spectra of SWCNTs/n-Si PV devices with and without Ag NPs IQE measurements, in which a localized increase at the plasmon peak’s wavelength would occur if Ag NPs con- tribute directly to photogeneration. On the other hand, if the Ag NPs strictly acted as an anti-reflection coating, the shape of IQE spectra of the SWCNTs/n-Si and AgNPs- SWCNTs/n-Si devices would be very similar but shifted to higher efficiency values after decoration. Figure 4 com- pares the IQE spectra for the SWCNTs-Ag NPs n-Si devices with pristine SWCNTs devices. The IQE of the pristine device is about 20 % in the UV range where it Fig. 5 a J–V curves of PV devices at selected values of N . Lp picks up and increases until the near infrared, where it b Comparison of the evolution of J , FF, V and the PCE as sc oc declines towards 1100 nm, close to the value of the functions of N Lp bandgap of silicon (1.11 eV). Overall, the IQE spectrum shape is similar to that of typical silicon-based devices, but current (J ) value (23 mA/cm ) and low fill factor values sc with a somewhat lower IQE in the near UV and visible (FF = 35 %), which lead to a PCE of 3.5 %. Upon the spectra [12]. This local decrease is attributed to the pres- addition of Ag NPs, PCE increases from a base value of ence of the SWCNTs film that has a strong absorption peak *3.5 % at N = 0, to 6 % at N = 1250 as shown in Lp Lp in the UV (as shown in Fig. 3b). In comparison, the IQE Fig. 5b. Over the same range, J also increases somewhat sc spectrum of the Ag NPs-SWCNTs device starts at about in value but stabilizes at 26 mA/cm . This variation in J sc 30 % and is overall 10 % higher until the IR range where it is too small to be the only driving force behind the increase becomes equivalent to that of the pristine devices. The in PCE. Open circuit voltage (V ) has a somewhat low oc absence of a clearly defined plasmon-induced peak in the value (0.37 V) at N = 0 that is improved upon with the Lp IQE spectra shows that the AgNPs act mostly as an anti- addition of Ag NPs up to 0.41 V, in line with previously reflection layer, trapping a larger number of high energy reported values [11]. FF follows a very similar pattern to photons close to the SWCNTs/n-Si interface. that of PCE, increasing from 35 % to 60 % between N = 0 and N = 1250 (Fig. 5b). Both the PCE and FF Lp Lp PCE optimization decrease rapidly after the optimal point and the PCE plummets down to 2.5 % for higher values of N . Addi- Lp Preliminary IQE measurements have demonstrated that tionally, it can be seen that while J decreases severely, sc there is a net positive effect of the addition of Ag NPs to V remains in the same range. oc the SWCNTs films. Therefore, PCE was optimized as a These measurements clearly show that N = 1250 Lp function of N . Figure 5a compares the J–V curves for Lp leads to an optimal value of PCE. There are several phe- several values of N of NHs. The pristine SWCNTs films Lp nomena that can contribute to PCE optimization: doping devices (N = 0, red circles) show a high short-circuit Lp and plasmonic effects, hot electron injection and electric 123 Mater Renew Sustain Energy (2016) 5:1 Page 7 of 9 1 effects through lowering of the overall film resistance. seen that the resistance decreases rapidly from 1200 to Doping as a means of increasing can effectively be ruled 200 X over the whole range of N , leading to the increase Lp out. Indeed, as shown in the Raman spectra of Fig. 1, little of the FF and thus of the PCE. However, partially coun- to no doping appears until N = 1500 where undesirable teracting this effect, optical transparency of the films Lp n-type doping happens. While this could account for the decreases significantly from 60 to 30 % over the same decrease in PCE at high N , it cannot be the underlying range of N as shown in Fig. 6. In addition to the decrease Lp Lp mechanism of the increase in PCE. Moreover, in addition of the NHs resistance, there is a degradation in the ability to doping effects, plasmon contribution effects should be of the NHs/n-Si junctions to act as a properly rectifying carefully considered. The IQE spectra of Fig. 4 show an diode, thus contributing to the decrease key parameters increased intensity in the near UV that carries over into the such as FF and PCE. Indeed, the reverse bias current (i.e. visible range. This localized increase is attributed to the the I-V curve’s slopes at negative voltages) is more than presence of Ag NPs and their plasmonic dispersion of light, five times smaller than the slopes of devices with low N Lp which increases the chance of photons interacting with the values. This also coincides with the switch from a discrete SWCNTs/n-Si heterojunction. In effect, the integrated IQE NPs Ag film to a continuous Ag NPs film as was evidenced (i.e., the area under the IQE curve) increases by about through UV–Vis Spectroscopy and SEM. Also, Raman 12 %. Given that IQE and J are directly correlated, it spectroscopy showed an n-type doping of the SWCNTs at sc should be expected that this increase in EQE will similarly N [ 1250, through the shifts of the G and 2D bands [38], Lp affect J . Indeed, J is about 13 % higher in SWCNTs/ which would also have in effect to degrade the rectification sc sc Ag-NPs PV devices with respect to pristine SWCNTs PV characteristics of the heterojunction with n-Si. To sum- devices. Finally, plasmon-enhanced hot electron injection marize, the PCE SWCNTs/Ag NPs PV devices appear to possibly contributes to the increase in IQE and J . In this be improved through a combination of NH film plasmonic sc model, the Ag NPs are considered to effectively act as a effects and resistance lowering. These net positive effects semiconductor from which sufficiently excited electrons are, however, eventually counterbalanced by Ag-induced could be injected directly in the SWCNTs films, rendering doping of the SWCNTs that leads to poorer rectification more charges available for collection. The efficiency of this and a decrease of the optical transparency of the Ag NP/ process is known to be low and potentially counts for only SWCNTs film. a small fraction of the J and IQE increase [40]. sc However, the increases in J and IQE can only partially sc explain the more large increase in PCE. As shown in Conclusion Fig. 5b, the main cause of the PCE improvement is the variation of the FF. The latter is directly related to the In conclusion, we report in this paper a straightforward and device properties (e.g., film resistance, contact resistances, clean way to decorate SWCNTs films with Ag NPs to form etc.). The device’s electric properties thus affect PCE more NHs that are directly integrated into PV devices. Using directly than it affects IQE or V . Figure 6 shows the Raman spectroscopy, we can pinpoint the conditions at oc electrical resistance of AgNPs/SWCNTs films. It can be which the Ag NP film percolates and act more as a bulk silver film. We were also able to concomitantly identify this point in optical spectroscopy measurements and cor- relate this with PCE measurements. PCE of SWCNTs/n-Si devices is indeed increased from 3.5 to 6 % through dec- oration of the SWCNTs films with Ag NPs. We show that the PV properties are directly related to the Ag NPs films’ properties and to the various interactions between SWCNTs and Ag. Indeed, there is a combination of factors that contribute to the improvement of PCE. Raman spec- troscopy measurements show that little to know the charge transfer occurs at low values of N but that major charge Lp transfer-like phenomena hamper photogeneration at high N values. However, through optical and electrical char- Lp acterizations, we show that the PCE improvement appears to arise from a synergy between improved optical absorp- tion and electrical conductivity. Future work should focus on numerical modeling of this device to better understand Fig. 6 Film resistance and optical transmission at 550 nm of the importance of each factor. SWCNTs/Ag-NPs hybrids as a function of N Lp 123 1 Page 8 of 9 Mater Renew Sustain Energy (2016) 5:1 Open Access This article is distributed under the terms of the Crea- nanotube-based field effect transistors. Nanotechnology 20(17), tive Commons Attribution 4.0 International License (http:// 175203 (2009) creativecommons.org/licenses/by/4.0/), which permits unrestricted 16. Li, Z., Kunets, V.P., Saini, V., Xu, Y., Dervishi, E., Salamo, G.J., use, distribution, and reproduction in any medium, provided you give Biris, A.R., Biris, A.S.: SOCl enhanced photovoltaic conversion appropriate credit to the original author(s) and the source, provide a link of single wall carbon nanotube/n-silicon heterojunctions. Appl. to the Creative Commons license, and indicate if changes were made. Phys. Lett. 93(24), 243117 (2008) 17. Bai, X., Wang, H.G., Wei, J.Q., Jia, Y., Zhu, H.W., Wang, K.L., Wu, D.H.: Carbon nanotube-silicon hybrid solar cells with hydrogen peroxide doping. Chem. Phys. Lett. 533, 70–73 (2012) References 18. Li, X., Jung, Y., Huang, J.-S., Goh, T., Taylor, A.D.: Device area scale-up and improvement of SWNT/Si solar cells using silver 1. Du¨rkop, T., Getty, S.A., Cobas, E., Fuhrer, M.S.: Extraordinary nanowires. Advanced Energy Materials 4(12), (2014) mobility in semiconducting carbon nanotubes. Nano Lett. 4(1), 19. Schaadt, D.M., Feng, B., Yu, E.T.: Enhanced semiconductor 35–39 (2004) optical absorption via surface plasmon excitation in metal 2. Hecht, D.S., Hu, L., Irvin, G.: Emerging transparent electrodes nanoparticles. Appl. Phys. Lett. 86, 063106 (2005) based on thin films of carbon nanotubes, graphene, and metallic 20. Chen, X., Jia, B., Saha, J.K., Cai, B., Stokes, N., Qiao, Q., Wang, nanostructures. Adv. Mater. 23(13), 1482–1513 (2011) Y., Shi, Z., Gu, M.: Broadband enhancement in thin-film amor- 3. Le Borgne, V., Gautier, L.A., El Khakani, M.A.: Figure of merit phous silicon solar cells enabled by nucleated silver nanoparti- based maximization of the quantum efficiency of SWCNTs/n-Si cles. Nano Lett. 12(5), 2187–2192 (2012) photovoltaic devices. Appl. Phys. Lett. 103, 073103 (2013) 21. Chen, L., Zhang, S., Chang, L., Zeng, L., Yu, X., Zhao, J., Zhao, 4. Ham, M.-H., Paulus, G.L.C., Lee, C.Y., Song, C., Kalantar- S., Xu, C.: Photovoltaic conversion enhancement of single wall Zadeh, K., Choi, W., Han, J.-H., Strano, M.S.: Evidence for high- carbon-Si heterojunction solar cell decorated with Ag nanopar- efficiency exciton dissociation at polymer/single-walled carbon ticles. Electrochim. Acta 93, 293–300 (2013) nanotube interfaces in planar nano-heterojunction photovoltaics. 22. Gautier, L.-A., Le Borgne, V., Delegan, N., Pandiyan, R., El ACS Nano 4(10), 6251–6259 (2010) Khakani, M.A.: Field electron emission enhancement of 5. Park, J.-G., Akhtar, M.S., Li, Z.Y., Cho, D.-S., Lee, W., Yang, graphenated MWCNTs emitters following their decoration with O.-B.: Application of single walled carbon nanotubes as counter Au nanoparticles by a pulsed laser ablation process. Nanotech- electrode for dye sensitized solar cells. Electrochim. Acta 85, nology 26(4), 045706 (2015) 600–604 (2012) 23. Ka, I., Le Borgne, V., Ma, D., El Khakani, M.A.: Pulsed laser 6. Ong, P.-L., Euler, W.B., Levitsky, I.A.: Hybrid solar cells based ablation based direct synthesis of single-wall carbon nanotubes/ on single-walled carbon nanotubes/Si heterojunctions. Nan- PbS-quantum-dots nanohybrids exhibiting strong, spectrally wide otechnology 21(10), 105203 (2010) and fast photoresponse. Adv. Mater. 24, 6289–6294 (2012) 7. Jung, Y., Li, X., Rajan, N.K., Taylor, A.D., Reed, M.A.: Record 24. Ka, I., Ma, D., El Khakani, M.A.: Tailoring the photoluminescence high efficiency single-walled carbon nanotube/silicon p–n junc- of PbS-nanoparticles layers deposited by means of the pulsed laser tion solar cells. Nano Lett. 13(1), 95–99 (2013) ablation technique. J. Nanopart. Res. 13(6), 2269–2274 (2011) 8. Kozawa, D., Hiraoka, K., Miyauchi, Y., Mouri, S., Matsuda, K.: 25. Moskovits, M.: Surface-enhanced Raman spectroscopy: a brief Analysis of the photovoltaic properties of single-walled carbon retrospective. J. Raman Spectrosc. 36(6–7), 485–496 (2005) nanotube/silicon heterojunction solar cells. Appl. Phys. Express 26. Rinzler, A.G., Liu, J., Dai, H., Nikolaev, P., Huffman, C.B., 5(4), 42304 (2012) Rodriguez-Macı´as, F.J., Boul, P.J., Lu, A.H., Heymann, D., 9. Jia, Y., Li, P., Gui, X., Wei, J., Wang, K., Zhu, H., Wu, D., Colbert, D.T., Lee, R.S., Fischer, J.E., Rao, A.M., Eklund, P.C., Zhang, L., Cao, A., Xu, Y.: Encapsulated carbon nanotube-oxide- Smalley, R.E.: Large-scale purification of single-wall carbon silicon solar cells with stable 10 % efficiency. Appl. Phys. Lett. nanotubes: process, product, and characterization. Appl. Phys. 98(13), 133115 (2011) A Mater. Sci. Process. 67(1), 29–37 (1998) 10. Tune, D.D., Flavel, B.S., Krupke, R., Shapter, J.G.: Carbon 27. Kataura, H., Kumazawa, Y., Maniwa, Y., Umezu, I., Suzuki, S., nanotube-silicon solar cells. Adv. Energy Mater. 2(9), 1043–1055 Ohtsuka, Y., Achiba, Y.: Optical properties of single-wall carbon (2012) nanotubes. Synth. Met. 103(1–3), 2555–2558 (1999) 11. Tune, D.D., Blanch, A.J., Krupke, R., Flavel, B.S., Shapter, J.G.: 28. Bachilo, S.M., Strano, M.S., Kittrell, C., Hauge, R.H., Smalley, Nanotube film metallicity and its effect on the performance of R.E., Weisman, R.B.: Structure-assigned optical spectra of sin- carbon nanotube-silicon solar cells. Phys. Status Solidi A 211(7), gle-walled carbon nanotubes. Science 298(5602), 2361–2366 1479–1487 (2014) (2002) 12. Le Borgne, V., Gautier, L.A., Castrucci, P., Del Gobbo, S., De 29. Valmalette, J.-C., Tan, Z., Abe, H., Ohara, S.: Raman scattering Crescenzi, M., El Khakani, M.A.: Enhanced UV photoresponse of linear chains of strongly coupled Ag nanoparticles on of KrF-laser-synthesized single-wall carbon nanotubes/n-silicon SWCNTs. Sci. Rep. 4, 5238 (2014) hybrid photovoltaic devices. Nanotechnology 23(21), 215206 30. Kneipp, K., Kneipp, H., Itzkan, I., Dasari, R.R., Feld, M.S.: (2012) Ultrasensitive chemical analysis by Raman spectroscopy. Chem. 13. Li, Z., Kunets, V.P., Saini, V., Xu, Y., Dervishi, E., Salamo, G.J., Rev. 99, 2957–2976 (1999) Biris, A.R., Biris, A.S.: Light-harvesting using high density 31. Kharlamova, M.V., Niu, J.J.: Donor doping of single-walled p-type single wall carbon nanotube/n-type silicon heterojunc- carbon nanotubes by filling of channels with silver. J. Exp. Theor. tions. ACS Nano 3(6), 1407–1414 (2009) Phys. 115(3), 485–491 (2012) 14. Shi, E., Zhang, L., Li, Z., Li, P., Shang, Y., Jia, Y., Wei, J., Wang, 32. Niu, Z.Q., Fang, Y.: Surface-enhanced Raman scattering of sin- K., Zhu, H., Wu, D., Zhang, S., Cao, A.: TiO -coated carbon gle-walled carbon nanotubes on silver-coated and gold-coated nanotube-silicon solar cells with efficiency of 15%. Sci. Rep. 2, filter paper. J. Colloid Interface Sci. 303(1), 224–228 (2006) 884 (2012) 33. Tsang, J.C., Freitag, M., Perebeinos, V., Liu, J., Avouris, Ph: 15. Aı¨ssa, B., El Khakani, M.A.: The channel length effect on the Doping and phonon renormalization in carbon nanotubes. Nat. electrical performance of suspended-single-wall-carbon- Nanotechnol. 2(11), 725–730 (2007) 123 Mater Renew Sustain Energy (2016) 5:1 Page 9 of 9 1 34. Lin, Y., Watson, K.A., Fallbach, M.J., Ghose, S., Smith, J.G., 37. Stapleton, A.J., Afre, R.A., Ellis, A.V., Shapter, J.G., Andersson, Delozier, D.M., Cao, W., Crooks, R.E., Connell, J.W.: Rapid, G.G., Quinton, J.S., Lewis, D.A.: Highly conductive interwoven solventless, bulk preparation of metal nanoparticle-decorated carbon nanotube and silver nanowire transparent electrodes. Sci. carbon nanotubes. ACS Nano 3(4), 871–884 (2009) Technol. Adv. Mater. 14(3), 035004 (2013) 35. Maciel, I.O., Campos-Delgado, J., Cruz-Silva, E., Pimenta, M.A., 38. Kelly, K.L., Coronado, E., Zhao, L.L., Schatz, G.C.: The optical Sumpter, B.G., Meunier, V., Lo´pez-Urı´as, F., Mun˜oz-Sandoval, properties of metal nanoparticles: the influence of size, shape, and E., Terrones, H., Terrones, M., Jorio, A.: Synthesis, electronic dielectric environment. Journal of Physical Chemistry B 107(3), structure, and Raman scattering of phosphorus-doped single-wall 668–677 (2003) carbon nanotubes. Nano Lett. 9(6), 2267–2272 (2009) 39. Djurisic, A.B., Li, H.: Optical properties of graphite. J. Appl. 36. Itkis, M.E., Perea, D.E., Niyogi, S., Rickard, S.M., Hamon, Phys. 85(10), 7404–7410 (1999) M.A., Hu, H., Zhao, B., Haddon, R.C.: Purity evaluation of as- 40. Clavero, C.: Plasmon-induced hot-electron generation at prepared single-walled carbon nanotube soot by use of solu- nanoparticle/metal-oxide interfaces for photovoltaic and photo- tion-phase near-IR spectroscopy. Nano Lett. 3(3), 309–314 catalytic devices. Nat. Photonics 8(2), 95–103 (2014) (2003)

Journal

Materials for Renewable and Sustainable EnergySpringer Journals

Published: Jan 9, 2016

References