Está en la página 1de 8

Article pubs.acs.

org/ac

Integrating Sphere Setup for the Traceable Measurement of Absolute Photoluminescence Quantum Yields in the Near Infrared
Christian Wurth, Jutta Pauli, Cornelia Lochmann, Monika Spieles, and Ute Resch-Genger*
BAM Federal Institute for Materials Research and Testing, Richard-Willstaetter-Strasse 11, 12489 Berlin, Germany
S * Supporting Information

ABSTRACT: There is an increasing interest in chromophores absorbing and emitting in the near-infrared (NIR) spectral region, e.g., for applications as fluorescent reporters for optical imaging techniques and hence, in reliable methods for the characterization of their signal-relevant properties like the fluorescence quantum yield (f) and brightness. The lack of well established f standards for the NIR region in conjunction with the need for accurate f measurements in transparent and scattering media encouraged us to built up an integrating sphere setup for spectrally resolved measurements of absolute fluorescence traceable to radiometric scales. Here, we present the design of this setup and its characterization and validation including an uncertainty budget for the determination of absolute f in the visible and NIR. To provide the basis for better measurements of f in the spectral window from ca. 600 to 1000 nm used, e.g., for optical imaging, the absolute f of a set of NIR chromophores covering this spectral region are measured and compared to relative values obtained using rhodamine 101 as f standard. Additionally, the absolute f values of some red dyes that are among the most commonly used labels in the life sciences are presented as well as the absolute quantum yield of an optical probe for tumor imaging.

entral advances in biochemical assays, molecular sensor technologies, and molecular optical imaging are related to fluorescent reporters that provide a high detection sensitivity for molecular processes in conjunction with a high selectivity. The most fundamental spectroscopic properties of such functional chromophores are the fluorescence quantum yield f, the molar absorption coefficient at the excitation wavelength (ex), and for some applications, also the fluorescence lifetime.14 Of special importance is the fluorometric key parameter f, i.e., the number of emitted photons Nem per number of absorbed photons Nabs that presents a direct measure for the efficiency of the conversion of absorbed into emitted light; see eq 1.5,6 Dye performance is typically assessed using the brightness ((ex) f) or the size of f7 and, for nanocrystalline chromophores like quantum dots, f presents a powerful tool to assess their quality.8,9

f =

Nem() Nabs()

(1)

Over the past decades, considerable efforts have been dedicated to develop reliable relative (standard needed) and absolute (standard-free) methods for the determination of f.6,1021 This included optical methods as well as photothermal methods like photoacoustic spectroscopy and thermal lensing that determine this quantity indirectly from the measurement of dissipated heat.11 Commonly, f is obtained optically comparing the absorption-weighted integral fluorescence intensities of a sample and a standard of known f.22 Main factors governing the accuracy of these relative optical
2011 American Chemical Society

methods are the accuracy of the radiometric spectrofluorometer characterization and the reliability of the standard's f value.4,10,2325 This value must be precisely known for the measurement conditions employed (e.g., solvent/matrix, excitation wavelength ex, temperature, etc.).26 However, at present, f values of only very few fluorophores emitting in the visible (vis) region like rhodamine 101, rhodamine 6G, fluorescein, and quinine sulfate dihydrate are well established with relative uncertainties of ca. 5%, typically for especially purified dyes not commercialized.11,22,27 Although functional chromophores absorbing and emitting above 600 nm are increasingly being used in biology, molecular imaging, and clinical diagnostics,2833 no dependable quantum yield standards for the near-infrared (NIR) spectral region are currently available.8,11 For example, in the case of the frequently employed NIR quantum yield standard indocyanine green (ICG), one of the most fundamental publication states that the f values provided present only trends due to questionable dye purity.34 Moreover, considerable uncertainties in the size of the f value of IR26 led to an overestimation of the photoluminescence quantum yields of many PbSe and PbS quantum dots determined relative to this dye as recently disclosed.8 Alternatively, the number of emitted photons Nem(ex) per number of absorbed photons Nabs(ex), see eq 1, can be measured absolutely with an integrating sphere setup, thereby
Received: August 31, 2011 Accepted: December 22, 2011 Published: December 22, 2011
1345
dx.doi.org/10.1021/ac2021954 | Anal. Chem. 2012, 84, 13451352

Analytical Chemistry circumventing standard-related uncertainties8,12,3537 Although such setups are more and more available, even calibrated, proper consideration of detector nonlinearities and reabsorption effects still present major challenges.22 The increasing importance of reliable fluorescence measurements caused us to assess methods, equipment, and standards for the measurement of f of application-relevant chromophores in the visible (vis) region and derive achievable uncertainties.10,22,24,38,39 These results, including the limitations of commercialized equipment,22,40 encouraged us to develop a new integrating sphere setup for the spectrally resolved measurements of absolute fluorescence quantum yields of transparent and scattering samples in the wavelength region of 400 to 1000 nm with measurement uncertainties 5% (for f values 10%). Other requirements included the suitability for direct and indirect sample illumination18 and the possibility to adjust the size of the excitation light beam to control the size of the illuminated sample volume. Here, we present the design of our integrating sphere setup and the calibration and validation strategies used including an uncertainty budget for the measurement of absolute fluorescence quantum yields of transparent dye solutions in the vis and NIR. As a first step to reliable f measurements in the long wavelength region, we assessed the potential of three commercialized red dyes as potential quantum yield standards for the spectral region of 600 to 1000 nm and present f values of common bioanalytically relevant vis and NIR emitters and an optical probe in mouse serum recently used for tumor imaging.

Article

MATERIALS AND INSTRUMENTATION Materials. Emission Standards. Quinine sulfate dihydrate (QS; equaling SRM 936a) was obtained from the National Institute of Standards and Technology (NIST).41 The spectral fluorescence standards F004 and F005 are from BAM and Sigma Aldrich.23,42 F007 is a new NIR emission standard currently evaluated by BAM. 4-(dicyanomethylene)-2-methyl6-(p-dimethylaminostyryl)-4H-pyran (DCM) was purchased from Lambda Physics (batch number 029401). All dyes were of the highest purity commercially available. Quantum Yield Standards. Rhodamine 101 (R101, batch number 019502), rhodamine 6G (R6G; batch number 119202), and the NIR dyes oxazine 1 (OXA1; batch number 090214), HITCI (batch number 029006), and IR 125 (batch number 10970) were obtained from Lambda Physics (see Supporting Information, Table 1S, for chemical structures). NIR Dyes, Labels, and Probes. Indocyanine green (ICG; 335601) was purchased from Pulsion Medical Systems, cryptocyanine (CRCY; batch 13715JZ) from Sigma Aldrich, Cy5, Cy5.5, and Cy7 from GE Healthcare Europe (batch numbers 392929, 398887 (Cy5.5 mono NHS ester, hydrolyzed; used for the preparation of RGD-Cy5.5), and 392393), Atto 740 from ATTO-Tec (batch number 131450324506P10), Atto 590 and Atto 680 from Fluka (batch numbers 70425 and 94875), and Alexa 647 and Alexa 750 from Invitrogen (batch numbers 822686 and 866762). The chemical structures of selected dyes are summarized in the Supporting Information (Table 2S). The synthesis of the new optical probe RGD-Cy5.5 was previously described.43 Solvents. The solvents used, i.e., ethanol for R101, R6G, OXA1, HITCI, CRCY, F004, F005, and F007, methanol for DCM, and dimethylsulfoxide (DMSO) for IR 125 were of spectroscopic grade and purchased from Sigma Aldrich and Merck. Phosphate buffer solution (PBS; Hank's balanced salt
1346

solution; Ca(II)- and Mg(II)-free without phenol red; pH = 7.8) used for ICG was obtained from PAA Laboratories (Pasching, Austria). Tris buffer (pH = 7.4) was prepared from tris-(hydroxymethyl)aminomethane and hydrochloric acid from Merck. Water equals always bidistilled water (pH ca. 7.0). Bovine serum albumin (BSA; fraction V) and silica particles were purchased from Merck. The mouse serum was selfextracted. All solvents were controlled for the presence of fluorescent impurities measuring blank spectra of the solvent for typical excitation wavelengths. Instrumentation. Relative Measurement of f. Absorption spectra were recorded on a calibrated Cary 5000 spectrometer. Fluorescence spectra were measured with a calibrated Spectronics Instruments 8100 with Glan Thompson polarizers in the excitation and the emission channel(s) set to 0 and 54.7.10,23,44 All fluorescence emission and excitation spectra shown are blank-corrected and corrected for instrument-specific effects (see Supporting Information)23,25,45 Relative f values using a chain of f transfer standards consisting of n dyes with R101 acting as gold standard40 were calculated according to the formula of Demas and Crosby6 (see Supporting Information, eq 1S). Absolute Measurement of f. Absolute f values were determined with the newly built integrating sphere setup detailed in the Results and Discussion section. Additionally, absolute f values of selected dyes were measured with an integrating sphere setup C9920-02 from Hamamatsu Photonics.22 Measurement Conditions. All absorption and fluorescence measurements were performed with air-saturated freshly prepared dye solutions46 at T = (25 1)C using 10 mm 10 mm as well as 4 mm 10 mm quartz cuvettes from Hellma GmbH. For measurements with the commercial integrating sphere setup, 10 mm 10 mm long-necked quartz cuvettes from Hamamatsu Photonics K.K. were employed, always filled with 3.5 mL of solvent or dye solution.22 For instrument calibration, at all times, expanded uncertainties (k = 2, confidence interval of 95%) are given. For all wavelengthdependent quantities (see Supporting Information, Table 3S), we used the relative standard deviation of the mean value, averaged over the relevant wavelength region. Dye Purity. HPLC analysis of the dyes detailed in the Supporting Information (Tables 4S and 5S) yielded the following purities: R101, 95.5% (525 nm) and 97.4% (565 nm); R6G, > 98.5% (480 nm, 530 nm); HITCI, 97.9% (760 nm); OXA1, 98.2% (665 nm); CRCY, 98.6% (725 nm); IR125, 99.1% (800 nm); and ICG, 98.8% (800 nm). The purity of QS of at least 98% follows from the NIST certificate and certification report,41 and the purity of the BAM spectral standards is at least 98.5%.23,42 All dyes were used without further purification. Safety Considerations. Proper safety procedures for handling, storage, and disposal of the organic dyes should be observed.

RESULTS AND DISCUSSION Design, Characterization, andHank Validation of the Integrating Sphere Setup. Instrument Design. For spectrally resolved measurements of absolute quantum yields in the wavelength region from ca. 400 to 1000 nm with versatile excitation and optimum spectral resolution, we used a 450 W xenon lamp, coupled to a single monochromator as excitation channel and an integrating sphere with a diameter of ca. 15 cm,
dx.doi.org/10.1021/ac2021954 | Anal. Chem. 2012, 84, 13451352

Analytical Chemistry coated with Spectraflect (Labsphere; sphere reflectivity of about 97% in the vis/NIR), as sample compartment (Figure 1). The

Article

Figure 1. Design principle of the integrating sphere setup (top view) and measurement principle.

integrating sphere is coupled with a quartz fiber bundle to a f/4 Czerny Turner imaging spectrograph (Shamrock 303i, Andor Technology PLC) attached to a Peltier cooled (183K) thinned back side illuminated deep depletion charge coupled device (CCD array; 1024 256 pixel, DU420A-BRDD, Andor Technology). The CCD has a peak quantum yield of 94% at 790 nm with pixel sizes of 26 m 26 m and an image area of 26.6 mm 6.7 mm. The optical fiber is shielded from direct reflexes with several baffles to guarantee the detection of only diffuse radiation. Measurements in the spectral range of 400 to 1000 nm require the use of multiple positions of the gratings in the detection channel, rendering the stability of the radiant power of the excitation light source a key feature for the accomplishable measurement uncertainty. To account for its

fluctuations, a Peltier-cooled reference detector (customdesigned silicon trap detector)47 is implemented into the setup which is triggered with the shutter pulse of the CCD. To realize multiple measurement configurations, the integrating sphere was equipped with several ports for the coupling of light into the sphere and for sample mounting. For the determination of the absolute f with direct and indirect sample illumination, the sample is center-mounted inside the integrating sphere using a Spectraflect-coated cuvette holder positioned with a HeNe laser. The excitation light is focused with two off-axis parabolic mirrors into the middle (sample position) of the integrating sphere, thereby imaging the exit slit of the excitation monochromator onto the middle of the cuvette. To enable measurements of small sample volumes, the exit slit of the excitation monochromator can be changed in width and height, thus allowing the adjustment of the illuminated volume. The radiant power entering the integrating sphere is controlled by an aperture in front of the integrating sphere. Instrument Characterization. Prerequisites for accurate measurements of absolute f are the reliable determination of the wavelength accuracy of the excitation and detection channel(s), the linearity of the detection channel, the long-term stability of the radiant power of the excitation light source reaching the sample, and the spectral responsivity of the integrating sphere-spectrograph-CCD ensemble. Wavelength Accuracy. Control of the wavelength accuracy of the emission monochromator with a low pressure mercury/ argon discharge lamp25,26 (see Supporting Information, Figure 1S) reveals maximum deviations <0.4 nm (uncertainty <0.14 nm). Linear Range of Detection. The linear range of the CCD detector was obtained via controlled variation of the radiant power reaching the detector using neutral density filters of known transmission (see Supporting Information, Figure 2S), with relative deviations from a perfectly linear behavior amounting to 0.015%. Long-Term Stability of the Excitation Light Source. For the determination of the long-term stability of the light source, the radiance of the sphere wall was recorded multiple times under

Figure 2. Left panel: Comparison of the corrected emission spectra of QS (SRM 936a), F004, F005 and the new BAM NIR standard candidate F007, either certified or measured with a calibrated fluorometer (lines) with the emission spectra recorded with the calibrated integrating sphere spectrometer (open circles). Right panel: Comparison of the emission spectra of DCM in methanol recorded with the integrating sphere setup (open circles), the calibrated fluorometer of the National Metrology Institute PTB (solid line), and with literature data1 (broken line); for clarity reasons, we omitted the corrected emission spectrum of DCM obtained with our calibrated spectrofluorometer that also matches with the corrected emission spectrum of this dyes measured with the integrating sphere setup.
1347
dx.doi.org/10.1021/ac2021954 | Anal. Chem. 2012, 84, 13451352

Analytical Chemistry

Article

Figure 3. Left panel: Influence of reabsorption effects on measured, spectrally corrected emission spectra of R101 in EtOH obtained with the calibrated BAM spectrofluorometer (solid symbols), the integration sphere setup from Hamamatsu (open symbols), and the BAM integrating sphere setup (solid line). Right panel: Corrected emission spectra of a 2.1 106 M ethanolic solution of R101 recorded with our integrating sphere using different sample volumes (3500 L: solid squares; 350 L: open circles) in comparison to the undisturbed emission spectrum (solid circles) determined with spectrofluorometer 8100 (absorbance < 0.05).

the same measurement conditions (integration time, number of accumulations) as used for typical f measurements without and with reference detector; see also Supporting Information (Figure 3S). The resulting relative deviations of the mean values (integration over the entire peak) of 2% without reference detector can be reduced to a maximum of <0.4% with the aid of the reference detector, with typical relative deviations of the mean value being <0.2%. This considerably reduces the measurement uncertainties of absolute f. Spectral Responsivity. The spectral responsivity of the integrating sphere-spectrograph-CCD ensemble was determined with a calibrated lamp as detailed in the Supporting Information (see Figure 4S). The accuracy of the emission correction curve was controlled by determining the corrected emission spectra of the established emission standards QS, BAM F004, and BAM F0052527,42 and our new NIR standard candidate F007 with the integrating sphere setup and a calibrated spectrofluorometer previously used for the certification of the BAM emission standards (Figure 2, left panel). All UV/vis/NIR fluorophores were chosen to show minimum spectral overlap between absorption and emission and broad and unstructured emission bands to minimize reabsorption and bandpass effects.23 The small deviations found between the corrected emission spectra obtained with the calibrated integrating sphere setup and the certified (QS, F004, F005) or corrected emission spectra obtained with our calibrated fluorometer (F007) shown in Figure 2 (left panel) underline the reliability of our emission correction curve for the wavelength region of 375 to 1000 nm. Additionally, we compared the corrected emission spectrum of DCM in methanol, a recommended emission standard,1 measured with our integrating sphere and our spectrofluorometer with the corrected emission spectrum obtained with the calibrated spectrofluorometer of the Physikalisch-Technische Bundesanstalt (PTB) and with literature data (Figure 2, right panel). These results underline the accuracy of our emission
1348

correction and, concomitantly, the need for reliable emission standards for instrument validation as indicated by the clear deviation of the literature spectrum of DCM1 from the corrected emission spectra obtained with three independently calibrated instruments. Sensitivity to Reabsorption Effects. For fluorophores that display a considerable spectral overlap between their absorption and emission bands like most bioanalytically relevant dyes,7 reabsorption effects have a considerable impact on fluorescence quantum yields determined with an integrating sphere setup, leading to underestimated f.22 Minimization of reabsorption effects is thus a prerequisite for the measurement of absolute f with minimum uncertainty. From the instrument side, the size of such effects is controlled by all parameters that affect the distance traveled by the emitted photons within the sample prior to detection like the wavelength- and material-dependent reflectivity of the sphere walls, the size of the integrating sphere, and the sample volume. The sensitivity of our integrating sphere setup to reabsorption effects was assessed by measuring the spectrally corrected fluorescence spectra of differently concentrated R101 solutions and comparing these data to the spectra of identical dye solutions (same sample volume) obtained with our spectrofluorometer and the integrating sphere setup from Hamamatsu (Figure 3, left panel). Although the emission spectra obtained with our integrating sphere setup (sphere diameter of ca. 15 cm; sphere reflectivity of about 97% in the vis/NIR) also reveal reabsorption-induced red shifts, the size of these effects is much less pronounced compared to the results obtained with the commercialized setup (sphere diameter of ca. 10 cm; sphere reflectivity of about 99% in the vis/NIR). Further minimization of reabsorption effects can be achieved by, e.g., reducing the sample volume. As shown in Figure 3 (right panel) for a 2.1 106 M ethanolic solution of R101, fdistorting reabsorption effects can be eliminated even for concentrated solutions of dyes with a small Stokes shift due to
dx.doi.org/10.1021/ac2021954 | Anal. Chem. 2012, 84, 13451352

Analytical Chemistry

Article

Figure 4. Left panel: Normalized absorption factors f() (solid lines) and normalized corrected fluorescence excitation spectra (open circles) of R101, OXA1, HITCI, and IR125 measured with spectrofluorometer 8100; Right panel: Normalized corrected emission spectra measured with spectrofluorometer 8100 (solid lines) and our integrating sphere setup (open symbols); for the latter measurement, very dilute dye solutions and cuvettes with a small sample volume/optical path length were used to minimize reabsorption effects. The excitation wavelengths employed for the determination of the absolute fluorescence quantum yields are indicated with arrows in the left panel.

Table 1. Absolute f Values of R6G, R101, and the Recommended NIR Standards OXA1, HITCI, and IR125 Determined with the BAM Integrating Sphere Spectrometer and Pairwise Relatively Using Spectrofluorometer 8100 and the Absolute Instrument-Related Measurement Uncertainties (f) for Each Methoda
dye R6G R101 OXA1 HITCI IR125
a

solvent EtOH EtOH EtOH EtOH DMSO

f absolute 0.916 0.915 0.148 0.297 0.228

f absolute 0.028 0.028 0.005 0.009 0.007

nabs 7 66 11 5 5

f relative 0.908 standard 0.915 0.138 0.316 0.207

f relative 0.033 standard 0.028 0.006 0.015 0.012

nrel 2 6 6 6

f lit 0.89740 0.913,48 0.940 0.13248 0.28348 0.22053

f lit 0.05 0.046, 0.05 0.008 0.017

The following excitation wavelengths (ex) were used R6G, ex = 500 nm; R101, ex = 525 nm; OXA1, ex = 600 nm; HITCI, ex = 688 nm; and IR125, ex = 730 nm. nabs and nrel are the number of independent measurements performed for the absolute and relative measurements. In the case of the absolute measurements, sample and blank were always removed from the integrating sphere after each measurement. For comparison, we also included selected literature data on relative f values. All measurements from our lab have been performed with identical dye batches.

the unique possibility to adjust the spot size of the excitation light beam, consequently enabling one to reduce the illuminated volume of the sample. Mathematical consideration of reabsorption effects was accomplished with eq 4S (see Supporting Information). Measurement of Absolute Quantum Yields. The determination of absolute f with our integrating sphere setup includes the following steps: (i) measurement of the transmitted incident radiant power and the emission spectrum of the sample (fluorophore in a matrix, here a solvent) and a blank (solvent-filled cuvette) under identical measurement conditions (e.g., excitation wavelength, temperature, cell position), each within a single scan, (ii) data evaluation including choice of the excitation and emission wavelength region for spectral emission correction and signal integration, and (iii) calculation of the fluorescence quantum yield using eq 2. Here, ex and em are the excitation and emission wavalength, respectively; I presents the measured, dark count-corrected and spectrally uncorrected spectrum, and s(ex) is the spectral responsivity of the detection channel (see Supporting Information, Figure 4S, top panel). The number of absorbed photons follows from the difference between the transmitted radiant power of the blank (Ib(ex)) and the sample (Is(ex)) within the spectral region of the excitation bandpass ex and the number of emitted photons from the integration of the blank-corrected and spectrallycorrected emission spectrum of the sample (Is(em) Ib(em)). For dyes with a small Stokes shift like most NIR dyes studied here, we employed a reabsorption correction of the measured
1349

f values (see Supporting Information, eq 4S).40


d em em N f = = em + Ib(ex) Is(ex) Nabs d ex s(ex) ex ex ex

em2 Is(em) Ib(em)


em1

s(em)

(2)

Uncertainties of Absolute f Measurements. Calibration-, method-, and sample-related uncertainties all contribute to the variation of f.22 Calibration-related uncertainties originate from (i) the inaccuracy of the wavelength scale (excitation; emission), (ii) the nonlinearity of the detection system(s), and (iii) the determination of the spectral responsivity of the detection system (see Supporting Information, Table 3S). Variations of the calibration-relevant radiometric quantities of the physical transfer standards (PTS) used were minimized by the choice of high quality PTS calibrated with minimum uncertainties.40 Other uncertainties arise from, e.g., cuvette positioning, dark- and background corrections, fluctuations of the radiant power reaching the sample, and the homogeneity of the reflectivity of the sphere wall coating. Main sample-related uncertainties include dye-specific, matrix-, and concentrationdependent reabsorption effects,40 the size of which is instrument-dependent for integrating sphere setups (e.g., Figure 3), and dye-specific impurities (see Supporting Information, Table 5S and Figure 6S). Such effects were minimized, and the size of their contributions were estimated here (see sections Sensitivity to Reabsorption Effects and Dye
dx.doi.org/10.1021/ac2021954 | Anal. Chem. 2012, 84, 13451352

Analytical Chemistry Purity and Supporting Information). With our instrument design and the procedures employed for calibration and data analysis, we could accomplish measurement uncertainties <3% (dye purity not considered) as compared to 5% at best reachable with the commercial setup recently evaluated by us.40 Fluorescence Quantum Yields of Candidate f Standards for the NIR. To provide fluorescence spectroscopists with reliable tools for the accurate determination of relative f also in the NIR, we determined the absolute fluorescence quantum yields of R101 and the f standard candidates OXA1, HITCI, and IR125 emitting in the wavelength region from ca. 500 to 1000 nm (Figure 4, right panel; Table 1 and Supporting Information, Table 1S). These dyes that were chosen because of their good photochemical and thermal stability and their commercial availability in a relatively high purity (see Supporting Information, Table 5S) present a quantum yield transfer chain for the vis/NIR spectral region, thereby extending our previously reported UV/vis quantum yield transfer chain.10,40 The f values of these potential quantum yield standards are independent of excitation wavelength as indicated by their matching absorption and corrected excitation spectra (Figure 4, left panel). Parallel to the absolute f measurements, we determined the relative f values of these NIR dyes with our calibrated spectrofluorometer (calibration uncertainty of 2%, measurement uncertainty of <3% for f > 0.10) using pairs of reference and standard dyes, i.e., OXA1 vs R101, HITCI vs OXA1, and IR125 vs HITCI, and identical excitation wavelengths for each sample-standard pair.40 For the calculation of the relative f values (see Supporting Information, eq 1S), we used a fluorescence quantum yield of R101 of 0.915 as determined absolutely (Table 1). As follows from Table 1, the absolute f values of R6G and R101 are in good agreement with previous measurements using a commercialized integrating sphere spectrometer.40 Due to the minimum calibration uncertainty accomplished here, the measurement uncertainties could be considerably diminished. The absolute and relative f values match within their stated uncertainties. In the case of IR125, absolute measurements result in smaller uncertainties than the use of the quantum yield transfer chain R101OXA1HITCIIR125 as the combined standard deviations of the f values increase with the number n of standard dyes employed (see Supporting Information, eq 2S).40 The fluorescence quantum yield of IR125 equaling in chromophore structure ICG (see Supporting Information, Tables 1S and 2S and next section) was confirmed to be 0.23 in DMSO.33,34 Second, to establish the need for indirect sample illumination for transparent solutions and to assess the overall sensitivity of our integrating sphere setup and the suitability of the recently suggested NIR f standard CRCY,48 we measured the absolute f of CRCY and our NIR dyes using both illumination geometries.22,35,4952 For all dye solutions, direct and indirect illumination as well as a combination of both geometries yielded identical results, suggesting that, for transparent matrixes, direct sample illumination is sufficient. In the case of CRCY, our absolute f value of 0.012 0.002 was in excellent agreement with the relative value of 0.012 0.001 recently published;48 however, the relative uncertainty of the f value of CRCY of 17% considerably exceeds the uncertainties accomplishable with the NIR f standards recommended by us, all showing f values >0.10. Hence, use of weakly emissive dyes
1350

Article

(f 0.10) with a strong overlap of absorption and emission as quantum yield standards is not recommended. Fluorescence Quantum Yields of Bioanalytically Relevant Red Dyes. As a further step to improve the reliability of f data in the life sciences, we determined the absolute f of common red fluorescent reporters. This includes the clinically approved NIR contrast agent ICG53,54 matching IR125 in chromophore structure IR125 and Cy5, Atto 590, Atto 680, Atto 740, Alexa 647, and Alexa 750 employed, e.g., as labels in fluorescence assays and for the design of optical probes for molecular imaging.5558 Available chemical structures are shown in the Supporting Information (see Table 2S). As follows from Table 2, all fluorescence quantum yields measured Table 2. Fluorescence Quantum Yields of NIR Fluorophores Commonly Used in the Life Sciences and Information from the Manufacturer's Homepage for Water As Well As Literature Dataa
dye Alexa 647 Alexa 750 Atto 590 Atto 680 Atto740 Cy5 ICG
a

solvent bidistilled water bidistilled water Tris buffer Tris buffer bidistilled water bidistilled water PBS

f 0.33 0.123 0.729 0.239 0.053 0.261 0.027

f 0.013 0.004 0.023 0.008 0.002 0.008 0.002

n 2 2 3 5 2 2 5

data from dye manufacturers or from the literature 0.33b 0.12b 0.80b 0.30b 0.10b 0.2758 0.0453

The pH values of bidistilled water, Tris buffer, and PBS were ca. 7.0, 7.4, and 7.8. n equals the number of independent measurements performed. The available chemical structures are shown in the Supporting Information (Table 2S). b Information from the manufacturer's homepage.

agree well with the data provided by the dye manufacturers, except for the f values of the Atto dyes. Here, our values deviate from the manufacturer's data by up to 53% in the case of Atto 740. Absolute Fluorescence Quantum Yields of Labels and Probes in Biologically Relevant Matrixes and Scattering Systems. As a representative example for a bioanalytically relevant system, we determined the fluorescence quantum yields of the new contrast agent Cy5.5-RGD43 and its parent fluorophore Cy5.5 in PBS, in PBS containing 5 mass % BSA modeling body fluids, and in slightly scattering mouse serum absolutely with our integrating sphere setup and relatively using OXA1 (f = 0.15)43 as standard (Table 3). As follows from Table 3, attachment of Cy5.5 to the cyclic peptide RGD that contains only a single fluorophore binding site in contrast to biomolecules like antibodies or antibody fragments53,5962 barely affects the dye's fluorescence quantum yield. Moreover, the presence of BSA and possibly other serum proteins as well results in a fluorescence enhancement. The relative and absolute f values agree well, except for Cy5.5 and Cy5.5RGD in slightly scattering mouse serum. Here, the relative f deviates by ca. 35% from the absolute f. The systematic underestimation of fluorescence quantum yields by relative measurements already in a slightly scattering medium is ascribed to an overestimation of the dye absorption at the excitation wavelength despite a baseline correction (i.e.,
dx.doi.org/10.1021/ac2021954 | Anal. Chem. 2012, 84, 13451352

Analytical Chemistry Table 3. Fluorescence Quantum Yields of Cy5.5 and Cy5.5RGD in Different Biologically Relevant Matrixesa
dye Cy5.5 Cy5.5-RGD Cy5.5 Cy5.5-RGD Cy55 Cy5.5-RGD
a

Article

solvent PBS PBS PBS/5% BSA PBS/5% BSA serum serum

f absolute 0.284 0.318 0.403 0.395 0.486 0.505

f absolute 0.006 0.008 0.007 0.009 0.008 0.009

n 2 2 4 5 6 6

f relative 0.2843 0.3243 0.4343 0.4143 0.34 0.31

n equals the number of independent measurements performed.

subtraction of the absorption spectrum of the mouse serum from the absorption spectrum of the dye-containing solutions; see Supporting Information, Figure 5S), in conjunction with a scattering-induced attenuation of the excitation light and a change in optical pathway for exciting and emitted photons. The latter affects the correct determination of the number of emitted photons. These results clearly underline the limitations of relative measurements of f for many biologically and bioanalytically relevant samples in scattering media like serum, blood, and tissue or for suspensions of nanometer-sized particles increasingly discussed only recently.37,63,64 Clearly advantageous is here the use of an integrating sphere that collects all scattered and emitted photons and enables the determination of the fraction of light absorbed by a chromophore even for an unknown ratio of the scattering and absorption coefficient of the fluorophore matrix. We also assessed the performance of our integrating sphere setup for the measurement of the absolute fluorescence quantum yields of dyes in scattering, yet nonabsorbing, matrixes. As model systems, ethanolic solutions of R101 (absorbance of 0.1 at ex) containing various amounts of 250 nm-sized silica particles as nonquenching scatterers were used. Here, the solvent without any scatter was used as blank, thereby deliberately avoiding the adaptation of the scattering properties of the blank to that of the sample by addition of the same concentration of the known scatter as often; neither the chemical composition(s)/scattering coefficient(s) nor the concentration(s) of scatterers are known. For solutions containing bead concentrations of 1 wt % or less, the resulting f values of R101 deviated by <2% from that of the dye in ethanol (Table 1). This underlines the suitability of our setup also for absolute f measurements of scattering systems.

volumes (e.g., 350400 L of minimum volume compared to 3.03.5 mL of required for other commercial integrating sphere setups22), our integrating sphere is especially suited for measurements of biological and bioanalytical samples that are often available only in small quantities. Its potential for the determination of the absolute fluorescence quantum yields of nm- and m-sized fluorescent particles and the determination of the absorption coefficients of such beads is currently being systematically assessed for several fluorophores and differently sized beads. In addition, to provide the necessary tools for reliable quantum yields in the near-infrared, we determined the absolute fluorescence quantum yields of the three NIR dyes OXA1, HITCI, and IR125 covering the spectral region of ca. 550 to 800 nm in excitation and ca. 600 to 1000 nm in emission. These dyes are recommended as candidate f standards due to their moderate and excitation wavelengthindependent f (f 0.1), good thermal and photochemical stability, and commercial availability in high purity. These measurements present an important step to the standardization of fluorescence measurements in the increasingly used NIR.

ASSOCIATED CONTENT

S * Supporting Information

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION

Corresponding Author

*Email: ute.resch@bam.de.

ACKNOWLEDGMENTS We express our gratitude to Dr. B. Ebert, PhysikalischTechnische Bundesanstalt (PTB), Germany, for the measurement of the corrected emission spectrum of DCM, to Dr. A. Lehmann, BAM-1.5, for help with the HPLC measurements, to Dr. W. Bremser for many fruitful discussions of the uncertainty budgets, and to Prof. Dr. F. Alves, MPI Experimental Medicine, Germany, for supply of the optical probe Cy5.5-RGD. Financial support from the German Ministry of Economics and Technology (BMWi; MNPQ-Vh 22-06 and 17/07) is gratefully acknowledged.
(1) Lakowicz, J. R. Principles of fluorescence spectroscopy, 3rd ed.; Springer Science+Business Media, LLC: New York, 2006. (2) Valeur, B. Molecular Fluorescence: Principles and Application; Wiley-VCH: Weinheim, 2002. (3) Jameson, D. M.; Croney, J. C.; Moens, P. D. J. Biophotonics, Part A 2003, 360, 143. (4) Resch-Genger, U.; Hoffmann, K.; Nietfeld, W.; Engel, A.; Neukammer, J.; Nitschke, R.; Ebert, B.; Macdonald, R. J. Fluoresc. 2005, 15, 337362. (5) Demas, J. N. Measurement of photon yields; Academic Press: New York, 1982. (6) Demas, J. N.; Crosby, G. A. J. Phys. Chem. 1971, 75, 9911024. (7) Resch-Genger, U.; Grabolle, M.; Cavaliere-Jaricot, S.; Nitschke, R.; Nann, T. Nat. Methods 2008, 5, 763775. (8) Semonin, O. E.; Johnson, J. C.; Luther, J. M.; Midgett, A. G.; Nozik, A. J.; Beard, M. C. J. Phys. Chem. Lett. 2010, 1, 24452450. (9) Ziegler, J.; Merkulov, A.; Grabolle, M.; Resch-Genger, U.; Nann, T. Langmuir 2007, 23, 77517759. (10) Grabolle, M.; Spieles, M.; Lesnyak, V.; Gaponik, N.; Eychmuller, A.; Resch-Genger, U. Anal. Chem. 2009, 81, 62856294.
1351
dx.doi.org/10.1021/ac2021954 | Anal. Chem. 2012, 84, 13451352

CONCLUSION AND OUTLOOK We designed a new integrating sphere setup for spectrally resolved measurements of absolute fluorescence quantum yields in the wavelength region of ca. 400 to 1000 nm traceable to the spectral radiance scale with relative measurement and calibration related uncertainties of <3% for samples with f 0.10. For weakly emissive chromophores with f in the order of 0.01, uncertainties of 20% could be realized. Such uncertainties cannot be reached with the commercialized equipment assessed by us.22,40 Prerequisites for this high accuracy are the use of high quality physical transfer standards (see Supporting Information), the unique design strategy of our setup that permits measurements of small sample volumes in the L range and minimizes variations related to sample misalignment, nonlinearities of the detection system, and reabsorption effects. Moreover, due to the possibility of measuring small sample

REFERENCES

Analytical Chemistry
(11) Rurack, K. In Standardization and Quality Assurance in Fluorescence Measurements I: Techniques; Resch-Genger, U., Ed.; Springer: Berlin-Heidelberg, 2008; Vol. 5. (12) Suzuki, K.; Kobayashi, A.; Kaneko, S.; Takehira, K.; Yoshihara, T.; Ishida, H.; Shiina, Y.; Oishi, S.; Tobita, S. Phys. Chem. Chem. Phys. 2009, 11, 98509860. (13) Vavilov, S. I. Z. Phys. 1924, 22, 266272. (14) Velapoldi, R. A.; Tonnesen, H. H. J. Fluoresc. 2004, 14, 465 472. (15) Magde, D.; Wong, R.; Seybold, P. G. Photochem. Photobiol. 2002, 75, 327334. (16) Eaton, D. F. Pure Appl. Chem. 1990, 62, 16311648. (17) Weber, G.; Teale, F. W. J. Trans. Faraday Soc. 1957, 53, 646 655. (18) de Mello, J. C.; Wittmann, H. F.; Friend, R. H. Adv. Mater. 1997, 9, 230232. (19) Beese, W.; Zink, J. I. J. Lumin. 1984, 29, 119122. (20) Galanin, M. D.; Kufenko, A. A.; Smorchkov, V. N.; Timofee, Y. P.; Chizhikov, Z. A. Opt. Spektrosk. (USSR) 1982, 53, 683689. (21) Melhuish, W. H. J. Res. Natl. Bureau Stand. Sect. C-Eng. Instrum. 1972, A 76, 547560. (22) Wurth, C.; Lochmann, C.; Spieles, M.; Pauli, J.; Hoffmann, K.; Schuttrigkeit, T.; Franzl, T.; Resch-Genger, U. Appl. Spectrosc. 2010, 64, 733741. (23) Resch-Genger, U.; Pfeifer, D.; Monte, C.; Pilz, W.; Hoffmann, A.; Spieles, M.; Rurack, K.; Hollandt, J.; Taubert, D.; Schonenberger, B.; Nording, P. J. Fluoresc. 2005, 15, 315336. (24) Resch-Genger, U.; Hoffmann, K.; Pfeifer, D. In Reviews in Fluorescence; Geddes, C. D., Ed.; Springer Science Businesss Media, Inc.: New York, 2009; Vol. 4, pp 132. (25) DeRose, P. C.; Resch-Genger, U. Anal. Chem. 2010, 82, 2129 2133. (26) Resch-Genger, U.; deRose, P. Pure Appl. Chem. 2010, 82, 2315 2335. (27) Velapoldi, R. A.; Mielenz, K. D. NBS Spec. Publ. 1980, 26064, 1115. (28) Weissleder, R.; Pittet, M. J. Nature 2008, 452, 580589. (29) Pauli, J.; Vag, T.; Haag, R.; Spieles, M.; Wenzel, M.; Kaiser, W. A.; Resch-Genger, U.; Hilger, I. Eur. J. Med. Chem. 2009, 44, 3496 3503. (30) Kobayashi, H.; Ogawa, M.; Alford, R.; Choyke, P. L.; Urano, Y. Chem. Rev. 2010, 110, 26202640. (31) Escobedo, J. O.; Rusin, O.; Lim, S.; Strongin, R. M. Curr. Opin. Chem. Biol. 2010, 14, 6470. (32) Kiyose, K.; Kojima, H.; Nagano, T. Chem.Asian J. 2008, 3, 506515. (33) Pons, T.; Lequeux, N.; Mahler, B.; Sasnouski, S.; Fragola, A.; Dubertret, B. Chem. Mater. 2009, 21, 14181424. (34) Benson, R. C.; Kues, H. A. J. Chem. Eng. Data 1977, 22, 379 383. (35) Porres, L.; Holland, A.; Palsson, L.-O.; Monkman, A. P.; Kemp, C.; Beeby, A. J. Fluoresc. 2006, 16, 267273. (36) Gaigalas, A. K.; Wang, L. L. J. Res. Natl. Inst. Stand. Technol. 2008, 113, 1728. (37) Boyer, J. C.; van Veggel, F. Nanoscale 2010, 2, 14171419. (38) Heckmann, A.; Dummler, S.; Pauli, J.; Margraf, M.; Kohler, J.; Stich, D.; Lambert, C.; Fischer, I.; Resch-Genger, U. J. Phys. Chem. C 2009, 113, 2095820966. (39) Lesnyak, V.; Lutich, A.; Gaponik, N.; Grabolle, M.; Plotnikov, A.; Resch-Genger, U.; Eychmuller, A. J. Mater. Chem. 2009, 19, 9147 9152. (40) Wurth, C.; Grabolle, M.; Pauli, J.; Spieles, M.; Resch-Genger, U. Anal. Chem. 2011, 83, 34313439. (41) Velapoldi, R. A.; Mielenz, K. D. Standard Reference Materials: A Fluorescence Standard Reference Material: Quinine Sulfate Dihydrate; U.S. Department of Commerce, National Bureau of Standards, 1980. (42) Pfeifer, D.; Hoffmann, K.; Hoffmann, A.; Monte, C.; ReschGenger, U. J. Fluoresc. 2006, 16, 581587.
1352

Article

(43) Mathejczyk, J. E.; Pauli, J.; Dullin, C.; Napp, J.; Tietze, L.-F.; Kessler, H.; Resch-Genger, U.; Alves, F. Mol. Imaging 2011, 10, 469 480. (44) Mielenz, K. D.; Cehelnik, E. D.; McKenzie, R. L. J. Chem. Phys. 1976, 64, 370374. (45) Braslavsky, S. E. Pure Appl. Chem. 2007, 79, 293465. (46) Seybold, P. G.; Gouterman, M.; Callis, J. Photochem. Photobiol. 1969, 9, 229242. (47) Hollandt, J.; Taubert, R. D.; Seidel, J.; Resch-Genger, U.; GuggHelminger, A.; Pfeifer, D.; Monte, C.; Pilz, W. J. Fluoresc. 2005, 15, 301313. (48) Rurack, K.; Spieles, M. Anal. Chem. 2011, 83, 12321242. (49) deMello, J. C.; Wittmann, H. F.; Friend, R. H. Adv. Mater. 1997, 9, 230232. (50) Greenham, N. C.; Samuel, I. D. W.; Hayes, G. R.; Phillips, R. T.; Kessener, Y. A. R. R.; Moratti, S. C.; Holmes, A. B.; Friend, R. H. Chem. Phys. Lett. 1995, 241, 8996. (51) Bourhill, G.; Palsson, L. O.; Samuel, I. D. W.; Sage, I. C.; Oswald, I. D. H.; Duignan, J. P. Chem. Phys. Lett. 2001, 336, 234241. (52) Rohwer, L. S.; Martin, J. E. J. Lumin. 2005, 115, 7790. (53) Pauli, J.; Brehm, R.; Spieles, M.; Kaiser, W. A.; Hilger, I.; ReschGenger, U. J. Fluoresc. 2010, 20, 681693. (54) Achilefu, S. Technol. Cancer Res. Treat. 2004, 3, 393409. (55) Fernandez-Suarez, M.; Ting, A. Y. Nat. Rev. Mol. Cell Biol. 2008, 9, 929943. (56) Marme, N.; Knemeyer, J. P.; Sauer, M.; Wolfrum, J. Bioconjugate Chem. 2003, 14, 11331139. (57) Kovar, J. L.; Simpson, M. A.; Schutz-Geschwender, A.; Olive, D. M. Anal. Biochem. 2007, 367, 112. (58) Schobel, U.; Egelhaaf, H.-J.; Brecht, A.; Oelkrug, D.; Gauglitz, G. Bioconjugate Chem. 1999, 10, 11071114. (59) Pauli, J.; Grabolle, M.; Brehm, R.; Spieles, M.; Hamann, F. M.; Wenzel, M.; Hilger, I.; Resch-Genger, U. Bioconjugate Chem. 2011, 22, 12981308. (60) Buschmann, V.; Weston, K. D.; Sauer, M. Bioconjugate Chem. 2003, 14, 195204. (61) Gruber, H. J.; Hahn, C. D.; Kada, G.; Riener, C. K.; Harms, G. S.; Ahrer, W.; Dax, T. G.; Knaus, H. G. Bioconjugate Chem. 2000, 11, 696704. (62) Lisy, M. R.; Goermar, A.; Thomas, C.; Pauli, J.; Resch-Genger, U.; Kaiser, W. A.; Hilger, I. Radiology 2008, 247, 779787. (63) Russin, T. J.; Altinoglu, E. I.; Adair, J. H.; Eklund, P. C. J. Phys.: Condens. Matter 2010, 22, 334217. (64) Martini, M.; Montagna, M.; Ou, M.; Tillement, O.; Roux, S.; Perriat, P. J. Appl. Phys. 2009, 106, 094304.

dx.doi.org/10.1021/ac2021954 | Anal. Chem. 2012, 84, 13451352

También podría gustarte