Quantum Coherent Modulation-Enhanced Single-Molecule ImagingMicroscopyHaitao Zhou,†,‡ Chengbing Qin,*,†,‡ Ruiyun Chen,†,‡ Yaoming Liu,§ Wenjin Zhou,†,‡

Guofeng Zhang,†,‡ Yan Gao,†,‡ Liantuan Xiao,*,†,‡ and Suotang Jia†,‡

†State Key Laboratory of Quantum Optics and Quantum Optics Devices, Institute of Laser Spectroscopy, Shanxi University,Taiyuan, Shanxi 030006, China‡Collaborative Innovation Center of Extreme Optics, Shanxi University, Taiyuan, Shanxi 030006, China§Scientific Instrument Center, Shanxi University, Taiyuan, Shanxi 030006, China

*S Supporting Information

ABSTRACT: In fluorescence imaging and detection, undesired fluorescence interference(such as autofluorescence) often hampers the contrast of the image and even prevents theidentification of structures of interest. Here, we develop a quantum coherent modulation-enhanced (QCME) single-molecule imaging microscopy (SMIM) to substantiallyeliminate the strong fluorescence interference, based on manipulation of the excited-state population probability of a single molecule. By periodically modulating the phasedifference between the ultrashort pulse pairs and performing a discrete Fourier transformof the arrival time of emitted photons, the decimation of single molecules from stronginterference in QCME-SMIM has been clearly determined, where the signal-to-interference ratio is enhanced by more than 2 orders of magnitude. This technique,confirmed to be universal to organic dyes and linked with biomacromolecules, paves theway to high-contrast bioimaging under unfavorable conditions.

Single-molecule fluorescence imaging and detection is apowerful method to understand the behavior of single

biomolecules in their nanoenvironments as well as their rolesin cellular processes.1−3 However, the sensitivity of conven-tional fluorescence imaging is still limited by the lowfluorescence intensity of probes and the presence of a strongbackground or interference, which often hampers imagingcontrast and even prevents the identification of interestedstructures. It is an enormous challenge to separate fluorescencesignals from interference (such as autofluorescence inbiological tissues4,5) in fluorescence-based biological detectionand biomedical diagnosis. To address such demands, hyper-spectral imaging5,6 and fluorescence lifetime imaging7,8 havebeen developed to reduce the influence of interference.Hyperspectral imaging presents an excellent effect when theemission spectra of interference are sufficiently different fromthose of interest, and so is lifetime imaging. However, neithermethod can distinguish fluorescence spectra and lifetimes thatare very similar, which limits their further applications.8−10

Recently, single-molecule imaging microscopy (SMIM)based on fluorescence modulation has been exploited tocircumvent this limitation and to enhance the sensitivity andselectivity of fluorescence imaging. Among fluorescencemodulations, photoswitchable fluorophores11−13 or fluorescentproteins (PS-FPs)14 are frequently used. The fluorescence ofPS-FPs can be optically switched either from one color toanother15 or from on to off11,16 by a secondary excitation,while the interference cannot be simultaneously modulated.

Thus, PS−PFs can be discriminated from the stronginterference even with identical spectra or lifetimes.15−27 Upto now, fluorescence modulation techniques, such as dual-lasermodulated synchronously amplified fluorescence image recov-ery (DM-SAFIRe)28−32 and that combined with out-of-phaseimaging after optical modulation (OPIOM),33 have been well-developed and have been used to reveal structures hiddenbelow the ambient interference by enhancing the imagingcontrast.34−37 However, these methods are merely suitable forthe fluorophores with photosensitive reactants and products,such as methoxyspiropyran (MSP) and methoxymerocyanine(MMC) or other PS-FPs.11,13,17 A universal technique thatapplies to all dye molecules is still lacking.Here, we develop a new fluorescence modulation technique

based on quantum coherent manipulation of single moleculesby modulating the phase difference between an ultrashort pulsepair, which is used to pump and probe single molecules.38−41

The periodic phase modulation substantially varies thepopulation probability of the excited state of single moleculesand thus changes their fluorescence emission. In contrast, theunwanted interference cannot be modulated. After demodulat-ing fluorescence photons, the contrast of fluorescence imagingin the frequency domain can be dramatically enhanced. Thesignal-to-interference ratio is improved from inundated (below

Received: November 30, 2018Accepted: January 1, 2019Published: January 1, 2019

Letter

pubs.acs.org/JPCLCite This: J. Phys. Chem. Lett. 2019, 10, 223−228

© XXXX American Chemical Society 223 DOI: 10.1021/acs.jpclett.8b03606J. Phys. Chem. Lett. 2019, 10, 223−228

Dow

nloa

ded

via

SHA

NX

I U

NIV

on

Janu

ary

8, 2

019

at 0

6:33

:06

(UT

C).

Se

e ht

tps:

//pub

s.ac

s.or

g/sh

arin

ggui

delin

es f

or o

ptio

ns o

n ho

w to

legi

timat

ely

shar

e pu

blis

hed

artic

les.

the background) at ∼0.054 in the time domain to a satisfactoryratio of 27.79 in the frequency domain, with a 515-foldcontrast enhancement. More importantly, the enhancementrepresents a power-law exponential increase over integrationtime, hinting that an unlimited contrast enhancement can beachieved in theory. Three types of organic dyes and that linkedwith a biomolecule protein have been tested successfully,indicating that our quantum coherent modulation-enhanced(QCME) SMIM is generally applicable to all dye molecules.Furthermore, the maintained high-contrast enhancement inthe region of 1 Hz to 500 kHz allows this technique to beapplied for both wide-field and confocal imaging with arbitrarymodulation speed. These outstanding features suggest that ourtechnique has great potential applications in bioimaging withstrong interference, as well as probing fast dynamics ofbiological processes.The principle of fluorescence modulation of single dye

molecules by ultrashort pulse pairs can be understood asmanipulation of the excited-state population probability.Considering the two-level approximation, the interactionbetween a single molecule and an ultrashort pulse pair,which is on resonance between its ground state (|0⟩) andexcited state (|1⟩), will prepare a coherent state, as shown inFigure 1a. This state can be illustrated by the Bloch vector on

the Bloch sphere, where the poles correspond to theeigenstates (“south”, |0⟩, “north”, |1⟩) and any other pointsindicate coherent superposition states between ground- andexcited-state electronic wave functions.39,42,43 The key toprepare a coherent state is to control the phase difference, Δφ,between the pulse pair. When the single molecule interactswith the first pulse, the process of preparing the coherent statecorresponds to rotation of the Bloch vector away from itsground state, as shown by the blue arrows in Figure 1b,c.Under ideal conditions where the single molecule has nodephasing, the final coherent state is determined by Δφ. WhenΔφ is fixed to 0 rad, the second pulse will rotate the Blochvector to the excited state, indicating the maximum excited-state population probability, as shown in Figure 1b. Thesecond pulse will rotate the Bloch vector to the ground statewhen Δφ is fixed to ±π rad, indicating the minimum excited-state population probability, as shown in Figure 1c. Theexcited-state population probability ρee with arbitrary Δφ canbe expressed as follows

ρ θ θ θ θ φ= · − + Δ12

(1 cos cos sin sin cos )ee 1 2 1 2 (1)

where θ1 and θ2 are the dimensionless pulse areas.42 For acertain single molecule and fixed power density, they areconstants (for detailed derivation, see the SupportingInformation, section S1). According to eq 1, the populationprobability of the excited state can be manipulated effectivelyby controlling Δφ. Considering that the fluorescence is directlycorrelated with its population probability, the fluorescenceintensity of the single molecule can be modulated by the phasedifference. However, for the interference (autofluorescence intissues) or background (ambient light or laser leaking), theintensities cannot be modulated effectively due to the lack ofcoherent interaction with the ultrashort pulse pairs.A schematic diagram of QCME-SMIM is presented in

Figure 2a, and a detailed description is supplied in the

Experimental Methods section. Briefly, a subpicosecond pulsepair with a tunable phase difference was generated by aMichelson interferometer structure, where an electro-opticalmodulator (EOM) was on one arm. On the other arm, thelaser passed through a quarter-wave plate. Thus, tworecombined beams had orthogonal polarizations, which couldeliminate the intensity modulation due to their owninterference (to eliminate the polarization modulation,modulation on unpolarized colloidal quantum dots wasperformed, which presented clear modulation; see Figure S1

Figure 1. Principle of the phase-controlled quantum coherent state.(a) The ultrashort pulse pairs with the tunable phase realizes controlof the coherent state between the ground state (|0⟩) and excited state(|1⟩) with resonance excitations. (b) Maximum excited-statepopulation probability is achieved with Δφ = 0. (c) The minimumexcited-state population probability is achieved with Δφ = ±π.

Figure 2. (a) Experimental setup of QCME-SMIM. QWP, quarter-wave plate; BS, beam splitter; EOM, electro-optical modulator; DM,dichroic mirror; SPAD, single-photon counting modulator; MCPET,multichannel picosecond event timer. (b) Periodic sawtooth voltageapplied on the EOM to change the phase difference of the pulse pairfrom −π to π, (c) which produces the population probability of theexcited state in sine form (waveform in gray). The blue lines indicatethe possible photon emission in each cycle. (d) Statistics offluorescence photons from a single molecule (SM., blue) andbackground (Back., gray) as a function of phase difference Δφ. (e)DFT of the arrival time of the photon series from a single moleculeand background.

The Journal of Physical Chemistry Letters Letter

DOI: 10.1021/acs.jpclett.8b03606J. Phys. Chem. Lett. 2019, 10, 223−228

224

for details). In the experiment, the phase difference wasperiodically modulated between −π and π by applying aperiodic sawtooth voltage on the EOM, as shown in Figure 2b.The modulation frequency was adopted to 1 kHz for mostcases. According to eq 1, the population probability of theexcited state will be modulated with a sinusoidal waveform as afunction of the phase difference, as shown in Figure 2c. Thus,in each cycle, the probability of fluorescence from a singlemolecule will also emit as a sinusoidal wave. However, due tothe photophysical properties of single molecules (such asphotoblinking) and the limited collection efficiency of theoptical system, only several photons can be detected in eachcycle (in 1 ms with a modulation frequency of 1 kHz). Byrecording the absolute arrival time of each photon and carryingout the statistics of all photons as a function of phase differencein an integration time, fluorescence photons in the modulationperiod can be achieved, as shown in Figure 2d. Note thatphotons emitted from a single molecule present a sinusoidalwaveform, as expected, while those from the background orinterference are almost constant in the full period. Demodulat-ing the arrival time of the photon series by discrete Fouriertransform (DFT), the single molecule shows a significant DFTsignal at 1 kHz, which is orders of magnitude larger than thatof the background, as shown in Figure 2e. Thus, we can expectthat reconstruction of single molecule imaging by DFT at themodulation frequency can dramatically suppress the back-ground/interference and significantly enhance the imagecontrast.To illustrate the contrast enhancement in the frequency

domain based on QCME-SMIM, squaraine-derived rotaxane(SR) molecules mixed with microspheres (based on melamineresin, marked with Nile Blue) were spin-coated on thesubstrate as a testing sample. Here, SR molecules with weakfluorescence emission (∼104 counts/s) are used as the signalsource, while microspheres with strong emission (∼106counts/s) are used as the interference. Due to the stronginteraction between closely packed dye molecules, the rapiddecoherence of microspheres results in no response toquantum coherent modulation.39,44 SRs and microsphereshave overlapping spectra, as shown in Figure S2. Thus, theycannot be distinguished by traditional fluorescence imaging.Figure 3 presents fluorescence imaging in the time domain(fluorescence intensity for each pixel) and frequency domain(DFT magnitude for each pixel) with different integrationtimes. As shown in Figure 3a, three bright patterns originatefrom the emission of microspheres, while SRs have been fullysubmerged in bright microspheres. Note that with increasingthe integration time no contrast improvement can be obtainedin the time domain. On the contrary, in the frequency domainfrom the same data, the strong interference has been wellsuppressed, and SRs (highlighted by three dashed circles) canbe clearly discerned, as shown in Figure 3b. More importantly,higher contrast can be achieved by increasing the integrationtime. This contrast enhancement can be explained as thenonmodulation of microspheres under periodic modulation ofthe phase difference.To further explore the contrast enhancement in the

frequency domain, we plot both fluorescence intensities andDFT magnitudes of the selected lines (labeled in Figure 3a,b),varied as the integration time, as shown in Figure 3c,d,respectively. From the frequency domain, we can find thatthere are two single molecules in the selected line, marked asS1 and S2, where the general background and interference

(strong emission from the microsphere) are marked as B and I,respectively. Owing to the fluorescence intensities of all fourselected areas (labeled as color bars) in the time domainsynchronously increasing with the integration time, no contrastimprovement can be determined. However, DFT magnitudesof single molecules increase much faster than those of theinterference/background, resulting in dramatic enhancement.This enhancement can be quantified by assuming averagephotons emitted per unit of signal and background/interference. The signal-to-background (S/B) ratio in thetime domain can be represented as

= · −R k k(Fluo.)S/B S B1

(2)

where kS represents photons emitted per second from a singlemolecule and kB represents photons emitted per second fromthe background. The S/B ratio in the frequency domain can bedescribed as45−48

ξ= · · ·−R k k t(DFT)S/B S B1/2 1/2

(3)

where t is the integration time (in s) and ξ is the slope factor ofmodulation, depending on the interaction strength betweenthe single molecule and the pulse pair (for detailed derivation,see the Supporting Information, section 4). Although the S/Bratio does not change in the time domain once the kS and kBare determined, as given in eq 2, the RS/B(DFT) can be

Figure 3. QCME-SMIM eliminates the strong interference. (a,b)Time domain and frequency domain images of SR molecules mixedwith microspheres recorded with various integration times. The threebright patterns in the time domain (a) are the strong fluorescenceemission from three microspheres, and the three emerging smallercircular patterns (highlighted by dashed circles) in the frequencydomain (b) are the DFT signal from three single SR molecules. Forclarity and comparison, all images are normalized by their maximumvalues. Scale bar: 1 μm. (c,d) Fluorescence intensities and DFTmagnitudes of labeled lines in (a) and (b) under four integrationtimes, which are 0.05, 0.2, 0.8, and 2.0 s, respectively. Three letters (α,β, and γ) denote the beginning, the middle, and the end of the lines.S1 and S2 denote the signals from the two single molecules,respectively. B and I (highlighted by solid circles) denote theintensities from the general background and strong interference,respectively.

The Journal of Physical Chemistry Letters Letter

DOI: 10.1021/acs.jpclett.8b03606J. Phys. Chem. Lett. 2019, 10, 223−228

225

unlimitedly improved in theory by increasing the integrationtime, as given in eq 3.To confirm the formulas, we plot the S/B ratio of time and

frequency domains as a function of integration time, as shownin Figure 4a. The fluorescence intensities and DFT magnitudes

of the signal and background are extracted from marked areasin Figure 3c,d, S1 and B, respectively. Note that RS/B(Fluo.)values are in the region of 8.8−9.8, consistent with eq 2. Thefluctuation mainly originates from the variation in the photonemission of single molecules (such as photoblinking). On theother hand, RS/B(DFT) is smaller at short integration time,while it is increased as a power-law function of integrationtime. The data can be fitted by the power-law model, R = a ×tb, yielding an exponent of b equal to 0.536, very close to 0.5(eq 3), which is in good agreement with the theoreticalpredictions.48 In this case, RS/B(DFT) equals RS/B(Fluo.) at anintegration time of 0.32 s and reaches 24.13 at 2.0 s. In orderto clearly reveal the contrast enhancement, we also present theratio of RS/B(DFT) to RS/B(Fluo.). The enhancement is alsoincreased in power-law with an exponent of 0.503, indicatingthat the enhancement can be improved unlimitedly in theoryby continuously increasing the integration time.This unlimited contrast enhancement is not only appropriate

for the S/B ratio but also suitable for the ratio of signal-to-interference (S/I). To illustrate the applicability, thefluorescence intensities and DFT magnitudes of singlemolecule S2 and interference of I (Figure 3c,d) are selected.Note that the fluorescence intensity of S2 is fully inundated bymicrospheres, and no signal can be found in Figure 3c. Byassuming that the fluorescence intensity of S2 is equal to that ofS1, the S/I ratios of time and frequency domains as a functionof integration time are shown in Figure 4b. The QCME-SMIMdramatically improves the S/I ratio from undetectable levels of0.054 to a favorable ratio of 3.1 even at an integration time ofonly 0.02 s. The fitting results proved that both the RS/I(DFT)and enhancement (the ratio of RS/I(DFT) to RS/I(Fluo.)) areincreased in power-law with exponents close to 0.5, coincidingwith eq 3. Due to the strong emission and nonmodulationfeature of microspheres, the enhancement reached 515 at anintegration time of 2.0 s. We also demonstrated the robustnessof this technique for the case of a laser leaking/ambient light-dominant background that drowns out the fluorescence ofsingle molecules of interest. More than 2 orders of magnitude

enhancement of the S/B ratio has been determined (fordetailsm see Figures S4 and S5).Specifically, QCME-SMIM is a universal method, applying

to all dyes with two-level approximation in theory, where thepopulation probability of the excited state can be manipulatedby the ultrashort pulse pairs. The huge contrast enhancementin the frequency domain has also been determined from twoother dye molecules (Cy5 and DiD) and that bounded withbiological proteins (streptavidin-bonded Cy5) (see Figure S6).Furthermore, by labeling SR molecules, the Escherichia coli canbe clearly distinguished from the large background. Figure 5

shows the comparison of the traditional wide-field fluorescenceintensity imaging of a coli bacillus cell and the correspondingQCME-SMIM. The fluorescence signal of the coli bacillus cellis drowned in the large background, while it becomes clearlyvisible in modulated imaging, as shown in Figure 5b. Theseattempts illustrate not only the universal application of thistechnique but also potential applications for biological imaging.The broad dynamic range of the modulation frequency and

potentially high imaging speed are other critical features of thistechnique. We have performed the modulation in thefrequency region of 1 Hz to 500 kHz, as shown in FigureS7. Apparently, the DFT magnitudes of both the backgroundand SR have no substantial change in almost 6 orders ofmagnitude. This broad region allows QCME-SMIM to beapplied to both wide-field imaging that can collect informationon the numbers of individual molecules simultaneously andconfocal fluorescence imaging that allows investigation of thereal-time dynamics of a single molecule for a long period withhigh temporal resolution.In conclusion, we have developed a novel fluorescence

modulation imaging technique with dramatic imaging contrastenhancement in the frequency domain by varying the phasedifference between the ultrashort pulse pairs that is used forexcitation, which can effectively manipulate the excited-statepopulation probability of single molecules and therefore theirfluorescence intensity. Imaging the single-molecule signal byQCME-SMIM can efficiently suppress the background orinterference, which will inevitably overwhelm the signal intraditional fluorescence microscopy. The S/I ratio hasdramatically improved from below the background (0.054)to a satisfactory ratio of 27.79, with contrast enhancement of515. More critically, this technique not only acts as a universaltechnique to the organic dyes and that linked withbiomacromolecule but also can be operated over themodulation frequency region from 1 Hz to 500 kHz. Thus,it offers great potential for discriminating single moleculesfrom autofluorescence and strong interference and simulta-

Figure 4. (a) Signal-to-background ratio (S/B) and (b) signal-to-interference ratio (S/I) in the time domain (Fluo.) and frequencydomain (DFT) as well as the corresponding enhancement (DFT/Fluo.) as a function of integration time. The signal and backgroundare selected from Figure 3. S1 and B are for S/B, and S2 and I are forS/I. All of the ratios and enhancements are fitted by the power-lawmodel, that is, R = a × tb. The corresponding parameters are in thefigures.

Figure 5. QCME-SMIM retrieves single SR molecule signal in a livecell. (a) Traditional fluorescence intensity imaging of a coli bacilluscell in the presence of a large background. (b) CorrespondingQCME-SMIM. Scale bars: 5 μm.

The Journal of Physical Chemistry Letters Letter

DOI: 10.1021/acs.jpclett.8b03606J. Phys. Chem. Lett. 2019, 10, 223−228

226

neously opens new opportunities for studying biologicalinteractions and dynamics with improved speed and sensitivity.

■ EXPERIMENTAL METHODSExperimental Procedures. The experimental setup is sketched inFigure 2a. The laser with a pulse width of 389 fs at awavelength of 635 nm was split by a beam splitter (BS). Onone arm, the laser passed through the electro-opticalmodulator (EOM) used for phase modulation, while on theother arm, the laser passed through a quarter-wave plate(QWP) to convert the laser into vertical polarization. Thus,two beams recombing at the BS have orthogonal polarizations,which can eliminate the intensity modulation due to their owninterference. The phase difference between pulse pairs wasperiodically modulated between −π and π by applying aperiodic sawtooth voltage on the EOM. The modulationfrequency was adapted between 1 Hz and 500 kHz, and themodulation voltage was set as the full wave voltage for bettermodulation depth. The delay of the pulse pair was adjustednearly to zero to address the coherent property of a two-levelsystem. The recombined light was reflected into the objectivelens via a dichroic mirror (DM) and then focused the lightonto the prepared sample. The fluorescence photons from thesingle molecules and background or interference were allcollected by the same objective, transmitted through the DM,and finally detected by a single-photon detector (τ-SPAD,).The absolute arrival time of each photon was recorded by thetime-tagged, time-resolved, and time-correlated single-photoncounting (TTTR-TCSPC) technique with a multichannelpicosecond event timer (MCPET). The temporal resolution ofthe MCPET was 1 ps. Single-molecule fluorescence imaging inthe frequency domain was obtained by demodulating thearrival time of the photon series for each pixel through DFT.Sample Preparation. In the experiment, the microspheres

(based on melamine resin, Nile Blue marked with 1 μm @Fluka Code 93887) were first dripped on the substrate (coverglass) and then irradiated with ultraviolet light for about 3 h toreduce the fluorescence intensity. After, the colloidal quantumdots (QDs, CdSeTe/ZnS, purchased from NanoOpticalMaterials Inc.) and three organic dyes, including squaraine-derived rotaxanes (SRs), 1,10-dioctadecyl-3,3,30,30-tetrame-thylindodicarbo-cyanine49 (DiD), and Cy5,50 as well asstreptavidin-bonded Cy5 (Cy5-SA), were prepared, respec-tively, by a spin-coated 20 μL solution (water for SR, Cy5 andCy5-SA and toluene for DiD) containing 10−9−10−10 M dyemolecules. Then, the poly(methyl methacrylate) (PMMA) filmwas spin-coated on the prepared dye molecules. Here, thePMMA film was used to prevent oxygen-induced fastphotobleaching. The final sample was heated under vacuumto 373 K for 3 h and then held overnight under vacuum atroom temperature to eliminate the remaining solvent and relaxthe stress induced by the spin-coating process.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.jp-clett.8b03606.

Derivation of the population probability under anarbitrary phase difference; fluorescence spectra of SRmolecules and microspheres; derivation of the S/B ratioin time and frequency domains; contrast enhancement

in QCME-SMIM under a strong background; contrastenhancement in QCME-SMIM under ambient light;universality of QCME-SMIM; applications of QCME-SMIM in living cells; and broad dynamic range ofQCME-SMIM (PDF)

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: chbqin@sxu.edu.cn (C.Q).*E-mail: xlt@sxu.edu.cn (L.X.).ORCIDHaitao Zhou: 0000-0001-5325-5275Chengbing Qin: 0000-0002-6822-5113Guofeng Zhang: 0000-0002-9030-0431Liantuan Xiao: 0000-0003-2690-6460NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThe authors gratefully acknowledge support from the NationalKey Research and Development Program of China (Grant No.2017YFA0304203), the National Natural Science Foundationof China (Grant No. 11434007), the Natural ScienceFoundation of China (Nos. 61527824, 61605104, 11504216,61675119, and U1510133), PCSIRT (No. IRT13076), and1331KSC.

■ REFERENCES(1) Lerner, E.; Cordes, T.; Ingargiola, A.; Alhadid, Y.; Chung, S.;Michalet, X.; Weiss, S. Toward Dynamic Structural Biology: TwoDecades of Single-Molecule Forster Resonance Energy Transfer.Science 2018, 359, eaan1133.(2) Dyla, M.; Terry, D. S.; Kjaergaard, M.; Sorensen, T. L. M.;Andersen, J. L.; Andersen, J. P.; Knudsen, C. R.; Altman, R. B.;Nissen, P.; Blanchard, S. C. Dynamics of P-type ATPase TransportRevealed by Single-Molecule FRET. Nature 2017, 551, 346−351.(3) Wang, X.; Li, X.; Deng, X.; Luu, D.-T.; Maurel, C.; Lin, J. Single-molecule Fluorescence Imaging to Quantify Membrane ProteinDynamics and Oligomerization in Living Plant Cells. Nat. Protoc.2015, 10, 2054−2063.(4) Monici, M. Cell and Tissue Autofluorescence Research andDiagnostic Applications. Biotechnol. Annu. Rev. 2005, 11, 227−256.(5) Mansfield, J. R.; Gossage, K. W.; Hoyt, C. C.; Levenson, R. M.Autofluorescence Removal, Multiplexing, and Automated AnalysisMethods for In-vivo Fluorescence Imaging. J. Biomed. Opt. 2005, 10,041207.(6) Cheng, J.-X.; Xie, X. S. Vibrational Spectroscopic Imaging ofLiving Systems: An Emerging Platform for Biology and Medicine.Science 2015, 350, 1054−1063.(7) Okabe, K.; Inada, N.; Gota, C.; Harada, Y.; Funatsu, T.;Uchiyama, S. Intracellular Temperature Mapping with a FluorescentPolymeric Thermometer and Fluorescence Lifetime Imaging Micros-copy. Nat. Commun. 2012, 3, 705.(8) Berezin, M. Y.; Achilefu, S. Fluorescence Lifetime Measurementsand Biological Imaging. Chem. Rev. 2010, 110, 2641−2684.(9) Cheng, J.-H.; Nicolai, B.; Sun, D.-W. Hyperspectral Imaging withMultivariate Analysis for Technological Parameters Prediction andClassification of Muscle Foods: A Review. Meat Sci. 2017, 123, 182−191.(10) Becker, W. Fluorescence Lifetime Imaging - Techniques andApplications. J. Microsc. 2012, 247, 119−136.(11) Vettenburg, T.; Corral, A.; Rodríguez-Pulido, A.; Flors, C.;Ripoll, J. Photoswitching-Enabled Contrast Enhancement in LightSheet Fluorescence Microscopy. ACS Photonics 2017, 4, 424−428.

The Journal of Physical Chemistry Letters Letter

DOI: 10.1021/acs.jpclett.8b03606J. Phys. Chem. Lett. 2019, 10, 223−228

227

(12) Querard, J.; Zhang, R.; Kelemen, Z.; Plamont, M.-A.; Xie, X.;Chouket, R.; Roemgens, I.; Korepina, Y.; Albright, S.; Ipendey, E.;et al. Resonant Out-of-Phase Fluorescence Microscopy and RemoteImaging Overcome Spectral Limitations. Nat. Commun. 2017, 8, 969.(13) Chen, Y.-C.; Dickson, R. M. Improved Fluorescent ProteinContrast and Discrimination by Optically Controlling Dark StateLifetimes. J. Phys. Chem. Lett. 2017, 8, 733−736.(14) Chen, Y. C.; Jablonski, A. E.; Issaeva, I.; Bourassa, D.; Hsiang, J.C.; Fahrni, C. J.; Dickson, R. M. Optically Modulated Photo-switchable Fluorescent Proteins Yield Improved Biological ImagingSensitivity. J. Am. Chem. Soc. 2015, 137, 12764−12767.(15) Wan, W.; Zhu, M. Q.; Tian, Z. Y.; Li, A. D. Q. Antiphase Dual-Color Correlation in a Reactant-Product Pair Imparts Ultrasensitivityin Reaction-Linked Double-Photoswitching Fluorescence Imaging. J.Am. Chem. Soc. 2015, 137, 4312−4315.(16) Wang, J.; Lv, Y. L.; Wan, W.; Wang, X. F.; Li, A. D. Q.; Tian, Z.Y. Photoswitching Near-Infrared Fluorescence from Polymer Nano-particles Catapults Signals over the Region of Noises andInterferences for Enhanced Sensitivity. ACS Appl. Mater. Interfaces2016, 8, 4399−4406.(17) Zhang, Y. L.; Zhang, K. Q.; Wang, J.; Tian, Z. Y.; Li, A. D. Q.Photoswitchable Fluorescent Nanoparticles and Their EmergingApplications. Nanoscale 2015, 7, 19342−19357.(18) Wu, Y.; Xie, Y. S.; Zhang, Q.; Tian, H.; Zhu, W. H.; Li, A. D. Q.Quantitative Photoswitching in Bis(dithiazole)ethene Enables Mod-ulation of Light for Encoding Optical Signals. Angew. Chem., Int. Ed.2014, 53, 2090−2094.(19) Li, C.; Hu, Z.; Aldred, M. P.; Zhao, L. X.; Yan, H.; Zhang, G.F.; Huang, Z. L.; Li, A. D. Q.; Zhu, M. Q. Water-Soluble PolymericPhotoswitching Dyads Impart Super-Resolution Lysosome High-lighters. Macromolecules 2014, 47, 8594−8601.(20) Tian, Z. Y.; Li, A. D. Q. Photoswitching-Enabled Novel OpticalImaging: Innovative Solutions for Real-World Challenges inFluorescence Detections. Acc. Chem. Res. 2013, 46, 269−279.(21) Yuan, L.; Lin, W.; Yang, Y.; Chen, H. A Unique Class of Near-Infrared Functional Fluorescent Dyes with Carboxylic-Acid-Modu-lated Fluorescence ON/OFF Switching: Rational Design, Synthesis,Optical Properties, Theoretical Calculations, and Applications forFluorescence Imaging in Living Animals. J. Am. Chem. Soc. 2012, 134,1200−1211.(22) Tian, Z. Y.; Wu, W. W.; Wan, W.; Li, A. D. Q. Photoswitching-Induced Frequency-Locked Donor-Acceptor Fluorescence DoubleModulations Identify the Target Analyte in Complex Environments. J.Am. Chem. Soc. 2011, 133, 16092−16100.(23) Li, A. D. Q.; Zhan, C. L.; Hu, D. H.; Wan, W.; Yao, J. N.Photoswitchable Nanoprobes Offer Unlimited Brightness in Fre-quency-Domain Imaging. J. Am. Chem. Soc. 2011, 133, 7628−7631.(24) Tian, Z. Y.; Wu, W. W.; Wan, W.; Li, A. D. Q. Single-Chromophore-Based Photoswitchable Nanoparticles Enable Dual-Alternating-Color Fluorescence for Unambiguous Live Cell Imaging.J. Am. Chem. Soc. 2009, 131, 4245−4252.(25) Hu, D. H.; Tian, Z. Y.; Wu, W. W.; Wan, W.; Li, A. D. Q.Photoswitchable Nanoparticles Enable High-Resolution Cell Imaging:PULSAR Microscopy. J. Am. Chem. Soc. 2008, 130, 15279−15281.(26) Zhu, L. Y.; Wu, W. W.; Zhu, M. Q.; Han, J. J.; Hurst, J. K.; Li,A. D. Q. Reversibly Photoswitchable Dual-color Fluorescent Nano-particles as New Tools for Live-cell Imaging. J. Am. Chem. Soc. 2007,129, 3524−3526.(27) Zhu, L. Y.; Zhu, M. Q.; Hurst, J. K.; Li, A. D. Q. Light-Controlled Molecular Switches Modulate Nanocrystal Fluorescence. J.Am. Chem. Soc. 2005, 127, 8968−8970.(28) Richards, C. I.; Hsiang, J.-C.; Dickson, R. M. SynchronouslyAmplified Fluorescence Image Recovery (SAFIRe). J. Phys. Chem. B2010, 114, 660−665.(29) Jablonski, A. E.; Hsiang, J.-C.; Bagchi, P.; Hull, N.; Richards, C.I.; Fahrni, C. J.; Dickson, R. M. Signal Discrimination BetweenFluorescent Proteins in Live Cells by Long-Wavelength OpticalModulation. J. Phys. Chem. Lett. 2012, 3, 3585−3591.

(30) Jablonski, A. E.; Vegh, R. B.; Hsiang, J. C.; Bommarius, B.;Chen, Y. C.; Solntsev, K. M.; Bommarius, A. S.; Tolbert, L. M.;Dickson, R. M. Optically Modulatable Blue Fluorescent Proteins. J.Am. Chem. Soc. 2013, 135, 16410−16417.(31) Hsiang, J. C.; Jablonski, A. E.; Dickson, R. M. OpticallyModulated Fluorescence Bioimaging: Visualizing Obscured Fluoro-phores in High Background. Acc. Chem. Res. 2014, 47, 1545−1554.(32) Petty, J. T.; Fan, C.; Story, S. P.; Sengupta, B.; Sartin, M.;Hsiang, J.-C.; Perry, J. W.; Dickson, R. M. Optically Enhanced, Near-IR, Silver Cluster Emission Altered by Single Base Changes in theDNA Template. J. Phys. Chem. B 2011, 115, 7996−8003.(33) Marriott, G.; Mao, S.; Sakata, T.; Ran, J.; Jackson, D. K.;Petchprayoon, C.; Gomez, T. J.; Warp, E.; Tulyathan, O.; Aaron, H.L.; et al. Optical Lock-in Detection Imaging Microscopy for Contrast-enhanced Imaging in Living Cells. Proc. Natl. Acad. Sci. U. S. A. 2008,105, 17789−17794.(34) Smith, D. A.; Holliger, P.; Flors, C. Reversible FluorescencePhotoswitching in DNA. J. Phys. Chem. B 2012, 116, 10290−10293.(35) Fan, C.; Hsiang, J.-C.; Dickson, R. M. Optical Modulation andSelective Recovery of Cy5 Fluorescence. ChemPhysChem 2012, 13,1023−1029.(36) Richards, C. I.; Hsiang, J. C.; Senapati, D.; Patel, S.; Yu, J. H.;Vosch, T.; Dickson, R. M. Optically Modulated Fluorophores forSelective Fluorescence Signal Recovery. J. Am. Chem. Soc. 2009, 131,4619−4621.(37) Mao, S.; Benninger, R. K. P.; Yan, Y. L.; Petchprayoon, C.;Jackson, D.; Easley, C. J.; Piston, D. W.; Marriott, G. Optical Lock-inDetection of FRET Using Synthetic and Genetically Encoded OpticalSwitches. Biophys. J. 2008, 94, 4515−4524.(38) Brinks, D.; Stefani, F. D.; Kulzer, F.; Hildner, R.; Taminiau, T.H.; Avlasevich, Y.; Mullen, K.; Van Hulst, N. F. Visualizing andcontrolling vibrational wave packets of single molecules. Nature 2010,465, 905−U905.(39) Hildner, R.; Brinks, D.; van Hulst, N. F. FemtosecondCoherence and Quantum Control of Single Molecules at RoomTemperature. Nat. Phys. 2011, 7, 172−177.(40) Hildner, R.; Brinks, D.; Nieder, J. B.; Cogdell, R. J.; van Hulst,N. F. Quantum Coherent Energy Transfer over Varying Pathways inSingle Light-Harvesting Complexes. Science 2013, 340, 1448−1451.(41) Brinks, D.; Hildner, R.; van Dijk, E.; Stefani, F. D.; Nieder, J. B.;Hernando, J.; van Hulst, N. F. Ultrafast Dynamics of SingleMolecules. Chem. Soc. Rev. 2014, 43, 2476−2491.(42) Feynman, R. P., Jr.; Vernon, F. L.; Hellwarth, R. W.Geometrical Representation of the Schrodinger Equation for SolvingMaser Problems. J. Appl. Phys. 1957, 28, 49−52.(43) Loudon, R. The Quantum Theory of Light; OUP Oxford:Oxford, 2000.(44) Zewail, A. H. Optical Molecular Dephasing: Principles of andProbings by Coherent Laser Spectroscopy. Acc. Chem. Res. 1980, 13,360−368.(45) Hu, J.; Jing, M.; Zhang, G.; Qin, C.; Xiao, L.; Jia, S.Performance of Single-Photons Communication Using the Multi-channel Frequency Coding Scheme. Opt. Express 2018, 26, 20835−20847.(46) Hu, J. Y.; Liu, Y.; Liu, L. L.; Yu, B.; Zhang, G. F.; Xiao, L. T.;Jia, S. T. Quantum Description and Measurement for Single PhotonModulation. Photonics Res. 2015, 3, 24−27.(47) Liu, Y.; Yu, B.; He, B.; Zhang, G.-F.; Xiao, L.-T.; Jia, S.-T.Single-photon Modulation Spectrum. Chin. Phys. B 2014, 23, 1674−1056.(48) Marvasti, F. Nonuniform Sampling: Theory and Practice;Springer Science & Business Media: Berlin, 2012.(49) Sasatsu, M.; Onishi, H.; Machida, Y. In Vitro and in VivoCharacterization of Nanoparticles Made of MeO-PEG Amine/PLABlock Copolymer and PLA. Int. J. Pharm. 2006, 317, 167−174.(50) Mujumdar, R. B.; Ernst, L. A.; Mujumdar, S. R.; Lewis, C. J.;Waggoner, A. S. Cyanine Dye Labelling Reagents: SulfoindocyanineSuccinimidyl Esters. Bioconjugate Chem. 1993, 4, 105−111.

The Journal of Physical Chemistry Letters Letter

DOI: 10.1021/acs.jpclett.8b03606J. Phys. Chem. Lett. 2019, 10, 223−228

228