Toward Subdiffraction Transmission Microscopy of Diffuse Materials

Feb 4, 2009 - (12) This pixel-type arrangement is exhibited in Figure 1d, where the top-surface of a single scale was exposed by focused ion beam (FIB...
0 downloads 4 Views 2MB Size
NANO LETTERS

Toward Subdiffraction Transmission Microscopy of Diffuse Materials with Silver Nanoparticle White-Light Beacons

2009 Vol. 9, No. 3 952-956

Debansu Chaudhuri,† Jeremy W. Galusha,‡ Manfred J. Walter,† Nicholas J. Borys,† Michael H. Bartl,*,‡ and John M. Lupton*,† Department of Physics and Department of Chemistry, UniVersity of Utah, Salt Lake City, Utah 84112 Received September 16, 2008; Revised Manuscript Received November 24, 2008

ABSTRACT We demonstrate high resolution transmission microscopy in a conventional two-photon wide-field fluorescence microscope by exploiting nonlinear white light generation from clusters of silver nanoparticles placed beneath the specimen. Surface-enhanced two-photon luminescence occurs at nanoparticle hot spots in the form of spectrally broad, spatially confined light which can be exploited to determine the transmission properties of a sample placed on the silver nanoparticles. We demonstrate the versatility of the technique by revealing individual crystalline domains formed in the diffuse biological photonic crystals of the scales of a beetle. We can identify submicron changes between photonic crystal facets as well as the occurrence of stacked domains invisible to surface-sensitive methods. Control over wavelength, polarization, and pulse shape promises selective addressing of hot spots in nanoparticle assemblies for motionless spatial scanning of the transmission properties with subdiffraction resolution.

The popular conception of optical microscopy generally involves transmission measurements, even though most contemporary high-resolution microscopes are based on fluorescence. Whereas luminescence has long surpassed the diffraction limit,1–4 high-contrast transmission spectroscopy with submicron spatial resolution is complicated by the need of broadband or tuneable light sources,5 and nanoscale apertures.6–11 Although fluorescence microscopy is unsurpassed in acquisition speed and spatial resolution,1–4 it is strictly limited to suitably autofluorescent or stained samples. This restriction makes fluorescence sensitive primarily to surfaces within a specimen (such as cell membranes) from which light emission occurs. Many material properties in the life sciences and in materials science alike are controlled by the bulk arrangement of matter, which is best probed in transmission.5 To demonstrate the need for versatile and facile high-resolution transmission microscopy, we turn to a specific example at the boundary between biology, physics and materials science. It is well known that a range of insects derive their iridescent coloration from photonic crystals formed in their cuticular exoskeleton.12–14 Two of the authors (J.W.G. and M.H.B.) recently demonstrated that the remarkable angular independence in the iridescent coloring of certain beetles such as Lamprocyphus Augustus (Figure 1a) * To whom correspondence should be addressed. E-mail: (J.M.L.) [email protected]; (M.H.B.) [email protected]. † Department of Physics. ‡ Department of Chemistry. 10.1021/nl802819n CCC: $40.75 Published on Web 02/04/2009

 2009 American Chemical Society

arises from a near-perfect overlap of photonic stop bands formed in a diamond-based photonic crystal structure,12 the “champion” photonic crystal morphology.15 Figure 1b shows an optical white-light reflection micrograph of the beetle’s exoskeleton scales, which possess an interior diamond-based photonic structure and hence are responsible for the green iridescent appearance. A corresponding transmission spectrum, derived from the microscope image shown in the inset, is displayed in panel c, exhibiting broad composite photonic stop bands between 520 and 620 nm. The reason that the beetle’s iridescence appears virtually independent of viewing angle lies in the fact that the diamond-based photonic crystal in individual scales is organized in differently oriented single-crystalline domains.12 This pixel-type arrangement is exhibited in Figure 1d, where the top-surface of a single scale was exposed by focused ion beam (FIB) milling and imaged by scanning-electron microscopy (SEM).12 Elaborate three-dimensional reconstruction of the crystal structure revealed that three dominant crystal domains of the diamond-based structure are oriented with their Γ-W, Γ-K, and Γ-X crystal axes normal or slightly off-normal to the scale top surface.12 Although band structure calculations and modeling of the photonic crystal with a lattice parameter of 450 nm (with a 10-20% variation of structural dimensions within single scales and between scales sampled from different parts of the beetle) and a dielectric constant of 2.5 for the cuticular material allows the decon-

Figure 1. Biological photonic crystals based on a cuticular exoskeleton diamond-structured lattice. (a) Photograph of the weevil Lamprocyphus Augustus. (b) Optical micrograph of iridescent beetle scales, recorded in reflection mode. (c) A typical bulk transmission spectrum of a single beetle scale exhibiting a broad composite photonic crystal stop band between 520 and 620 nm. The inset shows the white-light transmission image of a single scale. (d) SEM image of the top view of a single beetle scale, showing three distinct crystal facets. The interior photonic crystal structure was exposed by FIB milling. Images adapted from ref 12.

volution of the optical transmission spectrum (Figure 1c) into three stop bands corresponding to the observed Γ-W, Γ-K, and Γ-X crystal facets, conventional transmission microscopy cannot reveal isolated domains.12 Resolving the different crystalline domains in the specimen requires a spatial resolution in optical imaging of the order of the diffraction limit of light. Although this could, in principle, be achieved with near-field scanning microscopy,16–20 the technique does not always lend itself readily to the study of soft, inhomogeneous biological compounds. Besides, the sample itself is thick (∼10 µm) and highly scattering and therefore appears opaque in transmission (Figure 1c), implying that it would be nontrivial to couple in light from a tapered fiber.18–20 An alternative to nanoscale apertures7–10 previously explored in white-light transmission microscopy is given by broadband nanoscale light beacons excited at a wavelength to which the specimen appears transparent. Silver nanoparticle films as employed for single molecule surface-enhanced Raman scattering11,21–30 (SERS) can generate stable and spectrally broad light from subdiffraction spots,11 both under one- and two-photon excitation. Two-photon luminescence from fractal silver films involves surface plasmon-mediated enhancement of the fundamental field.11,30,31 We exploit the particularly broad luminescence generated under two-photon irradiation to probe transmission of light through a sample over a subdiffraction cross-sectional area. Figure 2a displays an SEM image of a single beetle scale placed on a silver nanoparticle film, prepared following the simple Tollens silver mirror reaction.21,22 Because of the strong van-der-Waals interactions the biological sample readily sticks to the nanoparticle film. All measurements were carried out at 300 K, in vacuo (10-6 mbar), in our homebuilt wide-field fluorescence microscopy setup22 with a longworking distance (7.7 mm) objective (N.A. 0.55). Similar Nano Lett., Vol. 9, No. 3, 2009

results were also obtainable in air at the cost of a somewhat reduced photostability of the silver nanoparticle emission. A fluorescence microscope allows three modes of imaging. Panel b of Figure 2 displays one-photon (440 nm) excitation of the beetle scale, which exhibits spectrally broad autofluorescence. Because of the thickness fluctuations over the beetle scale, the fluorescence intensity varies across the image. Panel c displays the same image under excitation at 920 nm (140 fs pulses, 80 MHz repetition rate) and detection of the back-scattered light. The Rayleigh scattering provides an image of the spatial nonuniformities on the surface of the scale, which is dominated by an amorphous cuticular shell enclosing the photonic crystal structure.12 By simply inserting a short-pass filter into the detection pathway of the microscope and blocking the Rayleigh scattered light, the broadband emission from the nanoparticle hot spots can be imaged through the beetle scale, as shown in panel d. Note that all three images show the same spatial position. Autofluorescence from the sample is not observed under twophoton (920 nm) excitation; the light detected is generated beneath the specimen and passed through it. We confirmed that the spatial distribution of the discrete spots of silver nanoparticle two-photon luminescence varied with polarization and wavelength of the incident laser, as expected for a nonlinear optical process resonantly enhanced at the fundamental frequency by coupled nanoparticle plasmons.11 Figure 2e compares the emission spectra of two different nanoparticle hot spots under two-photon excitation with and without a photonic crystal placed on top. Whereas the freespace spectrum (black) is broad and featureless, peaking at around 700 nm, the single hot spot emission transmitted through the crystal (red) exhibits significant structure while being attenuated approximately 2-fold by the scattering medium. A strong dip is observed at 530 nm, corresponding 953

Figure 2. Silver nanoparticle hot spot white-light transmission microscopy of biological photonic crystals. (a) SEM micrograph of a beetle scale on a Tollens silver SERS substrate. (b) Fluorescence image of a beetle scale at 440 nm excitation. (c) Rayleigh scattering of 920 nm light from the same beetle scale. (d) Two-photon white-light beacons shining through the beetle scale, revealing the same overall spatial structure of the scale. (e) A characteristic white-light spectrum (black) from the hot spot (typical image shown in the right-hand inset, size 3 × 3 µm2), excited at 880 nm, compared to a modified white-light spectrum (red) transmitted through the beetle scale (transmitted hot spot image shown in the left-hand inset). The transmitted light is attenuated by approximately a factor of 2 and displays a dip at 530 nm, characteristic of the Γ-W photonic crystal stop band.12 A peak is also observed at 440 nm, resulting from surface-enhanced second-harmonic generation. (f) Transmitted white-light spectra from different regions of the beetle scale indicating the presence of three distinct facets of the photonic crystal. (g) Histogram of transmission minima showing the inhomogeneous broadening of the photonic crystal stop bands due to slight structural and dielectric variations throughout the beetle scale.

to a transmission minimum. The narrow peak at 440 nm arises from surface-enhanced second-harmonic generation on the nanoparticle film. The insets in panel e show the Airy rings imaged from typical single hot spots, which appear slightly diffused in the case of transmission through the photonic crystal. As various spot locations are chosen from the hot spot transmission image (Figure 2d), different narrow minima in the broad white-light spectrum are identified (panel f). Figure 2g summarizes the data for 450 hot spots, recorded beneath 5 different beetle scales, by plotting a histogram of the transmission minima wavelengths. Three 954

distinct spectral ranges are identified, which match the calculated Γ-W, Γ-K, and Γ-X stop gaps of the diamondbased photonic crystal reported in ref 12 extremely well. The longest wavelength transmission stop bands corresponding to Γ-X top-oriented domains arise only occasionally. Because of the 10-20% lattice parameter variation of this natural photonic crystal structure observed in SEM,12 we find that the transmission minima of a particular domain vary locally, giving rise to the scatter seen in the histogram. This facile optical technique, which probes the bulk material, can therefore reveal structural inhomogeneities previously only Nano Lett., Vol. 9, No. 3, 2009

Figure 3. Spatially resolved hot spot transmission spectroscopy revealing different single-crystalline domains and crystal defects in a single beetle scale. (a) The lateral boundary between two differently oriented crystal domains is clearly identified when comparing hot spot transmission at either side of the dotted black line, separated by ∼800 nm. (b) Transmission spectra corresponding to 160 nm wide spatial regions illustrating a transition from the Γ-W to the Γ-K facet; red dotted lines in panel a mark the centers of each region. (c) Highresolution transmission microscopy can reveal features not accessible in fluorescence microscopy or SEM, such as the stacking of different crystal facets within the beetle scale. (d) This stacking results in the appearance of two stop bands at certain spatial coordinates on the sample. 800 nm wide regions are plotted, centered around the black dotted lines.

accessible with more elaborate surface-sensitive SEM methods. The hot spot emission under two-photon excitation is comparable in intensity to efficient fluorophores in single molecule experiments. It is surprising that the hot spot brightness is only attenuated approximately half-fold by the strongly scattering photonic crystal. We propose that the close proximity between the nanobeacon and the sample enables efficient near-field coupling, comparable to the case of a near-field microscope, following which the white light is effectively propagated within the photonic crystal. As the hot spot emission originates from a subdiffraction area (as documented by the Airy rings in the right-hand inset in Figure 2e), the transmission experiment can, in principle, probe subdiffraction cross sections of suitably thin samples. In the present specimen of thickness ∼10 µm we expect the light cone to broaden and diffuse. Indeed, the hot spots imaged through the beetle scale appear less sharp (Figure 2e). Nevertheless, we are able to resolve clear structural details and boundaries that are not visible with our conventional transmission microscope.12 Figure 3a shows the transmitted hot spot spectra as a function of position along the beetle scale on an intensity color-scale representation. When a hot spot is present at a given position, a dip in the emission occurs in the region 520-570 nm. Close inspection of the transmission spectra in panel b reveals a transition from the Γ-W to the Γ-K top-oriented facet. The transition between domains occurs within ∼800 nm of the dotted black line marked in the figure, which is in agreement with the general observations from electron microscopy (Figure 1d). Our technique can also identify vertical domain variations Nano Lett., Vol. 9, No. 3, 2009

within the beetle scale, which are masked in the surfacesensitive SEM. Figure 3c displays the spectra of a line of hot spots imaged along the scale. Whereas most spectra display just one discrete minimum, the two positions labeled in the figure indicate simultaneous light transmission through Γ-W and Γ-K top-oriented domains. The transmission spectra in panel d correspondingly reveal two dips at 540 and 575 nm. Our optical transmission technique combines the concepts of conventional multiphoton fluorescence with aperture microscopy with some resemblance of the initial proposal of subdiffraction imaging by Synge.32 The true power of using silver nanoparticle hot spots as spectrally broad nanoscale light sources lies in the fact that the silver particles can be selectively addressed, allowing a scanning of the sample illumination without moving the sample or the light source. Different hot spots can be excited by varying the polarization and wavelength of the light,11 and even by applying sophisticated pulse-shaping techniques that allow direct coherent control of the plasmon excitation.33,34 Even the random blinking in the hot spot emission35 can be exploited to construct a high-resolution transmission image using photobleaching localization microscopy. We expect to be able to dramatically enhance the addressability of whitelight hot spots by engineering artificial metal nanoparticle aggregates. Most high-resolution microscopy techniques are implicitly surface sensitive and often require staining, which may impede imaging capabilities due to phototoxicity. Noble metal nanoparticles, on the other hand, are more inert, although their potential has been explored mainly in the 955

context of staining.36,37 Many phenomena in the life sciences and in materials science are strictly controlled by the mesoscopic ordering of the bulk. Common examples where access to the overall morphology is crucial include tumor malignancy tests but also stress and thermal degradation processes in amorphous materials such as organic semiconductors. Our technique can image the absorption or scattering of light in volumes of subcellular cross sections. With the facile applicability to conventional multiphoton wide-field imaging microscopes, we expect nanoparticle hot spot transmission spectroscopy to provide a crucial addition to the ever-expanding palette of high-performance microscopy techniques. Acknowledgment. Funding by the Petroleum Research Fund (Grant 46795, J.M.L. and Grant 46482, M.H.B.), a DuPont Young Professor Grant (M.H.B.) and the National Science Foundation under Award No. CHE-ASC 748473 (J.M.L.) and Award No. ECS 0609244 (M.H.B.) is gratefully acknowledged, as is support by the University of Utah Research Foundation (Synergy program, J.M.L. and M.H.B.). References (1) Hell, S. W. Science 2007, 316, 1153–1158. (2) Moerner, W. E.; Fromm, D. P. ReV. Sci. Instrum. 2003, 74, 3597– 3619. (3) Michalet, X.; Kappanidis, A. N.; Laurence, T.; Pinaud, F.; Doose, S.; Pflughoefft, M.; Weiss, S. Annu. ReV. Biophys. Biomol. Struct. 2003, 32, 161–182. (4) Westphal, V.; Hell, S. W. Phys. ReV. Lett. 2005, 94, 143903. (5) Frank, J. H.; Elder, A. D.; Swartling, J.; Venkitaraman, A. R.; Jeyasekharan, A. D.; Kaminski, C. F. J. Microsc. 2007, 227, 203– 215. (6) Betzig, E.; Trautman, J. K.; Harris, T. D.; Weiner, J. S.; Kostelak, R. L. Science 1991, 251, 1468–1470. (7) Hecht, B.; Beate, S.; Wild, P. U.; Deckert, V.; Zenobi, R.; Martin, O. J. F.; Pohl, D. W. J. Chem. Phys. 2000, 112, 7761–7774. (8) Lezec, H. J.; Degiron, A.; Devaux, E.; Linke, R. A.; Martin-Moreno, L.; Garcia-Vidal, F. J.; Ebbesen, T. W. Science 2002, 97, 820–822. (9) Garcia-Vidal, F. J.; Martin-Moreno, L.; Lezec, H. J.; Ebbesen, T. W. Appl. Phys. Lett. 2003, 83, 4500–4502. (10) Stark, P. R. H.; Halleck, A. E.; Larson, D. N. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 18902–18906. (11) Shalaev, V. M. Nonlinear optics of random media: Fractal composites and metal-dielectric films; Springer: New York, 1999.

956

(12) Galusha, J. W.; Richey, L. R.; Gardner, J. S.; Cha, J. N.; Bartl, M. H. Phys. ReV. E 2008, 77, 050904. (13) Vukusis, P.; Sambles, J. R. Nature 2003, 424, 852–855. (14) Parker, A. R.; Townley, H. E. Nat. Nanotechnol. 2007, 2, 347–353. (15) Maldovan, M.; Thomas, E. L. Nat. Mater. 2004, 3, 593. (16) Pohl, D. W.; Denk, W.; Lanz, M. Appl. Phys. Lett. 1984, 44, 651– 653. (17) Michaelis, J.; Hettich, C.; Mlynek, J.; Sandoghdar, V. Nature 2000, 405, 325–328. (18) Schonbrun, E.; Wu, Q.; Park, W.; Yamashita, T.; Summers, C. J.; Abashin, M.; Fainman, Y. Appl. Phys. Lett. 2007, 90, 041113. (19) Flu¨ck, E.; van Hulst, N. F.; Vos, W. L.; Kuipers, L. Phys. ReV. E 2003, 68, 015601. (20) Koenderink, A. F.; Kafesaki, M.; Buchler, B. C.; Sandoghdar, V. Phys. ReV. Lett. 2005, 95, 153904. (21) Wang, Z.; Pan, S.; Krauss, T. D.; Du, H.; Rothberg, L. J. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 8638–8643. (22) Walter, M. J.; Lupton, J. M.; Becker, K.; Feldmann, J.; Gaefke, G.; Ho¨ger, S. Phys. ReV. Lett. 2007, 98, 137401. (23) Weiss, A.; Haran, G. J. Phys. Chem. B 2001, 105, 12348–12354. (24) Birke, R. L.; Lombardi, J. R.; Gersten, J. I. Phys. ReV. Lett. 1979, 43, 71–75. (25) Brouers, F.; Blacher, S.; Lagarkov, A. N.; Sarychev, A. K.; Patrice, G.; Shalaev, V. M. Phys. ReV. B 1997, 55, 13234–13244. (26) Dieringer, J. A.; Lettan, R. B.; Scheidt, K. A.; Van Duyne, R. P. J. Am. Chem. Soc. 2007, 129, 16249–16256. (27) Nie, S.; Emory, S. R. Science 1997, 275, 1102–1106. (28) Kneipp, K.; Wang, Y.; Kneipp, H.; Perelman, L. T.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Phys. ReV. Lett. 1997, 78, 1667–1670. (29) Shen, Y.; Friend, C. S.; Jiang, Y.; Jacubczyk, D.; Swiatkiewicz, J.; Prasad, P. N. J. Phys. Chem. B 2000, 104, 7577–7587. (30) Zolotavin, P.; Permoenova, E.; Sarkisov, O.; Nadtochenko, V.; Azouani, R.; Portes, P.; Chhor, K.; Kanaev, A. Chem. Phys. Lett. 2008, 457, 342–346. (31) Walter, M. J.; Borys, N. J.; Gaefke, G.; Ho¨ger, S.; Lupton J. M. J. Am. Chem. Soc. 2008, 130, 16830–16831. (32) Synge, E. H. Philos. Mag. 1928, 6, 356–362. (33) Stockman, M. I.; Faleev, S. V.; Bergman, D. J. Phys. ReV. Lett. 2002, 88, 067402. (34) Aeschlimann, M.; Bauer, M.; Bayer, D.; Brixner, T.; Garcia de Abajo, F. J.; Pfeiffer, W.; Rohmer, M.; Spindler, C.; Steeb, F. Nature 2007, 446, 301–304. (35) Andersen, P. C.; Jacobson, M. L.; Rowlen, K. L. J. Phys. Chem. B 2004, 108, 2148–2153. (36) Huang, X.; El-Sayed, I. H.; Qian, W.; El-Sayed, M. A. J. Am. Chem. Soc. 2006, 128, 2115–2120. (37) Yu, J.; Patel, S. A.; Dickson, R. M. Angew. Chem., Int. Ed. 2007, 46, 2028–2030.

NL802819N

Nano Lett., Vol. 9, No. 3, 2009