Subscriber access provided by Colorado State University | Libraries
Article
Multi-mode Surface Functional Group Determination – Combining Steady-state and Time-resolved Fluorescence with XPS and Absorption Measurements for Absolute Quantification Tobias Fischer, Paul M. Dietrich, Wolfgang E.S. Unger, and Knut Rurack Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.5b03468 • Publication Date (Web): 22 Dec 2015 Downloaded from http://pubs.acs.org on December 31, 2015
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 9
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
Multi-mode Surface Functional Group Determination – Combining Steady-state and Time-resolved Fluorescence with XPS and Absorption Measurements for Absolute Quantification. Tobias Fischer, Paul M. Dietrich, Wolfgang E. S. Unger, Knut Rurack* Bundesanstalt für Materialforschung und –prüfung (BAM), Unter den Eichen 87, 12205 Berlin (Germany) ABSTRACT: The quantitative determination of surface functional groups is approached in a straightforward laboratorybased approach with high reliability. Application of a multi-mode BODIPY-type fluorescence, photometry and XPS label allows estimation of the labeling ratio, i.e., the ratio of functional groups carrying a label after reaction, from the elemental ratios of nitrogen and fluorine. The amount of label on the surface is quantified with UV/Vis spectrophotometry based on the molar absorption coefficient as molecular property. The investigated surfaces with varying density are prepared by co-deposition of 3-(aminopropyl)triethoxysilane (APTES) and cyanoethyltriethoxysilane (CETES) from vapor. These surfaces show high functional group densities that result in significant fluorescence quenching of surface-bound labels. Since alternative quantification of the label on the surface is available through XPS and photometry, a novel method to quantitatively account for fluorescence quenching based on fluorescence lifetime (τ) measurements is shown. Due to the complex distribution of τ on high-density surfaces, the stretched exponential (or Kohlrausch) function is required to determine representative mean lifetimes. The approach is extended to a commercial Rhodamine B isothiocyanate (RITC) label, clearly revealing the problems that arise from such charged labels used in conjunction with silane surfaces.
The surface functional group density is perhaps the most decisive parameter for the application of analytical platforms in assays that rely on the covalent anchoring of small-molecule or biomacromolecular moieties to a substrate such as slides, chips or particles.1,2 Whereas for the latter NMR, IR, conductometry or zeta-potential measurements can be employed for direct surface group analysis,3,4 macroscopically flat substrates are significantly more demanding to analyze due to their low surface-tovolume ratio and often require employment of a specific chemical label that can be interrogated selectively and sensitively.5-7 Analysis of such substrates is mainly performed with sophisticated surface analytical techniques like X-ray photoelectron spectroscopy (XPS)8-13 or time-offlight secondary ion mass spectrometry (ToF-SIMS).14,15 Obtaining quantitative information on surface functional groups, however, is rather challenging as these techniques mostly produce relative results. Because of their intrinsically high sensitivity and convenient operation, fluorometric methods in combination with the fluorescence labeling of surface species (FLOSS) are frequently employed in the quantification of surface functional groups,16-19 yet the reliability of this method is also limited: Quantification based on fluorescence relies on the assumption that all—or at least a known amount, defined by labeling conditions and reaction yield, of—functional groups are labeled. Such quantitative labeling, however, is only achievable in rare cases of well-isolated functional groups on surfaces and high-conversion reactions.20 This
difficulty can be circumvented by claiming only a quantification of accessible surface functional groups rather than the total amount.21 The meaning of this value is yet debatable if a small-molecule dye is used for quantification of functional groups on surfaces which later on are intended for assays involving the binding of larger entities like DNA strands or proteins to the substrate. The dynamic range of fluorometric methods is usually limited by effects of self-quenching at higher loadings, resulting in a deviation from linearity or even a reduction in the measured intensity.22-24 The latter is not easily identifiable with commonly employed intensity-based (steady-state) instrumentation and can lead to an underestimation of surface functional groups. To account for such false negative results, different optical techniques like time-resolved fluorescence and absorption measurements need to be employed. Given the comparatively low amount of dye which is bound to such a surface, spectrophotometry is usually considered not sensitive enough. Moreover, many of the labels used for spectrophotometric quantification absorb in the UV, limiting their applicability on glass substrates that cut off transmission until ca. 300 nm.25,26 As polyenic or polyaromatic systems, UV-absorbing dyes are also commonly characterized by rather low molar absorption coefficients,27 rendering the direct measurement of monolayers of such dyes on a surface difficult. These drawbacks can be accounted for by cleaving the label dyes again from the surface and dissolving them in a solvent of choice before the absorption measurement.
ACS Paragon Plus Environment
Analytical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
This indirect method yet suffers from inherent uncertainties connected to the increased number of steps.28-31 On the other hand, time-resolved fluorescence measurements were, to the best of our knowledge, not yet employed in the context of surface functional group quantification. Besides spectral and intensity information as provided by steady-state fluorescence, fluorescence lifetime measurements reveal significantly more detail about the homogeneity of the excited ensemble and its deactivation processes, i.e., allow to understand (self-)quenching phenomena better. For an assessment of the dependence of changes in fluorescence intensity as a function of the number of labels bound to a surface it is essential to prepare sets of surfaces with defined yet varying amounts of functional groups. Techniques for the fabrication of surfaces that differ only in the number of functional groups are, however, limited.20,32-34 In a recent contribution, we presented a simple and robust vapor deposition method for the preparation of surfaces with variable functional group densities.35 These surfaces were employed in the verification of the performance of a dual-mode label dye that enables the correlation of fluorescence intensities with XPS contents via the unique fluorine content of the dye. Although the mono-alkoxy silanes employed in that study, 3-(aminopropyl)diisopropylethoxy silane (APDIPES) and 3-(cyanopropyl)dimethylmethoxysilane (CPDMMS), provided smooth monolayers of comparable density, they are only seldom used in actual (bio)chemical applications of functional slides.13 By far the most commonly employed silane for amino surfaces is 3-(aminopropyl)triethoxysilane (APTES) that, due to its three alkoxy residues, is known for its challenging binding chemistry and the formation of multilayer-like structures.21,36-41 To extend our approach of adjusting the functional surface group density through the use of a functional and an inert silane in a co-deposition process to industrially more relevant surfaces, we combined here APTES with 2cyanoethyltriethoxysilane (CETES) as the inert component (Scheme 1). Aiming at the development of rapid and versatile methods to quantify surface functional groups, we relied again on our dual-mode label 1 (Scheme 1), the results of which can be directly linked to traceable methods as shown recently.35 Here we will show that, in contrast to the well-defined monolayers of the APDIPES/CPDMMS system, the fluorescence intensity alone is a misleading parameter when dye 1 is used with the multilayer-forming APTES/CETES system, because significant fluorescence quenching occurs at higher labeling ratios. However, we will also show how to a certain extent the quenching can be accounted for by fluorescence lifetime measurements. In addition, the feasibility of UV/Vis-spectrophotometry will be shown and a new method presented in which the labeling ratio as determined by XPS is directly linked to absorption for quantification of surface amino groups on slides. Finally, the suitability of our approach for commercial surface labels will be discussed and the problems associated with the use of charged dyes pointed out.
Page 2 of 9
Scheme 1. Surface reactions including structures of APTES, CETES and label 1
EXPERIMENTAL SECTION Chemicals and materials. All reagents were obtained from commercial suppliers and used without further purification unless otherwise stated. APTES (Acros) and CETES (Alfa Aesar) were handled only under protective atmosphere. Rhodamine B isothiocyanate (RITC) was received as a mix of isomers from Sigma-Aldrich. 1Pentylamine was from Alfa Aesar. Dye 1 was synthesized according to ref 23. Corning plain, pre-cleaned microscope glass slides (75 × 25 mm) were used as substrates. Plasma activation was performed with a Zepto Plasma system (Diener Electronic) and sonication in an Elmasonic P 60 H (Elma Hans Schmidbauer). Vapor deposition was performed in 100 mL DURAN® premium bottles with high-temperature PPS lids and silicone/PTFE sealing. Surface modification of slides. The silane solutions (10 % (v/v) organosilane in dry toluene) were prepared under inert atmosphere by stepwise dilution of an APTES with a CETES solution for the mixed silane solutions, resulting in APTES:CETES (v/v) ratios of 1:0, 1:1, 1:3, 1:7 and 0:1. 50 µL of each silane solution were transferred into small open vessels inside 100 mL cleaned glass bottles under an inert atmosphere. Two conditioned slides were placed inside the bottle which was closed with a PTFEsealed lid before transfer into an oven pre-heated to 150 °C. After 4 h at 150 °C, the hot bottles were opened under a stream of argon, directly subjected to vacuum (< 0.1 mbar) to remove the excess of toluene/silane vapor and left to cool down to room temperature (RT, 1.5 h). After
ACS Paragon Plus Environment
Page 3 of 9
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
removal from vacuum, the slides were washed, dried and stored at -18°C under protective atmosphere. For liquid phase dye reactions, one slide was split in halves and the first half reacted with 1, by placing it in a 0.2 mM solution of 1 in MeCN for 24 h at RT under stirring, followed by cleaning with acetone in a Soxhlet extractor for 24 h and rinsing/drying. The second half was placed in a 0.2 mM solution of RITC in 40 mM phosphate buffer for 4 h at RT under stirring, followed by cleaning with acetone in a Soxhlet extractor for 24 h. Contact angle measurements. Static contact angle (CA) measurements were performed on an OCA 15Pro (Data Physics Instruments) under a nitrogen atmosphere. Drop shapes were analyzed with SCA20 software (Data Physics). All samples were dried in vacuo prior to measurement. At least ten drops (2 µL) of MilliQ water were measured for the static contact angle. Fluorescence scanning. All slides were scanned with 488 and 532 nm excitation (λex) and a green emission filter using a Molecular Devices Axon GenePix 4300A fluorescence scanner at 20 µm resolution and 16-bits-perpixel signal collection. The images were analyzed with GenePix Pro 7.1 software (Molecular Devices). The slides were scanned prior to and after reaction with the label. Absorption measurements. Absorption spectra were recorded on an Analytik Jena Specord 210 Plus spectrophotometer. The slides were scanned by recording spectra on ten different spots (approx. 1 × 15 mm). A cleaned and otherwise untreated slide was used as reference. The spectra were averaged and any background remaining from differences in scattering distortion was subtracted manually using the background correction from Origin 9.1pro. Solution spectra were obtained in 10 mm Hellma quartz cuvettes using spectroscopic-grade solvents. Time-resolved fluorescence measurements. Fluorescence lifetimes were determined with a unique customized laser impulse fluorometer with picosecond time resolution described elsewhere.42 The samples were measured in front-face configuration with a 30° angle of incidence; angle of 60° between surface normal and observation axis. The fluorescence lifetime profiles were analyzed with the software package DAS 6 (Horiba Scientific). X-ray photoelectron spectroscopy. XPS measurements were carried out with a Kratos Analytical AXIS Ultra DLD photoelectron spectrometer. XPS spectra were recorded using monochromatized Al Kα excitation at pass energies of 80 eV for survey and 20 eV for high-resolution core-level spectra. The charge neutralizer was used. The electron emission angle was 0° (with respect to the surface normal) and the source-to-analyzer angle was 60°. The samples were thoroughly cleaned before XPS analysis to remove carbonaceous contaminations.43 High resolution Si 2p, C 1s, N 1s, O 1s and F 1s core-level spectra were analyzed with CasaXPS (Casa Software). For curve fitting of the core-level spectra, a Gaussian/Lorentzian product function peak shape model (G/L=30) was used in combination with a Shirley background.
RESULTS AND DISCUSSION Characterization of silane surfaces. The bare silane films were analyzed by means of water contact angle (CA) measurements. CAs were found to span from 73±1° for the pure amino silane surface to 56±2° for the pure cyano silane surface. In agreement with our previous work on monoalkoxysilane-derived model surfaces,35 the contact angle decreases with reduced content of APTES in the deposited vapor. Accordingly, we attempted to calculate the ratio of the two silane components using the Cassie equation (Eq. 1) for a simple two component model.44,45 Eq. 1
cos = cos + (1 − ) cos
Here, is the contact angle of the surface containing both components, the contact angle of the surface purely consisting of component one, in our case APTES, the relative amount of component one and the contact angle of component two, here CETES. Table 1. XPS contents of relevant elements (in at%) and Cassie ratios ( ) for mixed APTES:CETES layers %APTES
C 1s
N 1s
O 1s
Si 2p
100
26.2
4.2
42.4
27.2
1.00±0.02
50
24.3
5.1
42.4
28.2
0.45±0.02
25
22.3
4.8
43.9
28.9
0.43±0.04
12.5
18.2
4.2
47.4
30.2
0.24±0.02
0
9.5
1.7
56.1
32.7
0.00±0.03
Although the determined ratios (Table 1) show the expected trend of decreasing amino ratio, a straightforward application of Cassie’s Law remains doubtful. One major requirement is that both components mix without differences in structure or silane density compared to the pure surfaces. To elucidate if this requirement is met, we investigated the elemental composition of the prepared surfaces with XPS. Careful examination of the data obtained shows that the surfaces containing (various amounts of) APTES exhibit a comparable content of nitrogen (4-5 at%), accompanied by a slight decrease in carbon content, whereas the pure CETES-derived surface shows a significantly lower carbon (~10 vs 20-25 at%) and nitrogen content (~2 vs. 4-5 at%), indicating that the layer thickness is significantly lower in case of CETES. In combination with the higher relative contents measured for the substrate elements (Si, O), it is reasonable to assume that the structure of the pure CETES layer deviates significantly from that of the APTES-containing layers, especially in terms of the substrate contribution to the measured CA, and Cassie’s equation cannot be reliably employed. Only minor amounts of ca. 10% of amino silane in the total silane content appear sufficient to catalyze condensation reactions leading to a thicker layer than in the absence of amino silane.34,37–39,46–48 It should be noted that for vapor deposition of APTES, the more rigorously the inert atmosphere is controlled during the process the thinner and more homogeneous is the resulting amino silane layer.23
ACS Paragon Plus Environment
Analytical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Characterization of surfaces labeled with 1. After reaction of the samples with dye 1 according to our previously described procedure,23,49 the fluorescence intensities were recorded with a fluorescence scanner. Figure 1 reveals that the fluorescence intensity increases slightly from the lowest applied amino content (12.5 %) to 25 %, followed by a decrease for 50% and 100%; the pure CETES slide does not show noticeable emission. Since the XPS data of the silane layer without the dye (Table 1) do not suggest an increase in amino groups on the surface with a reduced fraction of APTES in the vapor deposition process, we tentatively ascribe the negative correlation to significant fluorescence quenching at higher dye concentrations, similar to our earlier findings.23 Table 2. XPS contents of relevant elements (in at%), steady-state fluorescence intensities (F), fluorescence lifetimes (τ) and distribution factors (β) of samples containing label 1; % = % APTES %
C 1s
F 1s
N 1s
O 1s
Si 2p
F /a.u.
τ /ns (β)
100
27.3
1.3
4.1
41.9
25.4
17.4
0.34 (0.52)
50
27.1
0.6
4.9
41.3
26.1
22.5
0.79 (0.59)
25
21.5
0.3
4.0
46.0
28.1
24.4
1.19 (0.64)
12.5
19.5
0.2
3.8
46.8
29.6
22.2
1.70 (0.69)
0
11.2
0.0
1.6
55.3
31.8
0.7
-
Figure 1. Relative contents of organic fluorine (circles) and fluorescence intensities without (black squares) and with (blue squares) lifetime correction for mixed silane substrates labeled with 1 (λex = 488 nm; λem = 550-600 nm)
To analyze the actual amount of bound dye, XPS spectra of the samples were measured and the content of organic fluorine was determined (Table 2). As shown in Figure 1, these contents correlate well with the applied APTES ratio. It appears as if, in contrast to our previous findings on APDIPES/CPDMMS, the silane ratio on the substrate reflects the silane ratio used in the deposition process. However, the amino/cyano ratios that showed the strongest deviation between the applied and the resulting ratio (0.4-6% relative amino silane content) in the previous work were not considered here. Additionally, none of the present surfaces can safely be considered as
Page 4 of 9
monolayer-like structure, as even little amounts of water will lead to self-condensation of the trialkoxysilanes, especially in the presence of the catalyzing amino group. The deviation between the XPS and fluorescence data clearly suggests that an increase in amino content entails enhanced fluorescence quenching. A known method to account for concentration-dependent self-quenching is its verification by fluorescence lifetime measurements.24,50 Excited-state interactions of dyes with surface amino groups (like photoinduced electron transfer, PET) are unlikely as no indication of such a process was found in our previous work.35 The main processes leading to selfquenching of dyes attached to surfaces are dimer formation and homo-energy transfer.51,52 However, whereas homo-energy transfer involves an excited species, dimer formation is a ground-state process resulting in a nonfluorescent species, the dimer; this quenching process is effectively static and cannot be traced by changes in fluorescence lifetime. In solution, collisional self-quenching also has to be taken into account, yet such interactions are not expected on surfaces since the dyes are covalently bound to the surface and therefore restricted in motion. Most factors diminishing the fluorescence intensity can be accounted for in a simple manner by relating the intensity in the presence ( ) and the absence ( ) of quenching to the lifetime in the presence ( ) and the absence ( ) of quenching.50 ⁄ = ⁄ Eq. 2 Here, and are the fluorescence intensity and lifetime measured for the actual sample. can be determined for the sample with the lowest content of dye. Although this might not necessarily be the “true” unquenched value of an unperturbed dye molecule in an ideally diluted state, the defined sample with the least amount of anchored dye can serve as a reasonable reference point. In general, for fluorophores bound to a surface, distributed lifetimes can be expected rather than strictly mono-exponential decays, simply because the microenvironment around each and every fluorophore cannot be considered as identical or homogeneous. A suitable way to analyze such data thus invokes the stretched exponential (or Kohlrausch) function.24,53,54 (Unfortunately, the formation of non-fluorescent dimers cannot be determined using this method, vide supra.52) () = + ( ⁄ )
Eq. 3
This equation is commonly used for approximating lifetime distributions on surfaces, as it rather conveniently provides a lifetime and factor ! describing the distribution. For details on the distribution provided by ! see ref. 54,55. Because the possible single processes contributing to the non-exponential decay such as dimer formation, homo-energy transfer or interaction with surface groups/ adsorbed water cannot be disentangled or specified for our present system with reasonable effort, we allowed ! to vary and did not fix it to a particular model. Figure 1 compares the measured fluorescence intensity and the lifetime-corrected intensity with the applied amino ratio and the relative content of organic fluorine. It is obvious
ACS Paragon Plus Environment
Page 5 of 9
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
that the relative signal increase for the region of lower amino ratio shows a good overlap between the corrected fluorescence intensity and the XPS content. Only in the 100 % APTES case, for which τ is significantly reduced by a factor of five (Table 2), the correction is insufficient. A tentative explanation for the latter includes the appearance of (non-fluorescent) dimers at such high labeling densities that leads to a reduction of the overall fluorescence intensity through static quenching without affecting τ. To allow for an unambiguous interpretation of the results another optical method is required to link the contents of fluorine to the emission behavior of the fluorophoric system. That is why we carefully examined the possible use of conventional absorption spectroscopy.
Figure 2. Absorption spectra of slides labeled with 1 (APTES content: 100% (black), 50% (red), 25% (blue), 12.5% (green), 0% (cyan)), the small band around ~ 510 nm is a measurement artefact resulting from the lamp spectrum and a change of detector amplification in the instrument and is accounted for in the measurement uncertainty
As Figure 2 shows, especially at higher labeling densities, the spectra can be readily measured with acceptable noise. It is noteworthy that in contrast to the popular UVabsorbing label nitrobenzaldehyde, 1 shows absorption in the visible range, where the transparency of glass and silane films is considerably higher and the measurement thus less distorted. Especially
