1. Introduction

Matrix assisted Laser desorption/ionization mass spectroscopy (MALDI-MS) is a versatile technique to analyze bio-related molecules [1, 2]. For MALDI-MS measurements, matrix molecules are needed. The matrix molecules generate large numbers of fragments when irradiated by ultraviolet (UV) pulsed-laser light and assist the desorption and ionization of target molecules dispersed within matrix molecule microcrystals [2–4]. MALDI processes efficiently desorb and ionize larger molecules (> 1000 Da), but smaller molecules (< 500 Da) are difficult to detect because of the large background signals from the matrix fragments.

There are several techniques that aim to replace the matrix molecules. Surface-assisted laser desorption/ionization (SALDI) is an alternative method to MALDI-MS. In SALDI-MS measurements, plates with nanostructured surfaces are used to ionize target molecules [5, 6]. The absence of matrix molecules enables the detection of small molecules with molecular weights of less than 500.

Gold nanoparticles are the most studied nanoparticles because of their distinctive surface plasmon (SP) bands at around 520 nm. Gold nanoparticles are a preferable material to obtain well-designed functional nanostructures. Recently, gold nanorods, which are rod-shaped gold nanoparticles, were found to be a possible nanomaterial for SALDI-MS [7–9]. Castellana et al. reported near-IR induced SALDI-MS measurements using a combination of colorimetry and mass spectrometry. Conversely, we reported highly efficient SALDI processes using gold nanorods on an ITO plate. Trace amounts of oligopeptide (10⁻¹⁴ M of angiotensin I) could be detected using nanorod-deposited ITO plates [8].

For SALDI-MS measurements, however, gold is not the best material because many gold clusters are generated when the nanorods are irradiated by UV pulsed-laser light. Platinum nanoparticles seem to be promising material to produce an efficient SALDI plate [10, 11]; however, gold nanoparticles have advantages in preparation and evaluation of nanostructured surfaces. Because their SP bands are sensitive to size, shape, aggregation, and dielectric environment, the spectroscopic properties of gold nanoparticles on a SALDI plate provide information as to the uniformity of the deposited gold nanoparticles and of the nanostructured gold nanoparticle assemblies. Fusion and fragmentation of gold nanoparticles and their assemblies can also be detected in their spectral profiles.

The gold nanorods show two SP bands in the visible and near-infrared (IR) regions [12]. The near-IR SP bands, which originate from longitudinal SP oscillations, are very sensitive to shape changes and aggregation of only a few nanorods [13]. These optical properties are helpful in evaluating the degree of aggregation of small nanorod assemblies [14]. Thus, gold nanorods are suitable for studies designed to interrogate the relationship between SALDI processes and surface nanostructures.

In this work, we evaluated the relationship between gold nanorod assemblies and SALDI efficiency, and discuss the preferable assembly structures of gold nanorods for efficient SALDI processes. Spectroscopic profiles of the longitudinal SP bands, dark field optical microscopy, and scanning electron microscopy (SEM) were used to evaluate the degree of aggregation and morphological changes of the nanorods before and after the mass measurements.

2. Experimental

Gold nanorods were obtained from Dai-Nippon Toryo Co. Ltd., Japan. The gold nanorods were wrapped with poly(stylene sulfonate) (PSS), and deposited on the cationic ITO surface under the influence of electrostatic interactions [8]. To control the degree of aggregation of the gold nanorods on the ITO plates, the ITO plates were repeatedly immersed in a PSS-nanorod solution. We prepared thirty-one ITO plates (nanorod-deposited plates) on which the gold nanorods were deposited with various degrees of aggregation.

A solution of angiotensin I (Peptide Institute, DRVYIHPFHL) was cast on a nanorod-plate, and dried in air. MS measurements were performed using a MALDI-MS instrument (Autoflex, Bruker-Daltonics). Coupled Plasma mass spectrometry (ICP-MS) was used to evaluate the amount of gold on the plate. The extinction spectra for the ITO plates were obtained using a conventional spectrophotometer (V-570, JASCO). Scanning electron microscopy (SEM, JSM-6701F, JEOL) and an optical microscope (TE2000, Nikon) equipped with a dark field apparatus were used to observe the deposited plates.

3. Results and discussion

3.1 SALDI-MS signals and SP bandwidth

The extinction spectra of the thirty-one nanorod-deposited plates were recorded; four typical extinction spectra and SEM images of the corresponding samples are shown in Fig. 1. The extinction band in the near-IR region was assignable to the SP bands of the gold nanorods [15]. The longitudinal SP bands in the near-IR region were sensitive to the aggregation of nanorods [14]; the broadening of the bands in the longer wavelength region likely indicates formation of end-to-end nanorod assemblies [16, 17]. When the deposited gold nanorods displayed a narrower SP band (a) with a 205 nm bandwidth, few aggregates were found in the SEM image (b). The nanorod-deposited plates with broader SP bands (c, e) showed larger numbers of aggregates in the corresponding SEM images (d, f). End-to-end assemblies contributed to broadening of the SP bands. Side-by-side assemblies and random aggregates showed their SP bands in the shorter wavelength regions [14]. In the case of the plate (g) that showed the broadest SP band, large-scale aggregates were seen in the SEM image (h). The larger aggregates did not show remarkable extinction peaks but show a very broad band extending from the visible to near-IR regions [17]. Thus, the large aggregates are “black” in the extinction spectra. The baseline increase in spectrum (g), shown as the thin dotted lines, is the extinction of the larger aggregates. These results demonstrate that the width of the SP bands can be used as a qualitative indicator for small assemblies of nanorods, such as end-to-end assemblies.

 

Fig. 1 Extinction spectra (a, c, e, g) and SEM images (b, d, f, h) of typical nanorod-deposited ITO plates. The bandwidths were 205 (a), 259 (c), 366 (e), and 480 nm (g).

Download Full Size | PDF

To evaluate the SALDI efficiencies of the nanorod-deposited plates, a solution of angiotensin (1 µM, 1 µL) was cast on the nanorod-deposited plates. Figure 2 shows typical mass spectra (Reflector, positive) for the four nanorod-deposited plates. Molecular ions of the angiotensin were detected at m/z = 1297. The signal/noise ratios (S/N) of the molecular ion peaks are shown in the figures. Plate (a), which had a narrow SP band, showed a very low SALDI efficiency, but the other plates showed significant signals. The mass signals for the thirty-one plates were taken from seven randomly chosen spots on the plates. The largest and smallest S/N values were omitted, and the averages and standard deviations of the remaining five S/N values were used to describe the SALDI efficiencies. Figure 3 shows the S/N ratio plotted against the widths of the corresponding SP bands. For narrower bandwidths (200–250 nm), the S/N ratios were very small; the SALDI efficiencies were therefore very low when isolated gold nanorods were deposited on the plate.

 

Fig. 2 Mass spectra of the four nanorod-deposited ITO plates (a–d), corresponding to those shown in Fig. 1.

Download Full Size | PDF

 

Fig. 3 S/N values for the mass signals of angiotensin I (1 µM), plotted against the FWHM of longitudinal SP bands for gold nanorods on the ITO plates.

Download Full Size | PDF

For the plates that had 250–350 nm SP bandwidths, the SALDI efficiencies depended on which locations on the mass spectrometry was performed. Intense signals were randomly obtained, which strongly suggested that there were a few extraordinary spots that showed highly efficient SALDI processes. Enhancement of the SALDI process occurred in a small area. For the plates with bandwidths larger than 350 nm, moderate signals were obtained everywhere on the plates. In these cases, the reproducibility of the SALDI processes was good, but the enhancement disappeared. It was shown that large aggregates consisting of hundreds of gold nanorods—as shown in the SEM image in Fig. 1(h)—are not the origin of the efficient SALDI processes.

3.2 Microscopic observation of nanostructures on ITO plates

In the dark field images of the nanorod-deposited plate after the MALDI-MS measurements (Fig. 4), many bright spots were found. Because the image was obtained using a conventional color CCD camera (DXM-1200, Nikon) that has no sensitivity in the near-IR region, the white scattered spots showed the distribution of aggregates of nanorod. It should be noted that periodic darker areas could be found in the image. The center-to-center distances of the darker areas are about 200 μm.The dark areas are assigned to the laser-light irradiated areas to make the mass measurements. Thus, laser-light irradiation decreased the light-scattering of the gold nanorods on the plate. This strongly suggested that morphological changes were induced on the nanorod assemblies by the pulsed-laser irradiation of the mass imaging measurement.

 

Fig. 4 Dark field images of the nanorod-deposited plate.

Download Full Size | PDF

To investigate the morphological changes of the nanorods by laser irradiation, the plate was observed by SEM. A typical image of the nanorods on the plate is shown in Fig. 5(a). Assemblies and isolated nanorods, which are similar to those in Fig. 1(f), were found. In some places on the plate, spherical particles were found, as shown in Fig. 5(b). Because the gold nanorod solution does not contain many spherical particles, which show their SP bands at around 520 nm, the spherical nanoparticles must be originated from laser-irradiated gold nanorods that are transformed into spherical shapes. The isolated spherical nanoparticles do not show broad extinction in the visible region in comparison with that of the small aggregates. Thus, the periodic darker areas (Fig. 5(b)) are assignable to the areas where spherical particles form by fusion of the small aggregates.

 

Fig. 5 SEM images of the nanorod-deposited plate. (a): an area where aggregated nanorods were seen. (b): an area where spherical particles were seen. The insets are enlarged images of the corresponding areas.

Download Full Size | PDF

We have made evaluateion of suitable sizes of gold nanorod assemblies for the SALDI processes; that was image processing of SEM images using Image-J. Typical procedures of the analysis are shown in Fig. 6. Analysis of isolated gold nanorods gave the projected areas of the gold nanorods. When the gold nanorods formed assemblies, the projected areas were evaluated as the sum of the areas of the gold nanorods forming the assemblies. When the projected–area of an assembly is divided by the typical area of an isolated gold nanorod, the quotient reports the number of nanorods forming the assembly.

 

Fig. 6 Image processing of an SEM image. The projected areas of typical gold nanorods and their assembly are shown.

Download Full Size | PDF

SEM images of four typical nanorod-deposited plates, of which typical images are shown in Fig. 1, were analyzed to evaluate the size distribution of the gold nanorod assemblies (aggregates) on the plates. Figure 7 shows size histograms for the assemblies. The horizontal axes are normalized by the typical projected area of an isolated gold nanorod. Thus, the horizontal axes indicate the number of gold nanorods included in an assembly. For example, the value “1” in the horizontal axis corresponds to isolated nanorods. The “2” and the “3” correspond aggregates consisting of two and three nanorods, respectively (dimers and trimers). The vertical axis is relative fraction of nanorods forming the corresponding aggregates. The fractions were normalized to be unit at the largest peak in the distribution. Plate (a), which showed the narrowest SP band (FWHM = 205 nm), showed the narrowest size distribution, which indicated that most of the gold nanorods were deposited as isolated nanorods. In the case of plates (b) and (c), the FWHM values were 259 and 366, respectively; the broad distributions indicated that small assemblies, consisting of ten or fewer nanorods, formed on the plates. In the case of plate (d), the FWHM was 480, and there were some aggregates on the plate. Comparing the histograms and the SALDI efficiencies in Fig. 3, we concluded that the most efficient SALDI processes originated from small assemblies consisting of 4 to 10 nanorods. A few of these assemblies were responsible for the enhancement of the SALDI processes.

 

Fig. 7 Histograms of assembly (aggregate) sizes (a–d), corresponding to the samples shown in Fig. 1

Download Full Size | PDF

3.3 Photo-processes of gold nanorods

Under our experimental conditions, some gold ions, which Au⁺, Au₂⁺, and Au₃⁺, were detected in the SALDI-MS measurements. This indicated that the pulsed UV light in the MS measurements was able to ablate the gold. This ablation is accompanied by a morphological transformation of the gold nanorods into spherical nanoparticles, as are shown in Fig. 5. It was reported that micron-scale morphological changes are necessary for efficient SALDI processes [18]. At every spot where angiotensin molecular ions were found, gold ions were also detectable. It is probable that fusion and ablation of gold nanorods accompanied the SALDI of the angiotensin molecular ions.

The mass spectrometry simultaneously gives signals of angiotensin and gold ions at a certain spot. In Fig. 8, S/N values for the mass signals of angiotensin are plotted against S/N values of Au⁺ ions. When the S/N of the Au⁺ ions was less than 200, the S/N of angiotensin was almost independent to the S/N of Au⁺ ion. In contrast, when the Au⁺ ions were detected with higher S/N ratios than 400, ionization of angiotensin was suppressed. Moderate LDI of Au⁺ ions could accompany the SALDI of the angiotensin molecules.

 

Fig. 8 S/N values for the mass signals of angiotensin I (1 µM), plotted against the S/N values of Au⁺ ion.

Download Full Size | PDF

Here, we can illustrate the outlines of photo-processes of the nanorod-deposited plates. That is, fusion of the nanorods into spherical particles occurs everywhere in the laser-irradiated areas. The fusion would be helpful to desorb the angiotensin molecules on the nanorod surfaces. The laser-irradiation also induced ablation of gold nanorods and gave gold ions (Au_n⁺). The ablation was accompanied by the SALDI of the angiotensin molecules, but moderate LDI of gold was preferable for the SALDI processes, because large amounts of gold ions suppressed the SALDI processes. Consequently, the balance of the fusion of the gold nanorods and the LDI of gold ions is important to give the most efficient SALDI processes. Overall, the densities of the area of the enhancement were low.

4. Conclusion

Efficient SALDI processes occurred at small assemblies consisting of 4–10 gold nanorods. Microscopic observation did not reveal specific nanostructures for efficient SALDI processes, but it was found that fusion and ablation of gold accompanied the LDI of the oligopeptide. Isolated gold nanorods and large nanorod aggregates did not show efficient SALDI processes. The longitudinal SP bands could be used as an indicator to obtain size-controlled nano-assemblies of the gold nanorods for efficient SALDI measurements. That is, if we select a plate that has appropriate SP bandwidth (ca. 250 nm), it will show very high SALDI sensitivity. This is a new approach to optimize nano-assemblies for efficient SALDI processes using conventional extinction spectroscopy. A next step of this work should be increasing the enhancement of the SALDI process and evaluation of the spectroscopic properties of the area of the enhancement. Our studies will contribute to the preparation of highly sensitive and reproducible SALDI plates.