Contrast Mechanisms of Solution-Dispersed Graphene Compounds in

Article Cite This: Langmuir XXXX, XXX, XXX−XXX

pubs.acs.org/Langmuir

Contrast Mechanisms of Solution-Dispersed Graphene Compounds in Twilight Fluorescence Microscopy Yu-uki Ishikawa, Yu-uto Watanabe, and Masahito Sano* Department of Organic Materials Science, Yamagata University, 4-3-16 Jyonan, Yonezawa, Yamagata 992-8510, Japan

Downloaded via BUFFALO STATE on July 27, 2019 at 18:35:01 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: Twilight fluorescence microscopy is a newly developed technique that is capable of imaging a single-layer graphene compound dispersed in a liquid. A graphene solution mixed with a highly concentrated dye is placed on a glass plate and is irradiated by the excitation beam with an incident angle that has a finite width around the total internal reflection angle. Both the evanescence field and the faint refracted beam decay exponentially as they travel from the glass surface. The dye fluorescence excited by both beams is used as illumination. A simplified theory for dark contrast of graphene compounds is developed based on absorption and Förster resonance energy transfer (FRET), assuming that (1) FRET has a sharp cutoff distance, (2) FRET is independent of the number of layers, and (3) Dexter electron transfer is negligible. The contrast of a reduced graphene oxide multilayer, whose layer heights have been determined by atomic force microscopy, shows good agreement with the simplified theory under various dye concentrations. The FRET cutoff distance is found to be much shorter than one expected for graphene and similar to the distance between two small molecules. This short cutoff distance is the main reason for the assumption to be valid.

■

Scheme 1. Optical Configuration of TwiF Microscopya

INTRODUCTION A typical graphene sheet1 has a van der Waals thickness of 0.35 nm and a macroscopically large area extending over 10 μm2. Graphene oxide (GO)2 and reduced GO (rGO)3 have rough surfaces with an average thickness exceeding 1 nm. For fundamental studies and applications, it is often desirable to disperse these graphene compounds in liquid solvents. In many cases, the graphene dispersion is further manipulated, for instance, to form a thin film, to mix with other substances, and to perform chemical reactions.4 Despite its importance as a precursor, it is difficult to characterize the state of graphene compounds, such as a thickness distribution, an aggregation state, and a degree of chemical modifications, while dispersed in a liquid. In particular, although a graphene area is large enough to be visible under an optical microscope, its atomic thickness and the small absorbance5 of 2.3% pose problems in direct observations of a single-layer graphene compound floating freely in solution. Recently, we have developed twilight fluorescence (TwiF) microscopy which is capable of directly imaging a single-layer graphene compound floating in liquid.6 In a standard procedure, a highly concentrated solution of the fluorescent dye is added to a graphene dispersion. The mixture is placed on a viewing cell with a glass plate at the bottom. An excitation beam, which is configured with a mask and a slit for controlling the incident angle, is incident on the glass/solution interface from the glass side at an angle that has a small width around the total internal reflection (TIR) angle (Scheme 1). A ray with the incident angle larger than the TIR angle produces an evanescent field above the glass surface that decays exponentially within an order of the excitation wavelength. Although its intensity is much weaker than the TIR rays, a faint ray that enters the interface with the angle less than the TIR © XXXX American Chemical Society

a

A sample is mixed with a highly concentrated dye solution and the mixture is placed on a glass plate. The excitation beam incident on the interface with the angle of TIR is reflected completely (thick line). All rays that have the larger incident angles are also reflected and induce the evanescent field with a decay constant βTIR. The rays that enter with the smaller angles are refracted but decay due to the IFE with a decay constant βIFE

angle is refracted into the mixture. Because of strong absorption by the highly concentrated dye, however, the refracted ray cannot propagate a long distance and decays exponentially within an order of few micrometers, known as the inner filter effect (IFE).7 Both the evanescent field and the IFE beam excite only the dye molecules located near the glass surface. The fluorescent light emitted by these dye molecules, Received: May 7, 2019 Revised: July 1, 2019 Published: July 11, 2019 A

DOI: 10.1021/acs.langmuir.9b01349 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 1. TwiF images of GO sheets passing each other in aqueous RhB solution. The number indicates a temporal sequence and the entire process has taken over 30 s.

■

whose intensity is found to be suitable for viewing a single layer graphene sheet in liquid, is used as an illumination source in TwiF microscopy. In traditional fluorescence microscopy, dye molecules are attached to a target object and the stained object is dispersed in a solution that does not contain dye. Under the excitation light, only the stained object fluoresces and appears bright in the dark background. In TwiF microscopy, dye molecules are present everywhere in the solution and the entire viewing area fluoresces. The target object becomes visible if it causes changes in the fluorescence intensity or spectrum of dye molecules interacting with the object. Depending on the interaction, the object appears either brighter or darker than the surrounding solution. In other words, the contrast depends critically on the material properties of the viewing object. In this study, we are interested in developing a simple formula that describes the contrast of graphene compounds in various kinds of dye solutions. The result allows us to evaluate the number of layers in a stacked graphene sheet or the degree of chemical modifications, in addition to morphology and movements that are already possible to observe. A dark contrast mechanism of graphene compounds in TwiF microscopy involves light absorption, Förster resonance energy transfer (FRET), and specific chemical interactions between graphene compounds and dye molecules.6 It is possible to develop a detailed contrast theory based on the currently known properties of FRET and other quenching mechanisms. The detailed theory, however, contains too many parameters that it is difficult to extract the material properties of dispersed graphene compounds. In this study, we develop a simplified theory of the TwiF contrast which involves only two parameters. The theory is tested experimentally by measuring the TwiF contrast and the height of the identical rGO flakes using atomic force microscopy (AFM). The same relation is also examined using different dyes at various concentrations. Because it is difficult to prepare a dispersion of large area (>25 μm2) graphene sheets that are observable by the TwiF microscope, rGO sheets were used extensively.

EXPERIMENTAL SECTION

Preparation of rGO. GO was made by the Hummer’s method,8 followed by a brief sonication for 5 s. GO was reduced by L-ascorbic acid (50 mg/mL) in water for 1 h at 70 °C to afford rGO.9 The resulting rGO was collected on a filter paper and washed thoroughly with hot water. Decay Length in Epi-Fluorescence Microscopy. The absorbance at 560 nm of a given concentration of rhodamine-6G (Rh6G) in N-methyl-2-pyrrolidone (NMP) was measured with a UV−vis absorption spectroscopy system (V-570 spectrophotometer, Jasco) using 0.10, 1.0, 5.0, and 10.0 mm square cuvettes. A slope of the linear region in the absorbance versus cuvette length plot was taken as a decay constant βepi. Decay Length for IFE Beam in TwiF Microscopy. TwiF microscopy was operated with a 60× objective (TIRF lens, Nikon) with an NA of 1.49. For a decay constant of the refracted beam without IFE, GO sheets dispersed in NMP without any dye were observed by TwiF microscopy with the excitation wavelength at 330 < λex < 380 nm and the emission wavelength at 420 nm < λem. The solution was slightly agitated and the movement of a GO sheet that rotated front-to-back was monitored in real time. A tilt angle was determined from the imaged sheet width with respect to that of the horizontally oriented sheet. For a decay constant with IFE, the same procedure was applied to GO sheets in 250 μM Rh6G in NMP with 510 < λex < 560 and 590 nm < λem. Contrast Height Measurements of rGO. rGO dispersed in NMP was cast on a TwiF glass plate and dried completely. It was imaged by the tapping mode AFM (Agilent 5500 SPM) in the air. Then, the same rGO sheet was imaged by using a TwiF microscope in aqueous 250 μM rhodamine-b (RhB) solution with the same optical filter setting as Rh6G. For aqueous 250 μM fluorescein solution, 450 < λex < 490 and 520 nm < λem.

■

RESULTS AND DISCUSSION Simplified Theory of Dark Contrast in TwiF Microscopy. Figure 1 is a series of TwiF images showing a darker GO sheet passing behind the lighter colored GO sheets in solution (see also Video S1 for real-time observation in a different dye solution). GO was chosen because it disperses well in aqueous RhB solution. These features clearly demonstrate that TwiF microscopy is capable of imaging GO sheets floating freely in solution. In both the presented cases, GO appears darker than the surrounding solution. B

DOI: 10.1021/acs.langmuir.9b01349 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir Contrast η is defined as (I0 − I)/I0, where I0 is the light intensity reaching the lens through a pure dye solution and I is the light intensity passed through a graphene sheet. The excitation light consists of the evanescence field and the refracted IFE beam that will be shown to decay exponentially with the characteristic constants βTIR and βIFE, respectively. An n-layer graphene sheet of negligible thickness orients parallel to the glass surface and is located at a distance d from the surface (Figure 2).

where R is an order of 10 nm. Because the dye molecules are uniformly distributed over the entire solution, the dye’s absorption coefficient and quantum yield cancel out in the expression of η. Consequently, the fluorescence intensity is proportional to the local excitation intensity only. The excitation beam reaching the region (III) in Figure 2 is diminished by the graphene absorption. The emitted light in this region is also absorbed by the graphene sheet on the way back to the lens. For d > R, integrating over each region in Figure 2 yields η= Ie (e−βTIR (d − R) βTIR

− T 2n e−βTIR (d + R)) + Ie βTIR

Ir (e−βIFE(d − R) βIFE

− T 2n e−βIFE(d + R))

Ir βIFE

+

(1)

When the graphene sheet is freely floating, d is often much larger than 1/βTIR ≫ R. Then, e−βTIR(d±R) ≈ 0 and eq 1 is simplified to

Figure 2. Geometry of a floating graphene sheet in TwiF microscopy. A graphene sheet locates at a distance d from the surface, which quenches the fluorescence within a distance R around it by FRET. The fluorescence light from the regions (I) and (II) reaches a lens, located below the glass, without any modulation. Both the excitation and fluorescence beams in the region (III) are absorbed by the graphene sheet.

η = η0 e−βIFEd(1 − T 2n)

where η0 =

Ir / βIFE Ie I +βr βTIR IFE

(2)

is a constant that is determined by the

incident angle. Thus, when the graphene compound floats far from the glass surface, the contrast is produced by absorption of the IFE beam and its associated fluorescence intensity at d. For d < R, a similar integration gives

A graphene compound appears dark if I < I0. The dark contrast is caused by absorption and energy (or electron) transfer. Both the excitation and emitted lights are absorbed by the graphene sheet. Because a graphene compound has a relatively flat absorption spectrum over the visible region, each graphene layer is assumed to have a constant transmittance T at all wavelengths except UV. There are two fundamental processes to quench fluorescence: FRET10 and Dexter11 electron transfer. To develop a practical theory, we make the following simplifications (with the reported property in a parenthesis). (1) The fluorescence within a distance R from the graphene surface is completely quenched by FRET and no quenching takes place outside of R. (Unlike FRET between two small molecules, where the FRET efficiency drops sharply at the Förster distance, the efficiency between a molecule and a graphene sheet decreases gradually over the distance.12−14 Because the excitation beam decays exponentially in TwiF microscopy, the FRET term is given by a convolution of both distance dependences in the detailed theory.) (2) Layering has no effect on FRET. (There are reports suggesting that FRET depends on the number of layers in a graphene sheet.15,16 In the detailed theory, the nlayer FRET term involves a sum of FRET from each layer.) (3) A contribution from Dexter electron transfer is much smaller than that by FRET. (Fluorescence quenching of a dye monolayer directly in contact with GO has been reported.17 Also, depending on the kind of dye and solvent, a significant amount of dye can be adsorbed on the graphene surface.18 Dexter electron transfer adds an additional term that depends on the rate constant and the number of adsorbed molecules.) The evanescence field has an intensity of the form Iee−βTIRz and the IFE beam has Ire−βIFEz at a distance z from the glass surface. In standard operation, Ie > Ir and 1/βIFE > 1/βTIR ≫ R,

η=

Ie (1 βTIR

− T 2n e−βTIR (d + R)) + Ie βTIR

+

Ir (1 βIFE

− T 2n e−βIFE(d + R))

Ir βIFE

(3)

η decreases with decreasing d due to exclusion of the nonfluorescent solution. Because 1/βIFE ≫ (d + R), η=

Ie (1 βTIR

− T 2n e−βTIR (d + R)) + Ie βTIR

+

Ir βIFE

Ir (1 βIFE

− T 2n) (4)

Usually, the central incident angle is set to be slightly larger than the TIR angle to ensure the TIR condition. In this case, Ie/βTIR ≫ Ir/βIFE. Because a typical graphene compound has T close to 0.98, eq 4 is further simplified to η = 1 − T 2n e−βTIR (d + R)

(5)

When the graphene compound is located within several nanometers from the glass surface, the contrast is given by absorption and FRET. These expressions hold for any point on the graphene sheet at the distance d from the glass surface. Thus, the same equations can be applied to the contrast of a graphene sheet tilted relative to the glass surface. In the following, we determine each decay constant β first and then discuss the FRET cutoff distance R. Decay Length in Epi-Fluorescence Microscopy. Epifluorescence microscopy is a commonly employed technique in fluorescence microscopy.19 In this configuration, the excitation beam is incident perpendicular to the glass/solution interface and the whole beam is transmitted to the solution phase. Although TwiF microscopy operates on a different principle, it is instructive to find a decay constant βepi of epi-fluorescence microscopy under the condition of high dye concentrations C

DOI: 10.1021/acs.langmuir.9b01349 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir used in TwiF microscopy. The decay constant was measured as a slope of the absorbance against the path length plot using Rh6G in NMP. Figure 3 shows the concentration dependence

Figure 4. (a) Autofluorescence images of a GO sheet that is tilted 33° from the horizontal position (the rotating axis shown by the dashed blue line) and (b) its perpendicular profile (along the yellow line) is fitted by an exponential function. The tilt angle θ is calculated by measuring the apparent width. The excitation wavelength was 330 < λex < 380 nm.

Figure 3. Decay length of epi-fluorescence microscopy for Rh6G in NMP at 560 nm as a function of the dye concentration. A curve is drawn to show the Beer−Lambert law. The concentration typically used in TwiF microscopy is 250−500 μM.

βr, Ire−βrxtanθ. Analyzing several tilt angles yields 1/βr = 2.27 ± 0.53 μm. For the measurement of the decay constant βIFE of the refracted beam with IFE, the same procedure as above was taken with GO sheets dispersed in 250 μM Rh6G solution of NMP. In this particular system, GO appears brighter than the surrounding solution. The bright contrast is explained by the formation of fluorescent J-aggregates of Rh6G on the GO surface while the fluorescence in solution is suppressed by nonfluorescent H-aggregates.6 Similar to the autofluorescence case, local fluorescence intensity is proportional to the local excitation intensity at the same location. A video showing the rotational motion was presented in the previous paper.6 Figure 5 exhibits a video clip and the fluorescence profile. An

of the decay length at 560 nm. For the concentration below 250 μM, 1/βepi is inversely proportional to the concentration, indicating the Beer−Lambert law. Above this concentration, 1/ βepi deviates due to IFE. For the dye concentration typically used in TwiF microscopy, 1/βepi ≈ 1.5 mm. Decay Length for Evanescence Field in TwiF Microscopy. A decay constant βTIR for the evanescent field is given by20 βTIR =

2π λex

sin 2 θi ns 2

−1 (6)

where λex denotes the excitation wavelength, ns is the ratio of refractive index of the dye solution to that of the glass, and θi is the angle of incidence. In TwiF microscopy, the incident angle has a small width around the TIR angle. For a case of the glass/water interface with λex = 500 nm, 1/βTIR ≈ 0.2−0.3 μm. Decay Length for Refracted IFE Beam in TwiF Microscopy. Because of the long travel distance between the slit and the objective lens in the microscope, the crosssectional intensity profile of the incident beam is maximum at the central angle and diminishes rapidly as the angle deviates from it (Scheme 1). When the central angle is set to the TIR angle, the rays with the angles smaller than the TIR angle continue to propagate into the solution phase and travel with the small grazing angles to the glass surface. In order to evaluate an intensity profile of the refracted rays, we utilize autofluorescence of GO. With no dye in solution, GO itself fluoresces under UV light.21 Because autofluorescence originates from oxide species, rGO has much weaker autofluorescence than GO. A local autofluorescence intensity from a point on a GO sheet is proportional to the excitation intensity reaching to that point. Because no dye is involved in this case, the refracted beam propagates without IFE. The GO solution was agitated slightly and a rotating GO sheet in the front-to-back direction (the rotation axis parallel to the glass surface) was followed by the TwiF microscope. The tilt angle θ was calculated from measuring the apparent width x of the tilted sheet along the glass surface with respect to that in the horizontal orientation. As Figure 4 shows, its autofluorescence profile across a line perpendicular to the rotating axis can be fitted well with an exponential function with a decay constant

Figure 5. Rh6G fluorescence profile [along the blue line in (a)] of the GO sheet that is tilted 18° from the horizontal position is fitted by an exponential function (b). In (b), the intensity is plotted upside down to make a comparison with Figure 4 easier.

exponential fit, Ire−βIFExtanθ, over several angles gives 1/βIFE = 1.14 ± 0.28 μm. Thus, the effect of IFE is to limit the propagation distance of the refracted beam to half. The value of 1/βIFE is an order of magnitude larger than 1/ βTIR, but 3 orders of magnitude smaller than 1/βepi. In ordinary TIR fluorescence microscopy, the excitation beam incidents on the interface with the TIR angle and an object within the distance 1/βTIR from the interface is imaged. In practice, this situation corresponds to an object adhered on the glass surface. It is not possible to see floating and moving graphene sheets by TIR fluorescence microscopy. In TwiF microscopy, it is our common experience that the region around several times 1/ βIFE fluoresces with enough intensity to image a floating object D

DOI: 10.1021/acs.langmuir.9b01349 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 6. (a) AFM and (b) TwiF image of the same rGO sheets on the glass. AFM was measured in air and TwiF microscopy was performed in aqueous RhB solution. A top part of the right fragment of the rGO sheets in (b) was lost during handling of the sample.

away from the interface. This confined fluorescing region is also a key factor to image a single-layer graphene compound. In epi-fluorescence microscopy, all dye molecules within 1/βepi fluoresce. In this case, the fluorescence intensity is too strong for a graphene compound with a transmittance over 0.98 to produce a measurable contrast without causing the detector to saturate. Graphene Compounds at the Glass Surface. At d = 0, the n-layer TwiF contrast becomes η = 1 − T 2n e−βTIR R

height contains a non-negligible deviation from the layer thickness. On the other hand, the data points that are connected by a line indicate different layers belonging to the same rGO sheet. These points were observed under the same experimental condition in TwiF microscopy, implying a small relative error in determining the TwiF contrast. These considerations indicate that each line connecting the points can be shifted horizontally by the amount of uncertainty. Within these experimental errors, we conclude that eq 7 describes the TwiF contrast well. The curve is very sensitive to the value of T. The average transmittance at 600 nm is given by T = 0.992. Because ideal graphene has T = 0.977, rGO is more transparent as we expect. In order to test the assumptions concerning FRET, the multilayer rGO sheet of the known AFM heights was observed by TwiF microscopy under different dye concentrations. The TwiF contrast and the AFM height was fitted with eq 7 to obtain the transmittance and the FRET distance. The identical experiments were repeated using two different kinds of dye. In Figure 8, the points connected by a line refer to the same rGO sheet. Clearly, the transmittance of each rGO sheet is independent of the dye concentration for both RhB (λem ≈ 600 nm) and fluorescein (λem ≈ 530 nm). Because the contrast is produced by the evanescence field at d = 0, IFE does not play any role. Considering the violent oxidation process in Hummer’s method, there is a possibility that the variation among rGO sheets is caused by the different degree of oxidation/reduction. In Figure 8b, only the dye concentration was changed while the optical setting remained unchanged for a given rGO sheet, meaning that the data points connected by a line have the constant βTIR. The variation among rGO sheets is likely due to slightly different βTIR values set in each measurement as well as the degree of oxidation/reduction. Because Dexter electron transfer is caused by dye molecules that are adsorbed on the rGO surface, its effect should appear in the concentration dependence. Although there is a possibility of systematic changes with respect to the concentration, its dependence is small. Also, there is no dependence on the kind of dye. Thus, the effect of Dexter electron transfer is small in these systems. In order for Dexter electron transfer to be effective, molecular orbitals of the dye molecule and the graphene need to overlap. This makes the orientation of the dye molecule crucial and a small contribution of Dexter electron transfer between a flat molecule and the graphene plane has been suggested theoretically.13 In the original FRET theory between a donor and an acceptor molecule with a distance r apart, the Förster distance

(7)

There are only two material parameters: T and R. To verify this relation, an identical rGO sheet placed on the glass was imaged by both TwiF microscopy in aqueous RhB solution and AFM in the air (Figure 6). Here, rGO was used to avoid detaching from the glass surface as it was not dispersible in the aqueous RhB solution. We selected the multilayer rGO sheet so that nearly every layer was exposed partially. This allowed evaluating the TwiF contrast as well as the AFM height of each layer within the same rGO sheet. Figure 7 summarizes the results of all rGO sheets. The best fit was obtained by η = 1 − (0.985)h(0.99), where h is the

Figure 7. TwiF contrast vs AFM height. A best-fit curve is shown as a thick gray curve. The points connected by a colored line indicate the same rGO sheet with different layer surfaces.

AFM height. The thickness of a single layer graphene is 0.35 nm. When a single layer rGO sheet is observed by AFM, it typically shows an average AFM height of about 1 nm. The increased AFM height is due to the surface roughness caused by oxide groups and strains on the graphene plane as a result of inhomogeneous oxidation. In a multilayer rGO sheet, each layer does not need to be parallel and the spacing is not correlated with the layer number. Thus, the measured AFM E

DOI: 10.1021/acs.langmuir.9b01349 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 8. Dye concentration dependences on (a) transmittance T and (b) FRET distance βTIRR on the rGO sheets. The points connected by a line refer to the same rGO sheets. The reddish points connected by a solid line were taken for RhB and the greenish points connected by a dashed line for fluorescein.

R appears as a reference to r−6 dependence of the FRET efficiency, indicating that the efficiency drops sharply within a short distance around R. In the present system of a donor molecule and a graphene sheet, the distance dependence is given by r−4. This dependence implies that the efficiency drops gradually over a longer distance.22 Nevertheless, we approximated the change by the sharp cutoff distance R and obtained a reasonable result. If we take a value of 1/βTIR to be 250 nm, R ≈ 5 nm. This magnitude is found commonly in FRET between small molecules23 and is shorter than a quenching range predicted for graphene of 20−30 nm.12 This short distance also implies that only the outermost layer facing the dye solution contributes significantly to FRET, making the layer dependence of FRET much smaller than that of absorption. There is a report showing that GO quenches the fluorescence of the polymer dispersed dye that is 20 nm away.24 It took a separation distance of 200 nm to avoid the quenching. Such a long-distance quenching effect with GO cannot be explained by FRET, and other contrast mechanisms may be involved in their case. Graphene Compounds Floating in Solution. At d ≫ 1/ βTIR, the TwiF contrast is given by eq 2. The exponential term comes from the local excitation intensity and the term in the parenthesis signifies dark contrast by absorption. Although we could not perform experiments to examine this relation presently due to difficulty in controlling d, the exponential term is common to any contrast mechanisms and has been partially verified in the measurement of βIFE in Figure 5. For a stationary graphene sheet of a given layer number, eq 2 indicates that the TwiF contrast depends only on the transmittance. Because the transmittance is closely related to the electronic structure of a graphene compound, chemical modifications should induce a change in TwiF contrast. In fact, a reduction reaction of individual GO sheet can be followed in the as-dispersed state.25 Conditions for Valid Simplifications. The three simplifications made to develop the present theory are shown to be valid for rGO. The effective FRET distance was found to be short. In rGO, many small areas consisting of undamaged π-electron conjugation are scattered over the sheet surface and are surrounded by poorly conjugated oxidized fragments.26 The present result suggests that FRET between a dye molecule and rGO behaves as if it is between the dye molecule and a small aromatic molecule, that is, each scattered conjugated region in rGO acts as a molecule. This picture

indicates that the electronic structure of rGO is responsible for the sharp FRET cutoff and the layer independence. Although Dexter electron transfer of graphene compounds has not been studied in the solution phase, it appears to be not significant for flat dye molecules. These considerations suggest that the present simplifications are also valid for GO. On the other hand, graphene is predicted to induce FRET on a dye molecule located 20−30 nm away. There is a possibility that the variation of the evanescence intensity over such a distance may not be negligible, which requires to take a convolution as in the detailed theory. Because this FRET range is much larger than the single-layer graphene thickness, the inner layers are expected to contribute to FRET. Thus, we do not expect that the present simplifications hold for graphene. Presently, it is not possible to examine this prediction due to difficulty in preparing a dispersion of large area (>25 μm2) graphene sheets without using some kinds of dispersing agents such as surfactants or polymers. The agent molecules are adsorbed on the graphene surface and cause unpredictable changes in the dye concentration. In reality, nearly all dispersions of graphene compounds are made of either GO or rGO. Thus, the simplified theory is still useful in practice.

■

CONCLUSIONS A simplified theory of dark contrast on a graphene compound imaged by TwiF microscopy is developed and verified experimentally. Under certain conditions, TwiF contrast is given by only two parameters: absorbance and a FRET cutoff distance. The simplified theory allows us to evaluate not only morphology and movements of each graphene compound floating in solution but also the number of layers and the extent of electronic modifications. Because a small modification of dye’s fluorescence gives rise to measurable TwiF contrast, various materials other than graphene compounds become candidates for imaging in the as-dispersed state. In particular, a floating transparent object that is difficult to observe by other microscopies may be imaged without staining the object.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.9b01349. F

DOI: 10.1021/acs.langmuir.9b01349 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

■

(16) Chen, Z.; Berciaud, S.; Nuckolls, C.; Heinz, T. F.; Brus, L. E. Energy Transfer from Individual Semiconductor Nanocrystals to Graphene. ACS Nano 2010, 4, 2964−2968. (17) Treossi, E.; Melucci, M.; Liscio, A.; Gazzano, M.; Samorì, P.; Palermo, V. High-Contrast Visualization of Graphene Oxide on DyeSensitized Glass, Quartz, and Silicon by Fluorescence Quenching. J. Am. Chem. Soc. 2009, 131, 15576−15577. (18) Ramesha, G. K.; Vijaya Kumara, A.; Muralidhara, H. B.; Sampath, S. Graphene and Graphene Oxide as Effective Adsorbents toward Anionic and Cationic Dyes. J. Colloid Interface Sci. 2011, 361, 270−277. (19) Webb, D. J.; Brown, C. M. Epi-fluorescence microscopy. Methods Mol. Biol. 2012, 931, 29−59. (20) Hecht, E.; Zajac, A. Optics; Addison-Wesley Pub. Com.: Reading, Massachusetts, 1979. (21) Kozák, O.; Sudolská, M.; Pramanik, G.; Cígler, P.; Otyepka, M.; Zbořil, R. Photoluminescent Carbon Nanostructures. Chem. Mater. 2016, 28, 4085−4128. (22) Yun, C. S.; Javier, A.; Jennings, T.; Fisher, M.; Hira, S.; Peterson, S.; Hopkins, B.; Reich, N. O.; Strouse, G. F. Nanometal Surface Energy Transfer in Optical Rulers, Breaking the FRET Barrier. J. Am. Chem. Soc. 2005, 127, 3115−3119. (23) Ray, P. C.; Fan, Z.; Crouch, R. A.; Sinha, S. S.; Pramanik, A. Nanoscopic Optical Rulers Beyond the FRET Distance Limit: Fundamentals and Applications. Chem. Soc. Rev. 2014, 43, 6370− 6404. (24) Kim, J.; Cote, L. J.; Kim, F.; Huang, J. Visualizing Graphene Based Sheets by Fluorescence Quenching Microscopy. J. Am. Chem. Soc. 2010, 132, 260−267. (25) Kanazawa, K.; Sato, H.; Sano, M. Reaction Kinetics of Reducing Graphene Oxide at Individual Sheet Level Studied by Twilight Fluorescence Microscopy. J. Phys. Chem. C 2019, 123, 6881−6887. (26) Eda, G.; Lin, Y.-Y.; Mattevi, C.; Yamaguchi, H.; Chen, H.-A.; Chen, I.-S.; Chen, C.-W.; Chhowalla, M. Blue Photoluminescence from Chemically Derived Graphene Oxide. Adv. Mater. 2010, 22, 505−509.

Real-time observation of drifting GO sheets in NMP solution of coumarin (MPG)

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Masahito Sano: 0000-0002-9050-5269 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS A part of this work is based on the results obtained from a project commissioned by the New Energy and Industrial Technology Development Organization (NEDO).

■

REFERENCES

(1) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Electric Field Effect in Atomically Thin Carbon Films. Science 2004, 306, 666−669. (2) Dreyer, D. R.; Park, S.; Bielawski, C. W.; Ruoff, R. S. The Chemistry of Graphene Oxide. Chem. Soc. Rev. 2010, 39, 228−240. (3) Chua, C. K.; Pumera, M. Chemical Reduction of Graphene Oxide: A Synthetic Chemistry Viewpoint. Chem. Soc. Rev. 2014, 43, 291−312. (4) Tan, C.; Cao, X.; Wu, X.-J.; He, Q.; Yang, J.; Zhang, X.; Chen, J.; Zhao, W.; Han, S.; Nam, G.-H.; Sindoro, M.; Zhang, H. Recent Advances in Ultrathin Two-Dimensional Nanomaterials. Chem. Rev. 2017, 117, 6225−6331. (5) Nair, R. R.; Blake, P.; Grigorenko, A. N.; Novoselov, K. S.; Booth, T. J.; Stauber, T.; Peres, N. M. R.; Geim, A. K. Fine Structure Constant Defines Visual Transparency of Graphene. Science 2008, 320, 1308. (6) Matsuno, Y.; Sato, Y.-u.; Sato, H.; Sano, M. Direct Observations of Graphene Dispersed in Solution by Twilight Fluorescence Microscopy. J. Phys. Chem. Lett. 2017, 8, 2425−2431. (7) Fonin, A. V.; Sulatskaya, A. I.; Kuznetsova, I. M.; Turoverov, K. K. Fluorescence of Dyes in Solutions with High Absorbance. Inner Filter Effect Correction. PLoS One 2014, 9, No. e103878. (8) Hummers, W. S., Jr.; Offeman, R. E. Preparation of Graphitic Oxide. J. Am. Chem. Soc. 1958, 80, 1339. (9) Fernández-Merino, M. J.; Guardia, L.; Paredes, J. I.; Villar-Rodil, S.; Solís-Fernández, P.; Martínez-Alonso, A.; Tascón, J. M. D. Vitamin C Is an Ideal Substitute for Hydrazine in the Reduction of Graphene Oxide Suspensions. J. Phys. Chem. C 2010, 114, 6426−6432. (10) Fö rster, T. Zwischenmolekulare Energiewanderung und Fluoreszenz. Ann. Phys. 1948, 2, 55-75 DOI: 10.1002/ andp.19484370105 English translation by Knox, R. S. Biological Physics; Mielczarek, E. V., Greenbaum, E., Knox, R. S., Eds.; American Institute of Physics: New York, 1993; pp 148−160. (11) Dexter, D. L. A Theory of Sensitized Luminescence in Solids. J. Chem. Phys. 1953, 21, 836. (12) Swathi, R. S.; Sebastian, K. L. Long range resonance energy transfer from a dye molecule to graphene has (distance)−4 dependence. J. Chem. Phys. 2009, 130, 086101. (13) Malic, E.; Appel, H.; Hofmann, O. T.; Rubio, A. FörsterInduced Energy Transfer in Functionalized Graphene. J. Phys. Chem. C 2014, 118, 9283−9289. (14) Federspiel, F.; Froehlicher, G.; Nasilowski, M.; Pedetti, S.; Mahmood, A.; Doudin, B.; Park, S.; Lee, J.-O.; Halley, D.; Dubertret, B.; Gilliot, P.; Berciaud, S. Distance Dependence of the Energy Transfer Rate from a Single Semiconductor Nanostructure to Graphene. Nano Lett. 2015, 15, 1252−1258. (15) Xie, L.; Ling, X.; Fang, Y.; Zhang, J.; Liu, Z. Graphene as a Substrate to Suppress Fluorescence in Resonance Raman Spectroscopy. J. Am. Chem. Soc. 2009, 131, 9890−9891. G

DOI: 10.1021/acs.langmuir.9b01349 Langmuir XXXX, XXX, XXX−XXX