PHYSICAL REVIEW X 7, 031015 (2017) 1 2 2,* 3 3 4 4 4 5 F. Scarponi, S. Mattana, S. Corezzi, S. Caponi, L. Comez, P. Sassi, A. Morresi, M. Paolantoni, L. Urbanelli, 5 6 6 6 7 1 2,† C. Emiliani, L. Roscini, L. Corte, G. Cardinali, F. Palombo, J. R. Sandercock, and D. Fioretto Tablestable Ltd., Im Grindel 6, CH-8932 Mettmenstetten, Switzerland Dipartimento di Fisica e Geologia, Università di Perugia, Via Pascoli, I-06123 Perugia, Italy IOM-CNR, c/o Dipartimento di Fisica e Geologia, Università di Perugia, Via Pascoli, I-06123 Perugia, Italy Dipartimento di Chimica, Biologia e Biotecnologie, Università di Perugia, Via Elce di Sotto 8, I-06123 Perugia, Italy Dipartimento di Chimica, Biologia e Biotecnologie, Università di Perugia, Via del Giochetto, I-06123 Perugia, Italy Department of Pharmaceutical Sciences-Microbiology, University of Perugia, Borgo XX Giugno 74, 06121 Perugia, Italy University of Exeter, School of Physics and Astronomy, Exeter EX4 4QL, United Kingdom (Received 21 February 2017; revised manuscript received 8 June 2017; published 21 July 2017) Brillouin and Raman scattering spectroscopy are established techniques for the nondestructive contactless and label-free readout of mechanical, chemical, and structural properties of condensed matter. Brillouin-Raman investigations currently require separate measurements and a site-matched approach to obtain complementary information from a sample. Here, we demonstrate a new concept of fully scanning multimodal microspectroscopy for simultaneous detection of Brillouin and Raman light scattering in an exceptionally wide spectral range, from fractions of GHz to hundreds of THz. It yields an unprecedented 150-dB contrast, which is especially important for the analysis of opaque or turbid media such as biomedical samples, and spatial resolution on a subcellular scale. We report the first applications of this new multimodal method to a range of systems, from a single cell to the fast reaction kinetics of a curing process, and the mechanochemical mapping of highly scattering biological samples. DOI: 10.1103/PhysRevX.7.031015 Subject Areas: Interdisciplinary Physics, Optics, Soft Matter Brillouin and Raman spectroscopy are inelastic light investigations of fibers [3,8], nanoparticles [9,10], and scattering techniques that, by differing only in the probed structured materials [4,11], but it has only sparingly been frequency range, give complementary information on used in the biomedical field [12–15], where it can provide a mechanical and chemical properties of matter. Despite their high-frequency counterpart to traditional mechanical tech- origin being in the same years (1920s) [1,2], the two niques . Raman scattering (RS) involves light inelasti- techniques have been developed as separate tools, also in cally scattered by optical phonons or intramolecular modes, combination with optical microscopy. Brillouin light scatter- with frequency shifts typically larger than 1 THz. An RS spectrum is a truly chemical fingerprint of the material with ing (BLS) is the inelastic scattering of light from collective modes, such as acoustic waves (phonons) and spin waves information concerning the molecular composition and (magnons) propagating in condensed matter, that induce structure. RS is widely used in a number of fields ranging from analytical and physical chemistry to biophotonics and frequency shifts of the radiation in the range 0.1–100 GHz. biomedical sciences . Recently, the conventional barrier BLS spectroscopy, providing information on the elastic [3,4], viscoelastic , and magnetic properties of matter [6,7],has between the two techniques has been crumbling and increas- ing interest is devoted to simultaneous Brillouin and Raman been widely applied in condensed matter physics, including light scattering investigations. Extended depolarized light scattering (EDLS) based on separate Brillouin and Raman Corresponding author. measurements has been introduced to study solvation proc- email@example.com esses  and complex relaxation patterns in glasses  Corresponding author. and glass-forming materials . Also, Brillouin and Raman firstname.lastname@example.org microscopy have been applied through a site-matched Published by the American Physical Society under the terms of approach to obtain mechanical mapping with chemical the Creative Commons Attribution 4.0 International license. specificity of human tissues [21,22], thus opening the route Further distribution of this work must maintain attribution to to a wide range of biomedical and bioengineering applica- the author(s) and the published article’s title, journal citation, and DOI. tions. Although the contactless nature of light scattering 2160-3308=17=7(3)=031015(11) 031015-1 Published by the American Physical Society F. SCARPONI et al. PHYS. REV. X 7, 031015 (2017) considerably improves the potential for noninvasive real- the possibility to distinguish between heterogeneous time, in vivo applications, the full development of Brillouin and homogeneous broadening of Brillouin lines. microscopy has been hampered by long acquisition times Therefore, in those cases that require a detailed spectral [of the order of minutes with traditional Fabry-Pérot (FP) analysis, a scanning FP-based Brillouin system is the interferometers]. preferable method. On the Raman side, recent developments Advances in Brillouin microscopy have recently been in nonlinear techniques, e.g., coherent anti-Stokes Raman achieved by introducing a nonscanning virtually imaged scattering and stimulated Raman scattering, provide rapid phase array (VIPA) in place of the Fabry-Pérot interferom- optical sectioning for label-free imaging of biomaterials eter [14,15,23]. The low contrast of VIPA has been with limited depth of penetration. However, the reduced improved by multipass configurations , eventually used spectral resolution and the presence of a nonresonant in combination with a triple-pass Fabry-Pérot interferom- background still limit the sensitivity and hamper the eter as a bandpass filter  or a molecular or atomic gas applications of coherent anti-Stokes Raman scattering; cell as a notch filter . The latter solution has also been stimulated Raman scattering seems more promising, pro- adopted in a joint BLS-RS setup [27,28], although with vided new laser systems for the “fingerprint” region will improvements on spectral response (especially Raman) still become available. to be implemented. While VIPAs considerably improve the In this article, we present a high-resolution, high-contrast, Brillouin data gathering speed, this comes at a cost: (i) a and wide spectral range confocal microspectroscopic setup rather coarse spectral resolution, which is limited to based on a new concept of multipass FP interferometer and a ∼0.7 GHz by the fixed thickness of the etalon; (ii) a low conventional Raman spectrometer, enabling simultaneous contrast, which is 30 dB in the single-pass setup and reaches Brillouin-Raman microspectroscopy (BRMS) for a range of 85 dB in the filter-combined multipass configuration; (iii) a applications unapproachable by ordinary devices. reduced spectral range, which is limited to some tens of GHz Figure 1(a) shows a schematic of the setup. Visible by the repetition of orders in the etalon response function. light from a laser source (S) is focused onto a sample by the The VIPA-based approach remains limited to transparent same objective that is also used to collect the backscattered or moderately turbid media, e.g., tissue phantoms, because light. The sample is mounted onto a three-axes piezotrans- available spectrometers have not yet achieved the contrast lation stage (SH) for mapping measurements. A polarizing that is required to interrogate truly opaque media such as beam splitter (BS) transmits the depolarized backscattered most biological samples. Furthermore, a better resolution light through to the spectrometers. Immediately after, a of the spectrometer is needed to accurately measure the short-pass tunable edge filter (TEF) transmits the quasie- linewidth of Brillouin peaks, and to migrate from a purely lastic scattered light to a new concept of tandem Fabry- mechanical characterization (related only to the peak Pérot interferometer (TFP-2 HC; see Methods) and reflects frequency) towards a viscoelastic characterization (related the deeply inelastic scattered light into a Raman mono- to the peak frequency and linewidth) of biomaterials, with chromator (RM). (b) (a) (c) FIG. 1 (a) Schematic of the Brillouin-Raman microspectroscopy (BRMS) setup, consisting of a confocal microscope (CM-1), a tandem Fabry-Pérot interferometer (TFP-2 HC), and a Raman monochromator (RM) (see Methods for more details). (b) Scheme of the optical isolators used in the TFP-2 HC (side and front view) also represented by the small diodelike symbols in (a). The instrument uses circular polarizing techniques to prevent back reflections and so completely decouple the separate passes. (c) Spectral response of the TFP-2 HC interferometer. This yields an instrumental contrast better than 150 dB. 031015-2 HIGH-PERFORMANCE VERSATILE SETUP FOR … PHYS. REV. X 7, 031015 (2017) The tandem Fabry-Pérot interferometer, which we present Figure 2(a) shows a photomicrograph of a Candida here for the first time, reaches the unprecedented >150-dB albicans biofilm in which a multilayer cluster of round- contrast in just a 3 þ 3 pass configuration thanks to the use shaped cells is visible. The Brillouin and Raman maps of the of optical isolators, but also preserves a high luminosity sample are obtained using a 50× objective, 1-μmstep- through the use of an avalanche-photodiode detector. This size, 20 × 20 points. Spectra are collected simultaneously, configuration enables one to exploit all the scattered light, with an acquisition time of 10 s and an incident laser power of to improve by more than 50 dB the contrast with respect to 17 mW. Figure 2(d) shows that a contrast higher than 110 dB traditional 3 þ 3 pass interferometers, making it possible to is required to detect longitudinal phonons in this highly measure truly opaque samples and, eventually, to reduce the scattering sample. Raman maps are obtained by plotting the −1 acquisition time to values typical of Raman spectrometers. intensity (integral) of the bands in the range 2800–3000 cm This enables joint Brillouin-Raman rapid raster-scan map- [Fig. 2(b)], showing the prominent CH stretching mode of −1 ping and time-sampling experiments. proteins and lipids, and 1200–1530 cm [Fig. 2(c)], the so- To demonstrate the performances of the new instrument, called fingerprint region, comprising bands due to the amide here we report the results of Brillouin-Raman mapping III, cytochrome c, carbohydrates, proteins, and lipids. The of a microbial biofilm. Biofilms are ubiquitous structures observed heterogeneity in the concentration of CH groups in  formed by microbial cells growing onto solid sur- molecules across the scanned area can be attributed to a faces embedded in a polymer matrix, often made of eso- nonuniform thickness of the biofilm, with an increase in polysaccharides, produced by the cells themselves . intensity by a factor of 2–3 in the transition region from one to City of microbes  with a complex morphology, biofilms two or more Candida layers. The most interesting feature is are important because of their increased resistance to the high intensity of Raman bands in the central part of the antibiotics, antifungal drugs, and extreme conditions biofilm. In this region, a prominent maximum in the bands at −1 [5,32,33]. Despite being poorly understood, the mechanical 1200–1530 cm , whose intensity increases by a factor 20, characteristics of the biofilm are of primary interest to can not only be attributed to the multilayer structure of the elucidate the mechanisms governing the stability and the biofilm, but also to a modification in the cell status and/or in dispersion of cells within the biofilm [34,35]. the biofilm composition. In particular, we can expect the (a) (b) (c) (d) (e) (f) 19.5 18.5 18 0 FIG. 2. (a) Optical micrograph of a Candida albicans biofilm. The red box denotes the 20 × 20 μm area whereby Brillouin-Raman raster-scan maps are obtained. (b),(c) Raman map based on the integrated intensity of the bands in the region indicated in the label. (d) Brillouin spectrum extracted from the point denoted by an asterisk in (a), showing that a minimum contrast of approximately 110 dB is required to detect longitudinal phonons in this sample. The quasielastic signal (<2 GHz) is acquired through the same procedure as that used to obtain the spectrum in Fig. 1(c) (see Methods), i.e., with the laser beam attenuated by calibrated neutral density filters. (e),(f) Brillouin map based on the characteristic frequency ν (GHz) and linewidth Γ (GHz) of the Brillouin peak. 031015-3 F. SCARPONI et al. PHYS. REV. X 7, 031015 (2017) presence of zones where more extracellular polymeric Candida cells. This result is in line with the microbiological substance is present and where the survival of Candida cells evidence that the biofilm acts as a structure that increases is favored. In this respect, it is interesting to note that the the resistance of cells to challenging agents. In fact, it is marked (out-of-scale) increase of Raman signals in the range plausible that, in the region of larger thickness, the buried −1 cells are protected by the overlaying biofilm structure. The 1200–1530 cm could be at least partially attributed to the protecting layer is also able to preserve water inside the resonant scattering of cytocrome c, a marker of cell vitality inner cells, as suggested by the “softer” behavior revealed [36,37]. Crucial to understanding this feature is the notice- by Brillouin scattering. able change in the viscoelastic properties of the biofilm, A second experiment shows the potential of BRMS visible in the Brillouin maps [Figs. 2(e) and 2(f)]. An measurements to monitor fast mechanical and chemical elasticity map is inferred from the frequency map in changes occurring in reacting samples. The production Fig. 2(e) under the assumptions that the detected phonons process of technologically important materials is often are acoustic with linear dispersive behavior, and that the ratio multistage  or involves complex thermal cycles of ρ=n —with ρ the density of the sample and n its refractive reaction. Here, one major need for constructing mechan- index—is approximately constant through the scanned area ically optimized products is the synchronized determina- (see Methods). Greater care in the Brillouin spectrum tion of mechanical properties and extent of reaction interpretation is necessary when these conditions are not throughout the process. To test the capability of our system, satisfied, e.g., in microstructured materials with strong Fig. 3 shows the time evolution of Brillouin and Raman mechanical nonlinearity  or in mesoscopic structures such spectra during the isothermal polymerization of an epoxy- as opals and phononic crystals [4,11]. Refractive index and based thermoset resin made of diglycidyl ether of bisphenol density are not uniform throughout the sample; however, A (DGEBA) and diethylenetriamine (DETA), which is used ρ=n is constant with good approximation far from electronic as a matrix in high-performance composites . resonances, so that the relative variation of the frequency Brillouin spectra [Fig. 3(a)] are taken every 0.5 s, with squared ν gives a good estimation of the variation of the 30-mW laser intensity. The hardening of the resin is elastic modulus. The actual changes in refractive index and indicated by a progressive increase of frequency ν and a density typically amount to less than 2% uncertainty in the parallel decrease of linewidth Γ of the Brillouin peaks elastic modulus estimation [15,23,38], and therefore do not [Fig. 3(c)]; these parameters are related to the longitudinal affect the biomechanical characterization of the sample. In elastic modulus and longitudinal viscosity, respectively the central region of the biofilm, a ∼6% reduction of the (see Methods). Indeed, the hardening of the resin results Brillouin frequency corresponding to a ∼12% reduction in from the combination of two major effects, an increase of stiffness of the sample is observed, together with a twofold elastic energy density due to the formation of chemical increase in linewidth (related to the acoustic attenuation in bonds between DGEBA and DETA molecules and an the sample). We note that the presence of multiple scattering increase in structural relaxation time of the system as it in the Brillouin spectra of the biofilm, which is typical of approaches the glass transition. opaque scattering media, does not substantially affect the The molecular parameter we use to monitor the advance- linewidth interpretation. The combination of reduced stiff- ment of the reaction is the conversion α, i.e., the fraction ness and increased acoustic attenuation has already been of reacted epoxy rings during the curing process. Raman observed in dried biological tissues and attributed to the spectroscopy can directly access this parameter through plasticizing effect associated with an increase of residual the normalized intensity IðtÞ¼ I ðtÞ=I ðtÞ, where 1255 1610 water . In fact, both spectral changes we observe here can −1 I ðtÞ is the intensity of the epoxy band at 1255 cm be attributed to a reduction of the structural relaxation time, and I ðtÞ is the intensity of the phenyl ring stretching i.e., of local viscosity, associated to a local increase in −1 band at 1610 cm . In fact, the epoxy groups are hydration level. Note that, in the absence of independent consumed during the reaction, while phenyl rings are information on the local viscosity of the sample derived from unaffected by the curing process. The time evolution of the linewidth of Brillouin peaks, it is not possible to the conversion, measured as αðtÞ¼ 1 − IðtÞ=Ið0Þ,is distinguish between the true observed relaxational effect reported in Fig. 3(d) together with the result of a poly- and an alternative fallacious interpretation in terms of a nomial fit. A transition is observed in the evolution of ν merely elastic effect related to, e.g., a reduction in microfilm versus α at 8 min from the beginning, i.e., at about α ¼ 0.49 connectivity. At the same time, without foundational molecu- [Fig. 3(e)]. This decrease in the rate of change of the sound lar insight provided by simultaneous Raman spectra, it is velocity occurs in the reaction range where the sol-gel not possible to adequately interpret the observed micro- transition is expected (α ¼ 0.5 according to the Flory- mechanical properties, which makes the fully integrated gel Brillouin-Raman setup needed and necessary. Stockmayer theory ) and the system also goes through The observed physicochemical heterogeneity (sample the glass transition. In the absence of relaxation processes, thickness, viscoelastic behavior, and chemical modifica- ν is proportional to the elastic modulus (see Methods) that, tion) is compatible with the presence of a region with live in turn, is proportional to the energy density of chemical 031015-4 (arb. units) HIGH-PERFORMANCE VERSATILE SETUP FOR … PHYS. REV. X 7, 031015 (2017) (a) (b) (c) (d) (e) FIG. 3 (a) Brillouin and (b) Raman spectra of an epoxy-amine mixture (DGEBA-DETA 5:2) measured at different times of the isothermal polymerization process, at T ¼ 65 °C: red symbols, 1 s; green symbols, 200 s; blue symbol, 1000 s since the beginning. Lines are the results of curve-fit analysis applied to the Brillouin peaks (see Methods). (c) Temporal evolution of the characteristic frequency (ν) and linewidth (Γ) of Brillouin peaks. (d) Time evolution of the conversion α calculated from the intensity of Raman peaks (see text). (e) Plot of the Brillouin frequency squared versus α. The sol-gel transition is expected to occur at α ¼ 0.5. gel bonds, i.e., to the number of bonds formed during the various topics of condensed matter, going from hydration reaction. This condition is better satisfied in the second part of biological systems  to anharmonicity and boson peak in glasses  and highly viscous media [19,49].EDLS of the reaction, and can explain the linear behavior of ν measurements, until now carried out by means of separate versus α for α > 0.49. Conversely, the presence of the dispersive and interferometric setups, can now be per- structural relaxation and its evolution during the reaction formed with a unique instrument, simultaneously, and with , attested by the variation of Γ in Fig. 3(c), is possibly high spatial resolution, paving the way to the development responsible for the faster variation of ν versus α in the first of micro-EDLS. part of the reaction. How the complex interplay between Figure 4 shows the susceptibility representation of the structural relaxation and bond formation can generate the EDLS spectrum of a NIH/3T3 murine fibroblast, measured linear behavior of ν versus α for α < 0.49 is still unclear. If in the central region of the nucleus. The solvent-free this is found to be general, the frequency of Brillouin peaks spectrum obtained after the subtraction of the solvent could be used as an effective probe for the advancement of contribution is also reported, together with the spectrum a polymerization reaction. previously obtained for an aqueous solution of a protein While the separate monitoring of polymerization proc- using a conventional setup . The spectrum from the esses through Raman  and Brillouin scattering goes nucleus of the cell is dominated by a strong vibrational back a long way [45,46], to the best of our knowledge, this contribution around 1 THz. In addition to possible local is the first simultaneous measurement of chemical and vibrations in the complex structure of the nucleus, this mechanical evolution during a polymerization reaction. contribution can be associated to the boson peak ,a A third experiment benefits from the wide tunable spectral ubiquitous signature of collective vibrations in disordered range accessible to the tandem Fabry-Pérot interferometer and its polarization selectivity to perform extended depo- condensed matter . Conversely, the strong contribution larized light scattering investigations of biological matter. at 2–3 THz in the spectrum of the protein is mainly EDLS spectroscopy has been recently developed to attributed to the libration of methyl groups . The access both vibrational and relaxation dynamics of matter number of methyl groups in DNA is much smaller than in the wide frequency range between fractions of GHz and in proteins , and this can explain the suppression of the 2-THz contribution in the spectrum of the nucleus. Worthy tens of THz . In this range, it is possible to deal with 031015-5 F. SCARPONI et al. PHYS. REV. X 7, 031015 (2017) picoliter sample volumes); (iii) the collective dynamics of biological matter by extended depolarized light scattering with subcellular spatial resolution. This noninvasive method is now available for a variety of applications, opening the route for in situ investigation of diverse phenomena in material sciences and potentially for in vivo diagnosis of pathologies involving mechanical changes as well as altered structure and composition. Methods: Experimental setup.—Figure 1(a) shows a schematic of the experimental arrangement for Brillouin- Raman measurements. A 532-nm single-mode solid-state laser (SpectraPhysics Excelsior) is employed as the light source (S). The vertically polarized laser beam is first sent through a JRS Scientific Instruments TCF-1 temperature- controlled etalon, which reduces the intensity of laser sidelobes by a factor of about 600, and then to the input of a customized JRS Scientific Instruments CM-1 confocal microscope, specifically developed in collaboration with the manufacturer to allow Raman-Brillouin signal splitting. FIG. 4 Top: EDLS susceptibility of a fixed NIH/3T3 fibroblast cell line (red profile) and PBS solution (cyan profile), at Inside the microscope, the laser beam is first expanded T ¼ 25 °C. Each broadband spectral profile has been obtained and collimated (BE), then a broadband polarizing beam merging five interferometric spectra and one Raman spectrum in splitter (BS) sends it to an infinity-corrected apochromatic distinct measurements. Inset: Image of a NIH/3T3 single cell microscopic objective. In the present experiments, we use a acquired with a 50× Mitutoyo objective. Bottom: Solvent-free Mitutoyo M-Plan Apo 20× with a very long working spectrum obtained by subtracting the spectral profile of the PBS distance of 20 mm, a 2-μm depth of field, and a limiting solvent from that of the cell (red profile). The solvent-free numerical aperture of 0.42, and a Mitutoyo M-Plan Apo spectrum of a 100-mg=ml lysozyme aqueous solution  is 50× with working distance of 13 mm and numerical also shown for comparison (brown line). aperture of 0.55. The objective is used both to focus the laser beam onto the sample and to collect the backscattered of note is also the GHz range, where the spectrum is light. The laser spot diameter on the surface is ∼2 μm noticeably less intense than in the water-protein solution. measured with the 20× objective. Illumination of the In fact, the relaxation processes in DNA are known to be sample surface is obtained by means of a coaxial LED about 1 order of magnitude slower than in proteins ; illuminator (LI), providing blue light peaked at about hence, their spectral signature is out of this spectral range. 470 nm. A broadband λ=4 retarder wave plate can be The residual intensity in the 10–100-GHz region can be inserted upstream of the objective to switch from depolar- attributed to the relaxation of hydration water, which is ized to unpolarized scattering configuration. dynamically retarded by a factor greater than 3 with respect The polarizing beam splitter forwards the depolarized to bulk water. backscattered light towards the spectrometers. In the To our knowledge, BRMS gives the first evidence of collimated beam section of the microscope, a tunable boson peak and relaxational contributions from hydration ultrasteep short-pass filter (TEF, Semrock SP01-561RU) water within the nucleus of a single cell. These picosecond transmits the anti-Stokes quasielastic scattered light to a motions have been proposed to be responsible for medi- JRS Scientific Instruments TFP-2 HC Fabry-Pérot inter- ating biochemical reactions and energy transport in bio- ferometer and reflects the Stokes deeply inelastic scattered logical matter [55,56] and for controlling drug intercalation light towards the Horiba iHR320 Triax Raman spectrom- in DNA . eter (RM). The Fabry-Pérot spectrometer includes a narrow In conclusion, we demonstrate rapid and simultaneous band pass filter (NBF) to rule out broadband light signals, Brillouin and Raman experiments, with high contrast, high while a further RazorEdge long-pass filter (LPF) improves resolution, and extended spectral range. The capabilities of the rejection of residual elastic contributions on the Raman the novel technique are tested in three case studies: (i) the in situ raster-scan mapping of viscoelastic and chemical signal line. properties of biofilms, with the potential of revealing Given the different cross section for Raman and Brillouin “persister” cells, i.e., cells that are able to survive in the phenomena, Brillouin spectrometers are usually equipped presence of devitalizing agents and are amongst the major with the most sensitive detectors, namely, photomultipliers causes of hospital infections; (ii) the simultaneous mon- operating in the single-photon regime. In recent times, itoring of mechanical and chemical changes in reacting technologic improvements have led to the development materials with subsecond time resolution (and fractions of of single-photon avalanche photodiodes (SPAD), reaching 031015-6 [arb. units] HIGH-PERFORMANCE VERSATILE SETUP FOR … PHYS. REV. X 7, 031015 (2017) much higher quantum efficiencies than conventional photo- tuned to the laser frequency. The mirror spacing of the tubes. In our setup, a LaserComponents COUNT®-10 spectrometer is set at 2.20 mm in order to match the FSR of SPAD detector is used as a sensor inside the TFP-2 HC the etalon, and small input and output pinholes are used to spectrometer. This device provides a quantum efficiency of reduce contributions from uncollimated light. The output ∼70% at 532 nm, together with a maximum noise rate of power of the laser beam (∼70 mW) is attenuated by means 10 cps. The achievable signal-to-noise ratio is comparable of a set of neutral density filters and a variable one. with that of classical phototube detectors, with an approxi- Further investigations are in progress to understand the −15 mate fivefold gain in terms of data-gathering speed. origin of the residual features visible in Fig. 1(c) at the 10 Pin holes at the entrance of the TFP spectrometer provide level. The tiny inelastic structures may be attributed to a confocal arrangement for microscopic imaging: the broad- residual sidelobes of the original laser spectrum and also to band light gathered by the objective is imaged onto a USB spurious Brillouin scattering from optical components in the CMOS sensor camera (SC), which allows us to visualize the first part of the spectrometer. We can thus anticipate that the sample surface. The field-of-view diameter depends on the true contrast of the interferometer is greater than 150 dB. objective focal length: using the 20× objective, it is ∼87 μm. Moreover, it has to be noticed that the contrast [Fig. 1(c)] The use of both a polarizer beam splitter and a laser line OD6 is measured for vertically polarized light, because of the notch filter before the camera enables the image to be seen polarization selectivity of the instrument. Supposing to with no loss of signal during the joint measurements. change the polarization input, as for the effect of depolar- For mapping measurements, the sample is mounted onto ized elastic scattering, the five polarizers of internal groups a PI 611-3S Nanocube XYZ closed-loop piezoelectric of TFP-2 would strongly suppress the light intensity, while translation stage (SH), with resolution of 1 nm and a the instrumental contrast (related only to the Fabry-Pérot motion range of 100 μm per axis. The high voltage needed mirrors reflectance) would remain the same. This selectiv- by the positioner is generated by a PI E-664 controller, ity makes the instrument better suited for the study of voltage controlled through a homemade DAC=802.11 biological samples, owing to the intrinsically lower sensi- wireless control board. A software scripting system was tivity to depolarized elastic scattering. developed starting from the JRS automated measurement We summarize the characteristics of the new setup as application in order to execute sequences of simultaneous follows. Raman-Brillouin measurements and stage positioning oper- Contrast and luminosity.—The 150-dB contrast docu- ations. LabSpec 5 software is used for acquisition of the mented in Fig. 1(c) is made possible not only by the six-pass Raman data and JRS GHOST 7 for the Brillouin data. geometry of the interferometer, but also by the adoption of The TFP-2 HC is a very high contrast variant of the multiple optical isolators inside the device [Fig. 1(b)], which original Sandercock-type tandem multipass interferometer prevent back propagation of light inside the interferometer. . The tandem configuration considerably increases the The unprecedented contrast reached by the TFP-2 HC is by spectral range available for Brillouin and depolarized far superior to that of previous 3 þ 3 pass tandem Fabry- light scattering experiments, through the suppression of Pérot interferometers (∼100 dB for the TFP-1) .In replicas of inelastic signals due to the side orders. In the addition, the luminosity of our setup is ∼0.2 which, together high-contrast model, polarization and retardation optics with the factor 5 improvement in the efficiency of the [Fig. 1(b)] are used at each pass of the interferometer to detector, opens up the possibility to map the viscoelastic effectively decouple each pass from the next one, and a properties of truly opaque or even reflecting samples. spatial filter (SF) is used to optically decouple the first Resolution.—The full width at half height of the elastic Fabry-Pérot interferometer from the second one, improving line [Fig. 1(c)] gives the spectral resolution, which is related the extinction by a factor higher than 10 . Before each to the FPs’ mirror spacing and to their optical quality. With passage through a Fabry-Pérot interferometer, light passes mirror spacing of 15 mm or higher, a value of linewidth through a 45° polarizer and is back reflected with circular better than 0.1 GHz can be obtained. This is more than a polarization by a triangular prism. Light reflected from the factor 7 better than in VIPAs, enabling the study of acoustic next optical surface has opposite chirality and emerges at attenuation and relaxation processes in viscoelastic media. the polarizer with crossed polarization and is stopped. The Acquisition time.—The high efficiency of the setup in adoption of this method also allows the manufacturer to use collecting quasielastic scattered light, by using a polarizing a perfectly orthogonal incidence of light on all surfaces, cube and an edge filter as beam splitters and an avalanche thus aiming at the theoretical contrast limit of 10 . photodiode as the photodetector, enables the acquisition The laser spectrum in Fig. 1(c) is measured normalizing time for a single spectrum to be reduced to seconds. It is and merging several spectra of a laser beam at different thus possible to measure within a reasonable time both values of the input power. During these measurements, the collective (by BLS) and single-particle (by RS) properties sidelobes of the laser spectrum are attenuated by means of a of materials that are heterogeneous in space (e.g., biological series of two JRS TCF-1 temperature-controlled etalons, matter or geological samples) or changing over time (e.g., with a free spectral range (FSR) of about 68 GHz, both reactive or degradable systems). 031015-7 F. SCARPONI et al. PHYS. REV. X 7, 031015 (2017) Spectral range.—The tandem configuration is very purposes. Conversely, long and accurate measurements effective in suppressing (better than 40 dB) the replicas would be needed to obtain the shape of the relaxation of the transmission peaks of the two FP interferometers. By functions from the band shape of Brillouin spectra, extend- combining spectra recorded with mirror spacing in the ing far from the resonance . range of 20–0.2 mm, a spectral range from fractions of GHz Analysis of EDLS spectra.—The total EDLS spectrum to more than 1 THz can be obtained. The spectral range is reported in Fig. 4 is obtained by joining the experimental further extended to the high-frequency side by means of profiles collected by means of the interferometric and dispersive part of the setup. The interferometric portion the Raman spectrometer, which can operate in the range of the spectrum (1–2000 GHz) derives from the combina- 0.9–100 THz. The possibility of collecting a continuous spectrum in such a wide and informative range is not tion of different mirror separations, namely, d ¼ 14,4,1, possible with other setups, even with the most recent 0.5, and 0.2 mm; for the latter distance, three FSRs are used Raman-Brillouin ones , due to the very limited spectral in order to reach the thousands of GHz and the overlapping range accessible to VIPAs. with the dispersive segment. The high-frequency part of the Analysis of Brillouin spectra.—Brillouin spectra pre- spectrum (from about 900 GHz) has been recorded by sented in Figs. 2 and 3 originate from light inelastically using a 1800-l=mm grating with an aperture slit of 300 μm at the entrance of the monochromator. After the removal of scattered by thermally activated acoustic modes. In a backscattering experiment from ideal homogeneous and the dark count contributions, low- and high-frequency elastic solids, two single Brillouin peaks can be revealed at portions are spliced, benefiting from a large overlapping. frequency shifts ν given by ω ¼ 2πν ¼Vq, where V is Further, the imaginary part of the dynamic susceptibility the velocity of the longitudinal acoustic modes, q ¼ 2nk is χ ðνÞ is calculated as the ratio between the intensity of the momentum exchanged in the scattering process, n is the depolarized light I ðνÞ and ½nðνÞþ 1, where nðνÞ¼ VH −1 refractive index of the sample, and k is the wavevector of exp ½ðhν=k TÞ − 1 is the Bose-Einstein occupation incident light. In ideally elastic samples, the longitudinal number. elastic modulus M is simply given by M ¼ ρV , where ρ is Materials.—The Candida albicans CMC 1768 strain the mass density of the sample. In the case of viscoelastic we use in this study is isolated from abdominal fluid of a materials, damping mechanisms reduce the lifetime of patient in the Intensive Care Unit in Pisa Hospital. The acoustic modes, and the Brillouin spectrum can be identification is carried out by comparing the ITS and LSU described by a damped harmonic oscillator function: D1/D2 sequence with GenBank  and specific databases [62,63]. Biofilms are obtained by submerging disks of I ω Γ aluminum foil (4-mm diameter) in Petri dishes containing 0 b IðωÞ¼ · ; ð1Þ 2 2 2 2 2 YEPD medium (yeast extract 1%, peptone 1%, dextrose π ðω − ω Þ þ ω Γ b b 2% from Difco Laboratories, Detroit, MI) inoculated with where ω and Γ approximately give the frequency position yeast cells at OD ¼ 0.2. Petri dishes are placed in a b b 600 rocking incubator at 37 °C for 72 h, then samples are and the linewidth (FWHM) of Brillouin peaks. These washed twice with distilled sterile water and dried at room parameters are related to the real (M ) and imaginary temperature for 1 week prior to analysis. (M ¼ ωη ) parts of the longitudinal elastic modulus The epoxy-amine mixture we use in this study is made through the relations: of diglycidyl ether of bisphenol A (DGEBA, molecular 0 2 2 2 M ¼ ρω =q ; η ¼ ρΓ =q ; ð2Þ weight 348) and diethylenetriamine (DETA, purity ≥ 99%) L b in the 5:2 stoichiometric ratio. The two monomer types are Note that the elastic modulus can be expressed as mutually reactive and polymerize by stepwise addition of the 0 2 amino hydrogen to the epoxy group, with a rate of reaction M ¼ ρðλν=2nÞ , where λ is the wavelength of laser light. strongly controlled by the temperature. The mixture is The simple hydrodynamic model described by Eq. (1), prepared by mixing for 2 min at room temperature and then convoluted with the instrumental function, is appropriate to it is transferred into the measurement cell, keeping it at a fit the whole Brillouin spectrum of materials (liquids or fixed T ¼ 65 °C. Because of the presence of a multifunc- solids) with internal relaxation rates much faster than ω . In the case of relaxations close to ω , generalized hydro- tional reagent, the reaction of DGEBA with DETA grows branched molecules and finally yields a network-polymer dynamic models can be used to model the spectrum, glass. Before reaching the vitrification point, the system introducing a frequency dependence either of the modulus goes through the gel point, where a percolating network M or, equivalently, of the longitudinal viscosity η . of bonded particles has developed. The extent of reaction Luckily enough, also in this case Eq. (1) can be used to fit at the gel point, according to the classical theory of Flory- the spectrum in correspondence to the peak maximum and, Stockmayer ,is α ¼ 0.50 for our system. through Eq. (2), to obtain the real and imaginary part of the gel modulus at the single frequency of the Brillouin peak . The NIH/3T3 murine fibroblast cell line was purchased This fast procedure is the most appropriate for mapping from American Type Culture Collection (ATCC). Cells are 031015-8 HIGH-PERFORMANCE VERSATILE SETUP FOR … PHYS. REV. X 7, 031015 (2017)  T. Still, M. Mattarelli, D. Kiefer, G. Fytas, and M. grown in Dulbecco Modified Eagle’s Medium (DMEM) Montagna, Eigenvibrations of Submicrometer Colloidal containing 10% (v/v) heat-inactivated fetal bovine serum Spheres, J. Phys. Chem. Lett. 1, 2440 (2010). (FBS), 100 u=ml penicillin, 100 u=ml streptomycin  D. Schneider, P. J. Beltramo, M. Mattarelli, P. Pfleiderer, (SIGMA Aldrich, St. Louis, MO) and maintained at J. Vermant, D. Crespy, M. Montagna, E. M. Furst, and G. 37 °C in a 5% CO humidified atmosphere. Cells are seeded Fytas, Elongated Polystyrene Spheres as Resonant Building onto CaF substrates sterilized using 100% ethanol and UV Blocks in Anisotropic Colloidal Crystals, Soft Matter 9, irradiation, treated with poly-L-lysine solution. Fixation is 9129 (2013). performed by incubating cells with 4% paraformaldehyde  W. Cheng, J. Wang, U. Jonas, G. Fytas, and N. Stefanou, in phosphate buffer saline (PBS) solution for 10 min at Observation and Tuning of Hypersonic Bandgaps in room temperature, then cells are washed twice with PBS. Colloidal Crystals, Nat. Mater. 5, 830 (2006).  J. M. Vaughan and J. T. Randall, Brillouin Scattering, Density and Elastic Properties of the Lens and Cornea ACKNOWLEDGMENTS of the Eye, Nature (London) 284, 489 (1980).  F. Palombo, C. P. Winlove, R. S. Edginton, E. Green, N. S. Corezzi acknowledges financial support from MIUR- Stone, S. Caponi, M. Madami, and D. Fioretto, Biomechan- PRIN (Project No. 2012J8X57P). S. Caponi acknowledges ics of Fibrous Proteins of the Extracellular Matrix Studied support from PAT (Provincia Autonoma di Trento) (GP/ by Brillouin Scattering, J. R. Soc. Interface 11, 20140739 PAT/2012) “Grandi Progetti 2012” Project “MaDEleNA.” (2014). P. S., A. M., M. P. acknowledge financial support from  G. Antonacci , R. M. Pedrigi, A. Kondiboyina, V. V. Mehta, Centro Nazionale Trapianti (Project: “Studio di cellule per R. de Silva, C. Paterson, R. Krams, and P. Török Quanti- uso clinico umano, con particolare riferimento a modelli fication of Plaque Stiffness by Brillouin Microscopy in cellulari (liposomi) e linee cellulari in interazione con Experimental Thin Cap Fibroatheroma, J. R. Soc. Interface crioconservanti e con materiali biocompatibili”). L. C. 12, 20150843 (2015). and S. Caponi acknowledge financial support from  G. Scarcelli, W. J. Polacheck, H. T. Nia, K. Patel, A. J. Grodzinsky, R. D. Kamm, and S. H. Yun, Noncontact Consiglio Nazionale delle Ricerche-Istituto Officina dei Three-Dimensional Mapping of Intracellular Hydro- Materiali. F. P. acnowledges support from the UK mechanical Properties by Brillouin Microscopy, Nat. Meth- Engineering and Physical Sciences Research Council ods 12, 1132 (2015). (Grant No. EP/M028739/1 (F. P.)). The authors acknowl-  M. Kawabe, K. Fukui, M. Matsukawa, M. Granke, A. Saïed, edge Jacopo Scarponi for valuable help in setting up the Q. Grimal, and P. Laugier, Comparative Investigation of hardware and software system for simultaneous Raman and Elastic Properties in a Trabecula Using Micro-Brillouin BLS measurements. Scattering and Scanning Acoustic Microscopy, J. Acoust. Soc. Am. 132, EL54 (2012).  I. R. Lewis and H. G. M. Edwards, Handbook of Raman Spectroscopy: From the Research Laboratory to the Proc- ess Line (Marcel Dekker, New York, 2001).  L. Brillouin, Diffusion de la Lumiere et des Rayonnes X par  S. Perticaroli, L. Comez, M. Paolantoni, P. Sassi, A. un Corps Transparent Homogene; Influence del’Agitation Morresi, and D. Fioretto, Extended Frequency Range Depo- Thermique., Ann. Phys. (Paris) 9, 88 (1922). larized Light Scattering Study of N-Acetyl-leucine-methyla-  C. V. Raman and K. S. Krishnan, The Optical Analogue of mide–Water Solutions, J. Am. Chem. Soc. 133,12063 the Compton Effect, Nature (London) 121, 711 (1928). (2011).  K. J. Koski, Non-Invasive Determination of the Complete  L. Hong, B. Begen, A. Kisliuk, S. Pawlus, M. Paluch, Elastic Moduli of Spider Silks, Nat. Mater. 12, 262 (2013). and A. P. Sokolov, Influence of Pressure on Quasielastic  D. Schneider et al., Nonlinear Control of High-Frequency Scattering in Glasses: Relationship to the Boson Peak, Phonons in Spider Silk, Nat. Mater. 15, 1079 (2016). Phys. Rev. Lett. 102, 145502 (2009).  L. Comez, C. Masciovecchio, G. Monaco, and D. Fioretto,  N. J. Tao, G. Li, X. Chen, W. M. Du, and H. Z. Cummins, in Solid State Physics, 1st ed. (Elsevier, New York, 2012), Low-Frequency Raman-Scattering Study of the Liquid- Vol. 63, pp. 1–77. Glass Transition in Aqueous Lithium Chloride Solutions,  S. Tacchi et al., Universal Dependence of the Spin Wave Phys. Rev. A 44, 6665 (1991). Band Structure on the Geometrical Characteristics of Two-  F. Palombo, M. Madami, N. Stone, and D. Fioretto, Dimensional Magnonic Crystals, Sci. Rep. 5, 10367 (2015). Mechanical Mapping with Chemical Specificity by Confocal  M. Madami, S. Bonetti, G. Consolo, S. Tacchi, G. Carlotti, Brillouin and Raman Microscopy, Analyst 139, 729 G. Gubbiotti, F. B. Mancoff, M. A. Yar, and J. Åkerman, (2014). Direct Observation of a Propagating Spin Wave Induced by  F. Palombo, M. Madami, D. Fioretto, J. Nallala, H. Barr, Spin-Transfer Torque, Nat. Nanotechnol. 6, 635 (2011). A. David, and N. Stone Chemico-Mechanical Imaging of  J.-C. Beugnot, S. Lebrun, G. Pauliat, H. Maillotte, V. Laude, Barrett’s Oesophagus, J. Biophoton. 9, 694 (2016). and T. Sylvestre, Brillouin Light Scattering from Surface  G. Scarcelli and S. H. Yun, Confocal Brillouin Microscopy Acoustic Waves in a Subwavelength-Diameter Optical for Three-Dimensional Mechanical Imaging, Nat. Photon- Fibre, Nat. Commun. 5, 5242 (2014). ics 2, 39 (2008). 031015-9 F. SCARPONI et al. PHYS. REV. X 7, 031015 (2017)  G. Scarcelli and S. H. Yun, Multistage VIPA Etalons  R. Hardis, J. L. P. Jessop, F. E. Peters, and M. R. Kessler, for High-Extinction Parallel Brillouin Spectroscopy, Opt. Cure Kinetics Characterization and Monitoring of an Epoxy Express 19, 10913 (2011). Resin Using DSC, Raman Spectroscopy, and DEA, Com-  A. Fiore, J. Zhang, P. Shao, S. H. Yun, and G. Scarcelli, posites, Part A 49, 100 (2013). High-Extinction Virtually Imaged Phased Array-Vased  W. H. Stockmayer, Molecular Distribution in Condensation Brillouin Spectroscopy of Turbid Biological Media, Appl. Polymers, J. Polym. Sci. A 9, 69 (1952).  S. Corezzi, D. Fioretto, and P. Rolla, Bond-Controlled Phys. Lett. 108, 203701 (2016).  Z. Meng, A. J. Traverso, and V. V. Yakovlev, Background Configurational Entropy Reduction in Chemical Vitrifica- Clean-up in Brillouin Microspectroscopy of Scattering tion, Nature (London) 420, 653 (2002). Medium, Opt. Express 22, 5410 (2014).  B. Chu, G. Fytas, and G. Zalczer, Study of Thermal  A. J. Traverso, J. V. Thompson, Z. A. Steelman, Z. Meng, Polymerization of Styrene by Raman Scattering, Macro- M. O. Scully, and V. V. Yakovlev Dual Raman-Brillouin molecules 14, 395 (1981). Microscope for Chemical and Mechanical Characterization  S. Kondo, T. Igarashi, and T. Nakamura, Rayleigh-Brillouin Scattering from Reacting Epoxy Resins: Comparison with and Imaging, Anal. Chem. 87, 7519 (2015).  Z. Meng, S. C. Bustamante Lopez, K. E. Meissner, and Torsional Braid Analysis, J. Appl. Polym. Sci. 26, 2337 V. V. Yakovlev, Subcellular Measurements of Mechanical (1981). and Chemical Properties Using Dual Raman-Brillouin  I. Alig, D. Lellinger, K. Nancke, A. Rizos, and G. Fytas, Microspectroscopy, J. Biophoton. 9, 201 (2016). Dynamic Light Scattering and Ultrasonic Investigations  S. Y.-E. Tan, S. C. Chew, S. Y.-Y. Tan, M. Givskov, and L. during the Cure Reaction of an Epoxy Resin, J. Appl. Yang, Emerging Frontiers in Detection and Control of Polym. Sci. 44, 829 (1992).  L. Comez, L. Lupi, A. Morresi, M. Paolantoni, P. Sassi, and Bacterial Biofilms, Curr. Opin. Biotechnol. 26, 1 (2014).  C. E. Zobell, The Effect of Solid Surfaces upon Bacterial D. Fioretto, More Is Different: Experimental Results on the Activity, J. Bacteriol. 46, 39 (1943). Effect of Biomolecules on the Dynamics of Hydration Water,  Y. A. Nikolaev and V. K. Plakunov, Biofilm—”City of J. Phys. Chem. Lett. 4, 1188 (2013). Microbes” or an Analogue of Multicellular Organisms?,  S. Caponi, A. Fontana, F. Rossi, G. Baldi, and E. Fabiani, Microbiology 76, 125 (2007). Effect of Temperature on the Vibrational Density of States  K. Drescher, Y. Shen, B. L. Bassler, and H. A. Stone, in Vitreous SiO : A Raman Study, Phys. Rev. B 76, 092201 Biofilm Streamers Cause Catastrophic Disruption of (2007). Flow with Consequences for Environmental and Medical  S. Caponi, S. Corezzi, D. Fioretto, A. Fontana, G. Monaco, Systems, Proc. Natl. Acad. Sci. U.S.A. 110, 4345 (2013). and F. Rossi, Raman-Scattering Measurements of the  L. Corte et al., Exploring Ecological Modelling to Inves- Vibrational Density of States of a Reactive Mixture During tigate Factors Governing the Colonization Success in Polymerization: Effect on the Boson Peak, Phys. Rev. Lett. Nosocomial Environment of Candida Albicans and Other 102, 027402 (2009). Pathogenic Yeasts, Sci. Rep. 6, 26860 (2016).  S. Perticaroli, L. Comez, M. Paolantoni, P. Sassi, L.  M. Trejo, C. Douarche, V. Bailleux, C. Poulard, S. Mariot, Lupi, D. Fioretto, A. Paciaroni, and A. Morresi, Broad- C. Regeard, and E. Raspaud, Elasticity and Wrinkled band Depolarized Light Scattering Study of Diluted Morphology of Bacillus subtilis Pellicles, Proc. Natl. Acad. Protein Aqueous Solutions, J. Phys. Chem. B 114, 8262 Sci. U.S.A. 110, 2011 (2013). (2010).  R. Janissen, D. M. Murillo, B. Niza, P. K. Sahoo, M. M.  P. C. Painter, L. Mosher, and C. Rhoads, Low-Frequency Nobrega, C. L. Cesar, M. L. A. Temperini, H. F. Carvalho, Modes in the Raman Spectrum of DNA, Biopolymers 20, A. A. de Souza, and M. A. Cotta, Spatiotemporal Distribu- 243 (1981). tion of Different Extracellular Polymeric Substances and  T. S. Grigera, V. Martín-Mayor, G. Parisi, and P. Verrocchio, Filamentation Mediate Xylella fastidiosa Adhesion and Phonon Interpretation of the “Boson Peak” in Supercooled Biofilm Formation, Sci. Rep. 5, 9856 (2015). Liquids, Nature (London) 422, 289 (2003).  M. Okada, N. I. Smith, A. F. Palonpon, H. Endo, S. Kawata,  S. Perticaroli et al., Painting Biological Low-Frequency M. Sodeoka, and K. Fujita, Label-Free Raman Observation Vibrational Modes from Small Peptides to Proteins, Phys. of Cytochrome c Dynamics during Apoptosis, Proc. Natl. Chem. Chem. Phys. 17, 11423 (2015). Acad. Sci. U.S.A. 109, 28 (2012).  S. Khodadadi and A. P. Sokolov, Protein Dynamics: From  S. Caponi et al., Raman Micro-Spectroscopy Study of Living Rattling in a Cage to Structural Relaxation, Soft Matter 11, SH-SY5Y Cells Adhering on Different Substrates, Biophys. 4984 (2015). Chem. 208, 48 (2016).  D. A. Turton, H. M. Senn, T. Harwood, A. J. Lapthorn,  G. Scarcelli, P. Kim, and S. H. Yun, In Vivo Measurement of E. M. Ellis, and K. Wynne Terahertz Underdamped Vibra- Age-Related Stiffening in the Crystalline Lens by Brillouin tional Motion Governs Protein-Ligand Binding in Solution, Optical Microscopy, Biophys. J. 101, 1539 (2011). Nat. Commun. 5, 3999 (2014).  K. Ioannidou, M. Kanduč, L. Li, D. Frenkel, J. Dobnikar,  S. Perticaroli, J. D. Nickels, G. Ehlers, and A. P. Sokolov, and E. D. Gado, The Crucial Effect of Early-Stage Gelation Rigidity, Secondary Structure, and the Universality of the on the Mechanical Properties of Cement Hydrates, Nat. Boson Peak in Proteins, Biophys. J. 106, 2667 (2014). Commun. 7, 12106 (2016).  K.-C. Chou and B. Mao, Collective Motion in DNA and  J. Suhr, N. Koratkar, P. Keblinski, andP. Ajayan, Viscoelasticity Its Role in Drug Intercalation, Biopolymers 27, 1795 in Carbon Nanotube Composites, Nat. Mater. 4, 134 (2005). (1988). 031015-10 HIGH-PERFORMANCE VERSATILE SETUP FOR … PHYS. REV. X 7, 031015 (2017)  F. Nizzoli and J. R. Sandercock, Dynamical Properties  http://blast.ncbi.nlm.nih.gov/Blast.cgi. of Solids (North-Holland, Amsterdam, 1990), Vol. 6,  L. Irinyi et al., International Society of Human and Animal pp. 281–335. Mycology (ISHAM)-ITS Reference DNA Barcoding  N. J. Tao, G. Li, and H. Z. Cummins, Brillouin-Scattering Database—The Quality Controlled Standard Tool for Study of the Liquid-Glass Transition in CaKNO : Mode- Routine Identification of Human and Animal Pathogenic Coupling Analysis, Phys. Rev. B 45, 686 (1992). Fungi, Med. Mycol. 53, 313 (2015).  D. Fioretto, L. Comez, G. Socino, L. Verdini, S. Corezzi,  C. L. Schoch et al., Finding Needles in Haystacks: Linking and P. A.Rolla, Dynamics of Density Fluctuations of a Scientific Names, Reference Specimens and Molecular Data Glass-Forming Epoxy Resin Revealed by Brillouin Light for Fungi, Oxford Database (University of Oxford, Oxford, Scattering, Phys. Rev. E 59, 1899 (1999). 2014), DOI:10.1093/database/bau061. 031015-11
Physical Review X – American Physical Society (APS)
Published: Jul 1, 2017
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera