Synthesis of Gd₂O₃:Ho³⁺/Yb³⁺ upconversion nanoparticles for latent fingermark detection on difficult surfaces

Abstract

Infrared to visible upconversion fluorescent nanoparticles of Gd₂O₃ codoped with Ho³⁺/Yb³⁺ ions are synthesized via thermal decomposition process. The X-ray diffraction analysis of as-synthesized nanoparticles and annealed sample at 1000 °C has shown body-centered cubic phase of Gd₂O₃. The synthesized phosphor has shown intense green emission upon 980-nm excitation. High-contrast latent fingermarks on some difficult semi-porous and non-porous surfaces under 980-nm diode laser excitation were developed through powder dusting and colloidal solution spraying techniques and the results are compared with the commercial green luminescent fingermark powder. The latent fingermarks were developed on transparent (biological glass slides), single-color (aluminum foil) and multicolor (plywood, plastic bottle and book cover page) background surfaces. The present study depicts that the upconversion-based latent fingermarks detection using Gd₂O₃:Ho³⁺/Yb³⁺ phosphor material is suitable over the other conventional powders and has potential for practical applications in forensic science.

1 Introduction

Upconversion emission in rare earth (RE) containing materials is an interesting phenomenon for many applications where conventional downconversion emission is not a well suitable process [1–6]. This unique frequency conversion capability of rare earth ions imports advantage that it exhibits no background interference due to low auto-fluorescence from the background [2]. The RE upconversion nanoparticles (UCNPs) have been assumed to have potential for display device, temperature sensing, biological imaging and random lasing applications [3–6]. New advanced chemical preparation routes are capable enough to produce shape and size confined nanoparticles. Owing to the suitability of UCNPs for above said applications, intense research is going on to improve the emission efficiency of chemically prepared UCNPs.

One of the interesting applications of UCNPs is in the field of forensic science [2]. To identify the physical evidence of crime in forensic science investigations, fingermark plays an important key role because each finger has its unique ridge pattern like the DNA system in biological tissues. However, latent fingermark pattern is invisible to bare eyes and hence it is enhanced/developed by the use of various techniques [2, 7–11]. For the visualization of latent fingermarks, conventional techniques like powder dusting [12], cyanoacrylate fuming [13], ninhydrin spraying [14] and silver nitrate soaking [15, 16] are being utilized by crime investigators. Apart from these techniques, several reports of fingermark detection based on down converting fluorescence materials through UV excitation source have also came into light [17, 18]. But the conventional technique using downconversion fluorescent phosphors fails to enhance latent ridges on multicolor background surfaces due to auto-fluorescence of background that interfere in achieving high contrast [18–20].

The UCNPs-based detection of fingermarks has been demonstrated by several researchers [21–25]. The Er³⁺/Yb³⁺ and other ions doped/codoped UCNPs were used for the visualization of fingermarks on various surfaces [22–26]. Ma et al. [22, 23] have utilized commercially available NaYF₄:Er³⁺/Yb³⁺ and YVO₄:Er³⁺/Yb³⁺ UCNPs for the detection of latent fingermarks on various non-porous and semi-porous surfaces. The particles used by them were irregular in shape and size (range from 0.2 to 2.0 μm). They have indicated the potential of upconverting materials in developing fingermarks with high contrast on substrates that would otherwise prove to be difficult. Recently, Wang et al. [24] have used solvothermal approach and successfully synthesized spherical-shaped NaYF₄:Er³⁺/Yb³⁺ UCNPs having ~15 nm average size. Authors have achieved excellent clarity in patterns with high developing sensitivity and selectivity using these nanoparticles. In particular, they have efficiently developed one-year-old fingermarks using UCNPs. Tiwari et al. [25] in their paper have demonstrated the same using the combustion route synthesized La₂O₃:Er³⁺/Yb³⁺ upconversion system. These researchers found that powder prepared via combustion route has also good sticking property and hence provide good clarity of ridges in the developed fingermark. The researchers are working on different hosts and dopants for the efficient development of latent fingermarks using upconversion emission-based recording. Above results suggest that upconversion phosphors may have potential to replace the conventional downconversion phosphors for forensic science; however, an intense research is required in this direction. It is a big challenge for forensic scientists that how a latent fingermarks can be read out on-spot with high accuracy to produce it as an evidence of committed crime. For efficient detection of fingermarks using powder dusting or colloidal spraying of UCNPs, several conditions such as low phosphor cost, high emission efficiency in visible region, optimized shape and size need to be optimized. The fine particles have higher capability of sticking at the ridges of latent fingermarks [24], and it may increase the uniformity and distinguishability of developed ridge patterns with high contrast.

In the present work, authors had the motive to use thermal decomposition route [26] to prepare size and shape controlled UCNPs of Gd₂O₃:Ho³⁺ (0.3 mol%)/Yb³⁺ (3.0 mol%) phosphor and then to test its efficacy in fingermark development. Previous studies have shown that Gd₂O₃ is good host to achieve high upconversion emission. The powder dusting and colloidal solution spraying techniques are used to sprinkle the UCNPs on fingermarks made upon porous and non-porous surfaces.

2 Materials and methods

2.1 Chemical synthesis

Fine spherical particles of Gd₂O₃ codoped Ho³⁺/Yb³⁺ were synthesized via modified thermal decomposition method [27]. Hexa-hydrated rare earth nitrates Gd(NO₃)₃·6H₂O (96.7 mol%), Ho(NO₃)₃·6H₂O (0.3 mol%) and Yb(NO₃)₃·6H₂O (3 mol%) were taken along with oleic acid (C₁₈H₃₄O₂, 60–88 %), oleylamine (C₁₈H₃₇N, >95 % conc.), n-octadecene (C₁₈H₃₆, >90 % conc.) and ethanol. The whole synthesis process was performed in oxygen free environment (in presence of nitrogen gas). Here oleic acid (OA) and oleylamine (OM) were used as chelating as well as capping agent to stabilize surface of the particles and to control the size of particles [28]. In a typical synthesis process, 10 mmol of oleic acid (OA), 10 mmol of oleylamine (OM) and 20 mmol of n-octadecene (ODE) were mixed homogenously in a 100 ml round bottom three-neck flask and then 1 mmol of hexa-hydrated rare earth nitrate was added to this solution. Initially, solution was heated at 130 °C for 80 min under constant stirring to remove the water content with gentle flow of N₂ gas. After this, temperature was increased up to 330 °C with continuous flow of N₂ gas. At this temperature, the decomposition takes place and crystal particles start to form. After completion of the reaction, mixture was cooled down to room temperature. After cooling, the ethanol solution was added in excess amount and centrifuged at around 11,000 rpm for 10 min to collect the precipitate. The raw precipitated sample was then washed three times by ethanol and dried at 80 °C over night to obtain nanoparticles of Gd₂O₃:Ho³⁺/Yb³⁺. The as-synthesized sample was further annealed at 1000 °C for 3 h.

3 Characterization tools

Crystal phase of the samples was investigated using powder X-ray diffraction (XRD) acquired on Bruker D8 advanced X-ray diffractometer. Surface morphology and size of the particles were studied by field emission scanning electron microscopy (FESEM) technique using Supra 55, Carl Zeiss instrument. Upconversion spectra of the samples were monitored on SP-2300 grating spectrometer (Princeton Instruments, USA) equipped with a photomultiplier tube. The 980-nm diode laser was used as an excitation source.

The latent fingermarks on variety of surfaces such as transparent biological glass plate, single-color background surface (aluminum foil) and multicolor background surfaces (plastic bottle, book cover and ply wood surfaces) were deposited after the donor wiped his fingers over forehead or facial area just prior to fingermark deposition. The synthesized powder sample was dusted over the area of fingermarks and then 980-nm diode laser beams were manually scanned on the fingermark area. The images of fingermarks were recorded using a general purpose Nikon D 3200 (Japan) camera. The shutter of camera was opened for 20 s, and within this time span scanning process was completed. To block the scattered excitation radiation, a short-pass filter (model: FE50750, Thor labs, USA) was used in front of camera lens. The quality of developed images of fingermarks (contrast, ridges pattern etc.) was compared with commercial green luminescent fingermark powder (Crime Scene, Arizona, USA). The UV excitation source of 380-nm wavelength was used to illuminate fingermarks dusted with commercial green phosphor.

3.1 Structural characterization

3.1.1 X-ray diffraction analysis

The XRD pattern of as-prepared as well as annealed samples at 1000 °C for 3 h is shown in Fig. 1. The peak positions corresponding to different diffraction angles (2θ) values at 20.28°, 28.49°, 33.07°, 35.16°, 42.95°, 47.54°, 56.57°, 59.22°, 77.03° and 79.26° were obtained at (211), (222), (400), (411), (332), (440), (622), (444), (662) and (840). Sharpness in peak intensity increases under the influence of annealing temperature which can be seen in the figure. These XRD peaks were matched well with JCPDS File No. 43-1014 and no extra peaks due to impurity or secondary phases are observed. Phase of the sample was found to body-centered cubic crystal (bcc) system with Ia-3 (206) space group. The crystalline size was estimated with the help of Debye–Scherrer formula [29].

$$d = \frac{0.9\lambda }{\beta \cos \theta }$$

(1)

where β is a full width at half maxima (FWHM) corresponding to each half diffraction angle θ. λ is the wavelength of diffracted X-ray for Cu-K_α radiation having the value of 0.154 nm. The average crystallite size was calculated around 30 and 45 nm for as-synthesized and annealed (1000 °C for 3 h) samples, respectively. From the XRD pattern shown in Fig. 1, the FWHM of diffraction peaks is seen to decrease for annealed sample and as a result average crystallite size is found to increase around 1.5-fold compared to as-synthesized sample.

Fig. 1

X-ray diffraction (XRD) pattern of Gd₂O₃:Ho³⁺/Yb³⁺ nanoparticles a annealed at 1000 °C and b as-synthesized and c reference peaks of Gd₂O₃ sample matched with JCPDS Card No. 43-1014

3.1.2 FESEM analysis

The size of particle is a very important factor in fingermark detection. The sample having large or irregular particle size may not stick properly at ridges of the fingermarks [30]. This causes difficulty in reading of deposited pattern. Hence the particles with small size and uniform shape can develop regular uniform bright ridge patterns and that has been demonstrated in the present paper. In SEM image, shown in Fig. 2, uniform size spherical-shaped particles are seen. The average particle size of as-synthesized sample is found around 40 nm. Here it is to be mention that SEM image gives the actual particle size while XRD analysis gives idea of crystallite size (distance up to what lattice periodicity holds). During the image recording at high resolution, it was found that impinging electron beam causes sample heating and because of it particles were found stick to each other with movement and thus results de-focused image.

Fig. 2

The field emission scanning electron microscopy (FESEM) image of Gd₂O₃:Ho³⁺/Yb³⁺ nanoparticles. Inset shows particle size distribution obtained using image J software

3.2 Optical characterizations

3.2.1 Upconversion measurement

The upconversion emission spectra of the prepared sample were recorded in the visible region using 980-nm excitation [31–33]. Figure 3a, b shows upconversion spectra for annealed sample at various excitation power densities. The emission intensity is seen to increase with the pump power. Good intensity green glow was seen to the eyes. The camera images of green color emission from samples are given in the inset of Fig. 3a. First inset figure is the image of fluorescence for powder form of the sample while the second inset figure is for the colloidal suspension of particles in hexane. Good thing here is that most part of the emitted intensity falls to the green emission, and other emission bands are found very weak relative to green emission. This property of sample indicates its suitability for fingermarks detection. The prominent emission peak is found at 550 nm with weak intensity peaks at 667 and 757 nm corresponding to the ⁵F₄/⁵S₂ → ⁵I₈, ⁵F₅ → ⁵I₈ and ⁵I₄ → ⁵I₈ transitions. It is also observed that the background noise around 667 and 757 nm emission bands are very high. By observing the slopes of ln–ln plot (Fig. 4) between the emission intensity (I) and pump power (P) for green, red and NIR emission bands, it can be confirmed that these bands originate due to the absorption of two NIR photons [30].

Fig. 3

The upconversion emission spectra corresponding to dominant band at 550-nm wavelength (a). The green color of the powder and colloidal dispersed samples are captured after illuminating 980-nm diode laser (inset of a). b shows the upconversion emission spectra corresponding to 667- and 757-nm wavelength

Fig. 4

Shows ln–ln plot between emission intensity (I) and laser power density (P)

The UC mechanism for the present sample has been explained in our previous work [27]. The possible transitions have been shown in Fig. 5. On 980-nm excitation, Yb³⁺ ions first absorb the incident light and excited into ²F_5/2 level due to higher concentration of Yb³⁺ ions and also its higher absorption cross section compared to Ho³⁺ ion [32]. The ²F_5/2 level then decays to ground level ²F_7/2 via energy transfer to ⁵I₅ level of Ho³⁺ ion. The ⁵I₅ level of Ho³⁺ ion then decays to ⁵I₆ level through non-radiative emission. The Ho³⁺ ions in ground state also absorb 980-nm excitation and excite to ⁵I₅ level. The ⁵I₆ level of Ho³⁺ ion has meta-stable character and hence this level further absorbs incident photon and excites to ⁵F₄/⁵S₂ levels. The ⁵F₄/⁵S₂ and just lower ⁵F₄ levels decay radiatively to ground state and give emission at 550 and 667 nm, respectively. There is also a probability that Yb³⁺ ions in ²F_5/2 levels cooperatively transfer its energy to ⁵F₃/³K₈ levels of Ho³⁺ ion. These ⁵F₃/³K₈ levels then decay to lower levels through non-radiative transition. The ⁵I₄ level is populated by non-radiative relaxation of higher states, and its decay to ground level gives emission at 757 nm. Whole process is shown in Fig. 5.

Fig. 5

The excitation and emission mechanism between sensitizer (Yb³⁺) and activator (Ho³⁺) dopant in Gd₂O₃ nanoparticles

The color coordinates of the emission was also calculated using Go CIE software, and obtained values are shown in Fig. 6 for both as-synthesized as well as annealed samples. The CIE color coordinates [34] of as-synthesized sample are found to 0.35, 0.64 with 94 % color purity whereas for 1000 °C annealed sample it is found to 0.30, 0.69 with 99 % color purity at the excitation density of 90 W/cm². Having 99 % color purity, the annealed sample yields an intense green upconversion emission. However, intensity of as-synthesized sample is sufficient to record fingermarks images. The green to red color ratio for annealed sample is found to increase with increasing the pump power as shown in Fig. 6b. The CIE plot confirms that there is almost nil contamination of other colors. Therefore, the present sample seems appropriate for fingermarks detection and other purposes.

Fig. 6

a CIE (International Commission on Illumination chromaticity diagram) color coordinates (x, y) representation on CIE color diagram of as-synthesized and annealed samples at 1000 °C and b variation of green to red ratio with pump power density

4 Application of sample as fingermark detection

4.1 Deposition protocol and fingermarks detection

In Fig. 7 the pictorial demonstration of fingermarks deposition and detection is shown. The whole process is shown in three steps [35, 36]. In step 1, the fingermark was deposited on a plane surface. The second step is the powder dusting/colloidal spraying on the fingermark area using squirrel brush. The colloidal spraying method needs washing of extra particles with low density liquid such as cyclohexane. In third and final step, a beam of 980-nm wavelength diode laser was manually scanned over the region and images were recorded with the help of a camera.

Fig. 7

The demonstration technique of fingermark deposition. Nikon D 3200 DSLR camera coupled with a short-pass filter (model: FE50750, Thor labs, USA) is used for detection purpose by illuminating 980-nm diode laser source

The specimen of fingermarks was supplied by a volunteer of 28 years old. Prior to fingermark deposition, volunteer was said to clean his hands with soap then dry it and do not touch anything before taking fingermarks on the different surfaces. The donor wiped his fingers at forehead area to attach the natural secretions like sebaceous fat. The fingermarks were left at different non-porous and semi-porous surfaces mentioned elsewhere in this manuscript. The annealed Gd₂O₃:Ho³⁺/Yb³⁺ powder was dusted on fingermark area. The excess powder was removed by squirrel brush and light air flow. The colloidal solution of sample was made by the dispersion of dry powder in cyclohexane followed by ultra sonication for 10 min. The 0.5 ml solution was filled in a plastic sprayer and sprayed upon the area containing fingermark. The specimen was left for 10 min and after that fingermark area was dried for recording purpose. For comparative study of fingermarks, commercial green luminescent fingermark powder (Crime Scene, Arizona, USA) was used on the same surfaces discussed above and fingermarks were excited by 380-nm UV source. A 980-nm diode laser was used to illuminate the fingermarks at fixed excitation power (24 W/cm²) under dark condition for upconversion emission mode recording. Since the laser beam covers fewer area (8 × 5 mm²) of selected mark, beam was scanned over entire fingermark. Consequently, a long shutter speed (20 s) camera (Nikon D 3200) was used to record the images.

To detect the latent fingermarks different typical surfaces were chosen. The first surface was transparent biological glass slide. Apart from this we have chosen some semi-porous (book cover page) and non-porous (aluminum foil, plywood and plastic bottle surfaces) surfaces with single- and multicolor surfaces. The same fingermarks were also developed and recorded in downconverting emission mode through 380-nm UV lamp excitation using commercial green luminescent fingermark powder. The complete description of work is given in the next section.

4.1.1 Development of latent fingermarks on transparent glass surface

First trial was made on glass slide as it is easy surface. Three fingermarks with light thumb impression were made on glass slides and developed using powder/colloidal particles using the process described above. A comparison of recorded images using laboratory synthesized and commercial powder samples is shown in Fig. 8. Figure 8a is the bright field image of biological glass surface where the fingermark was made. Using commercial sample, the recorded fingermark is shown in Fig. 8b. Figure 8c is the image of fingermark developed using Gd₂O₃:Ho³⁺/Yb³⁺ powder sample and illuminated by 980 nm diode laser. Figure 8d was recorded using colloidal suspension of Gd₂O₃:Ho³⁺/Yb³⁺ nanoparticles. After seeing these images, it can be interpreted that Fig. 8b has lowest contrast and also has less clear ridge patterns. Strong background is present that reduces the contrast of the ridges. Moreover, at some places in Fig. 8b sample distribution is not uniform. Due to this ridge pattern is not detectable at those places. The images, shown in Fig. 8c, d, have, however, good contrast, clearer ridge patterns and good sample uniformity. The color intensity in Fig. 8d is somewhat lower compared to the Fig. 8c because the use of nanoparticles in hexane medium. However, ridges are equally clear and readable. The result shows that upconversion recording mode produces better contrast and clarity than the commercial green downconversion phosphor. The results can also say that wet spraying of colloidal UCNPs is also suitable for clear detection of fingermarks.

Fig. 8

Developed latent fingermarks on biological glass surface a bright field mode, b with commercial green luminescent fingermark powder excitation with 380-nm UV lamp, c, d fingermarks developed with Gd₂O₃:Ho³⁺/Yb³⁺ nanoparticles using powder and colloidal dispersed samples, respectively, after illuminating 980-nm diode laser. A general purpose camera [Nikon D 3200 coupled with a short-pass filter (model: FE50750, Thor labs, USA)] was used to record the images

4.1.2 Development of latent fingermarks on reflecting surface

After the glass plate, single-color reflecting surface (aluminum foil) was selected to develop the fingermark using prepared upconversion sample. The bare aluminum foil image is shown in Fig. 9a. The fingermark image developed by commercial green powder (illuminated by 380 nm) is shown in Fig. 9b. Due to the high reflection of UV excitation light, the fingermark color recorded by the camera looks cyan. Also background is seen with mix of blue and red colors. As a result the fingermark ridge pattern is not distinguishable. The image recorded in upconversion emission mode using powder Gd₂O₃:Ho³⁺/Yb³⁺ phosphor is shown in Fig. 9c. This image shows dark background and readable ridge pattern. Again here upconversion emission shows advantage over the downconversion emission.

Fig. 9

Development of latent fingermarks on aluminum foil surface a under bright field, b with commercial green luminescent fingermark powder excitation with 380-nm UV lamp and c fingermarks developed with Gd₂O₃:Ho³⁺/Yb³⁺ nanoparticles, recorded by a short-pass filter with Nikon D 3200 camera having 20 s of shuttering speed setup. Illumination of 980-nm diode laser was used

4.1.3 Development of latent fingermarks on multicolor background surface

In the next step, more difficult multicolored surfaces were selected. Figures 10, 11 and 12 are showing the recorded images of fingermarks on book cover page, plastic bottle and plywood surfaces, respectively. In Fig. 10 the fingermark in upconversion mode on two color book cover page is compared with the commercial green phosphor. Figure 10a is the book cover page on which fingermarks were developed. Here fingermark developed with commercial powder shows serious auto-fluorescence, and hence fingermark ridges are not detectable at all. The fingermark developed by powder dusting of Gd₂O₃:Ho³⁺/Yb³⁺, as shown in Fig. 10c, however, shows readable ridge pattern. Very little background is seen here. Similar observation is seen for next two surfaces viz. multicolored non-porous plastic bottle surface and two color semi-porous plywood surface. The images of fingermarks developed using upconversion sample are only readable, Fig. 11c for plastic bottle and Fig. 12c for plywood. Here situation is worse for images recorded using commercial phosphor.

Fig. 10

Latent fingermark development on a book cover surface a image of book cover page before dusting, b with commercial green luminescent fingermark powder illuminated by 380 nm and c fingermark with the help of Gd₂O₃:Ho³⁺/Yb³⁺ nanoparticles via excitation through 980-nm laser diode. All images were recorded using a general purpose short-pass filter and camera (Nikon D 3200) having short shutter speed 20 s

Fig. 11

The latent fingermarks developed on plastic bottle surface a in presence of bright field, b with commercial green luminescent fingermark powder, 380-nm illumination was used and c fingermark with Gd₂O₃:Ho³⁺/Yb³⁺ nanoparticles via 980-nm diode laser excitation. Images were recorded through a short-pass filter and camera (Nikon D 3200) setup

Fig. 12

The latent fingermarks developed on a ply wood surface a under bright light mode, b with commercial green luminescent fingermark powder with 380-nm excitation and c fingermark developed with Gd₂O₃:Ho³⁺/Yb³⁺ nanoparticles with 980-nm diode laser excitation. Images were recorded through a short-pass filter and camera (Nikon D 3200) combination

On comparing of all the images of fingermarks, it can be concluded that on multicolor hard surfaces conventional green phosphor is not able to produce readable ridge patterns; however, on single-color glass and reflecting aluminum surface conventional green phosphor produce readable patterns. On the contrary, upconversion emission mode using Gd₂O₃:Ho³⁺/Yb³⁺ powder/colloidal form produce easily readable patterns with low background on all the studied surfaces. The background red hue seen in upconversion mode images of Fig. 10c and 12c is due to the incomplete rejection of laser excitation at objective lens of the camera, and it can be improved using better filters. Hence there is a possibility of improving the image quality.

5 Conclusions

Spherical shape Gd₂O₃:Ho³⁺/Yb³⁺ upconversion particles in powder and colloidal form were successfully prepared. Very intense green emission around 550-nm wavelength on 980-nm excitation is observed in this prepared phosphor. The green color purity is calculated using CIE color software and is found high in purity (CIE: x = 0.30, y = 0.69). Having strong green emission and fine particle size, this phosphor has produced readable fingermarks in upconversion mode on non-porous/semi-porous surfaces like biological glass plate, aluminum foil, ply wood, book cover and plastic bottle. The comparison of developed fingermarks using upconversion phosphor with commercial green luminescence powder has shown the superiority of synthesized nano sized upconversion particles. In conclusion, Gd₂O₃:Ho³⁺/Yb³⁺ upconversion nanoparticles can be used in latent fingermark detection with high selectivity, high contrast and low background noise, and thus, the rare earth doped phosphor nano-materials may be potential candidates for forensic science practitioners.