Simultaneous Determination of Stable Carbon, Oxygen, and Hydrogen

Dec 11, 2014 - Department of Geography, College of Science, Swansea University, Swansea, SA2 8PP, United Kingdom. ‡School of Geography, Earth and En...
7 downloads 15 Views 1MB Size
Technical Note pubs.acs.org/ac

Simultaneous Determination of Stable Carbon, Oxygen, and Hydrogen Isotopes in Cellulose N.J. Loader,*,† F.A. Street-Perrott,† T.J. Daley,‡ P.D.M. Hughes,§ A. Kimak,∥ T. Levanič,⊥ G. Mallon,# D. Mauquoy,¶ I. Robertson,† T.P. Roland,∇,§ S. van Bellen,¶ M.M. Ziehmer,∥ and M. Leuenberger∥ †

Department of Geography, College of Science, Swansea University, Swansea, SA2 8PP, United Kingdom School of Geography, Earth and Environmental Sciences, University of Plymouth, Plymouth, PL4 8AA, United Kingdom § School of Geography, University of Southampton, Southampton, SO17 1BJ, United Kingdom ∥ Climate and Environmental Physics, Physics Institute and Oeschger Centre for Climate Change Research University of Bern, 3012 Bern, Switzerland ⊥ Slovenian Forestry Institute, Vecna pot 2, SI-1000 Ljubljana, Slovenia # Department of Geography, University of Sheffield, Sheffield, S10 2TN, United Kingdom ¶ School of Geosciences, University of Aberdeen, Aberdeen, AB24 3UF, United Kingdom ∇ Department of Geography, College of Life & Environmental Sciences, University of Exeter, Exeter, EX4 4RJ, United Kingdom ‡

S Supporting Information *

ABSTRACT: A technological development is described through which the stable carbon-, oxygen-, and nonexchangeable hydrogen-isotopic ratios (δ13C, δ18O, δ2H) are determined on a single carbohydrate (cellulose) sample with precision equivalent to conventional techniques (δ13C 0.15‰, δ18O 0.30‰, δ2H 3.0‰). This triple-isotope approach offers significant new research opportunities, most notably in physiology and medicine, isotope biogeochemistry, forensic science, and palaeoclimatology, when isotopic analysis of a common sample is desirable or when sample material is limited.

A

Many plant-based physiological, palaeoclimatological, archeological, and forensic studies are constrained by a lack of sample material or experience difficulty in measuring all three isotopes. This has often forced researchers to “trade-off” between the range of isotopic analyses possible and the need for replication. One consequence is that, while models describing carbon- and oxygen-isotope fractionation in plants are well constrained,1−3,11 equivalent models explaining hydrogen-isotope fractionation remain relatively uncertain and would benefit from additional data coupled, where possible, with other isotopic measures.12 Hence, it would be both scientifically and economically advantageous to be able to determine all three isotope ratios simultaneously. A diverse range of analytical techniques has been developed for the separate analysis of carbon-, oxygen-, and nonexchangeable hydrogen-isotopic ratios in cellulose.13 However, the chemical composition of cellulose, like many simple carbohydrates, makes sample preparation for simultaneous analysis of

s the most abundant biopolymer, cellulose represents an important substrate for isotopic investigation. Variations in the relative abundance of the stable carbon, oxygen, and nonexchangeable hydrogen isotopes from which the cellulose is made are inextricably linked to the global carbon and water cycles. As a consequence, the isotopic study of these elements, both as labeled compounds and at natural-abundance levels, has contributed significantly to the characterization of plantphysiological processes and the study of past and contemporary environmental changes.1−5 This environmental and physiological linkage between carbon-, oxygen-, and hydrogen-isotope fractionation means that, when viewed together, these three indicators have the capacity to provide far greater insights into plant physiology and the functioning of the Earth system than might be possible if considered in isolation. An improved understanding of this three-way relationship is of increasing importance in the evaluation and development of dynamic vegetation and water-isotope models, as the measurement of stable isotopes in natural archives (e.g., tree rings, peat macrofossils, pollen) can provide a unique perspective on isotopic variability across spatial and temporal scales not accessible from direct observations or remote sensing.4−10 © 2014 American Chemical Society

Received: July 11, 2014 Accepted: November 20, 2014 Published: December 11, 2014 376

dx.doi.org/10.1021/ac502557x | Anal. Chem. 2015, 87, 376−380

Analytical Chemistry

Technical Note

steam is passed in a flow of helium. The steam is produced from water of known isotopic composition, which is supplied to a heater block at 110 °C by a peristaltic pump operating at 2 μL min−1. The sample is allowed to equilibrate for 600 s, after which it is dropped into a low-blank autosampler purged with helium at 150 mL min−1 immediately prior to dropping into a Flash HT elemental analyzer (Thermo GmbH, Germany), where it is pyrolyzed at >1400 °C. Pyrolysis is contained within a ceramic tube lined with a glassy-carbon insert part-filled with glassy carbon chips. Residue from the silver capsules is collected in an octagonal graphite crucible (OEA Laboratories, Ltd., UK.). The tube is flushed with (Grade N4.6) helium (150 mL· min−1). The resultant products of cellulose pyrolysis, H2 and CO, are resolved gas-chromatographically via a 5 Å molecular sieve maintained at 50 °C, prior to mass-spectrometric analysis using a Thermo Delta V mass spectrometer (Thermo GmbH, Germany). Carbon resulting from pyrolysis of cellulose (C6H10O5) remains in the graphite crucible and has been shown to have a negligible effect upon the δ13C signal or the carbon-isotopic ratio of subsequent analyses.17,18,20 Mass-spectrometric analysis comprises determination of three reference-gas pulses (H2, grade N5.5, BOC Ltd., UK) followed by determination of the isotope ratio of the hydrogen gas, which is first to elute (at ca. 170 s). Once the hydrogen peak has passed, a peak-jump is performed (at ca. 230 s) to enable detection and measurement of the carbon monoxide peak (at ca. 350 s). The run concludes with determination of three reference-gas pulses (CO, grade N4.7, BOC Ltd., UK) (Figure S1, Supporting Information). Samples run sequentially, overlapping temporally during equilibration, pyrolysis, and isotope analysis, resulting in an effective run time of 630 s. Multiple analyses can be performed in batches of ca. 200 before deashing of the furnace crucible is required. Isotope data are expressed as ‰ deviations from the VPDB and VSMOW standards. Data may be corrected using a singleor dual(multiple)-point correction. In this paper, to facilitate comparison with Hafner et al.,21 we present the single-point corrected data, verified by additional standard materials analyzed within the run for which the nonexchangeable hydrogen-isotope composition is known. To accommodate different calibration schemes and temperature dependence of δ18O, the oxygen data developed using low-temperature (1090 °C) pyrolysis were adjusted by +0.6‰ (based on measurement of IAEA-C3 cellulose) and the equilibrated hydrogen isotope data by −3.0‰ (based on measurement of IAEA-CH7).16−19,21 During routine analyses, a total of ten IAEA-CH7 polyethylene standards (nonexchangeable hydrogen only) are run at the start and end of the analysis batch to test for drift in the reference gas and evaluate precision over time. For every 19 samples, four Sigma and one IAEA-C3 holocellulose are analyzed to assess machine performance and provide a means for linear-drift correction in the equilibrated samples if required. Normally, the system is stable within a run (standard precision within limits of ±0.15‰ δ13C, ± 0.30‰ δ18O, and ±3.0‰ δ2H) and no drift correction is required. However, it is important to conduct these periodic checks to detect changes in equilibration conditions or instability during a run. We obtain the best results when samples are weighed to within a fixed size range (typically 0.30−0.35 mg), as this provides consistent peak heights and precludes the need for size-related (linearity) corrections. For the polyethylene standard, samples of 0.125− 0.15 mg are prepared to yield equivalent peak heights to the

all three elements challenging, as the products of thermal degradation are subject to the water−gas shift reactions and Boudouard equilibrium. These two processes vary the nature and relative proportions of the reaction products, thereby making isolation and quantitative conversion of the cellulose to a single analyte problematic. However, if the sample is pyrolyzed at high temperature (>1400 °C), it is possible to convert cellulose to carbon, carbon monoxide, and hydrogen. A further consideration when analyzing cellulose for hydrogen isotopes is the proportion of exchangeable and nonexchangeable hydrogen. Three out of ten hydrogen atoms in the cellulose polymer are hydroxyl-bound and capable of isotopic exchange. This component must be adequately quantified or removed prior to hydrogen-isotope determination if a measure of the hydrogen (water) used by the plant during cellulose synthesis is to be obtained. Previous successful attempts to circumvent this problem have included derivatization of the cellulose to cellulose nitrate14 and isotopic equilibration of cellulose with water of known isotopic composition.15 Preparation of cellulose nitrate is not always desirable or feasible; it requires large quantities (several milligrams) of cellulose, and the unstable nature of the end product makes long-term storage or transport a potential safety hazard. Recent analytical developments in sample preparation and continuous-flow isotope analysis now permit online equilibration of exchangeable hydrogen. Filot et al.16 describe a system whereby cellulose samples are weighed into silver capsules and individually equilibrated at 110 °C immediately prior to online analysis of the hydrogen-isotope ratio. Other studies have convincingly demonstrated that both oxygen- and carbon-isotope ratios of cellulose can be reliably measured simultaneously on carbon monoxide from pyrolysis of cellulose.17,18 A significant methodological advance is described which integrates technologies to enable simultaneous determination of carbon-, oxygen-, and nonexchangeable hydrogen-isotope ratios from a single cellulose sample. The instrument system comprises an automated equilibration unit interfaced with an elemental analyzer and an isotope-ratio mass spectrometer. The performance of the instrumentation was evaluated through a comparison of reference cellulose materials and cellulose prepared from absolutely dated tree rings measured using standard alternative methods. The first test sequentially analyzed repeated blocks of 5 standard or in-house reference materials, including the two International Atomic Energy Agency (IAEA) standards (IAEA-CH7 Polyethylene Foil (nonexchangeable H only) and IAEA-C3 (holocellulose)), an in-house Sigma cellulose standard, and two samples of cellulose prepared from Juniperus procera Hochst. ex Endl. (African Juniper) and a single ring of Pinus longaeva Bailey (Great Basin Bristlecone Pine), respectively. The isotopic range of these test materials enabled measurement precision to be compared against that typically reported using standard methods,19 as well as detection and quantification of any significant isotopic memory effects or carry-over between groups of standards.



METHOD Dry sample material (0.30−0.35 mg) is weighed into a silver foil capsule (OEA Laboratories Ltd., UK), lightly crimped, and loaded into the autosampler of a fully automated equilibration unit based on the manual design described by Filot et al.16 The sample is dropped from the autosampler tray into an equilibration chamber heated to 110 °C, through which 377

dx.doi.org/10.1021/ac502557x | Anal. Chem. 2015, 87, 376−380

Analytical Chemistry

Technical Note

cellulose samples. Once weighed and packed, samples are freeze-dried overnight (−45 °C at 20 mbar) prior to analysis. The resulting carbon and oxygen data are evaluated using a single-point correction using Sigma α-cellulose (−23.89‰ δ13C, 27.3‰ δ18O, −99.16‰ δ2Hnonexchangeable) and a secondary verification is conducted using IAEA-C3 holocellulose standards (−24.55‰ δ13C, 32.6‰ δ18O, −67.43‰ δ2Hnonexchangeable). Typically, hydrogen isotopes are qualitychecked and, if necessary, corrected first against the IAEA-CH7 standard (−100.6‰ δ2Hnonexchangeable) and then for equilibration using the values of the nonexchangeable cellulose and the isotopic composition of the equilibration water by mass balance, assuming full equilibration. Back-calculation using standard materials for which the nonexchangeable hydrogen content has been previously assessed provides a further verification of the data set. Some workers using a single-point correction have identified a difference in the variance of carbonisotope data derived from carbon monoxide compared to values obtained through combustion of the cellulose.17−19 This does not constitute a problem when data sets are developed exclusively from either combustion or pyrolysis methods, but when data from both methods are to be combined, we recommend an assessment of the carbon data to determine whether or not a secondary scaling of one data set to the other is required prior to their combination. Using a dual- or multipoint correction based upon standards covering the full range of sample variability will dispense with the need for such secondary scaling. This suppressed variance may reflect a position-dependent isotopic proclivity during cellulose synthesis or an isotopomer effect in the formation of carbon monoxide and deposition of carbon during cellulose pyrolysis.20,22,23 Carbon-isotopic data presented here have been variance-scaled to the target (combustion data set). Safety Considerations. The online method presented here has low associated risk since no extremely harmful chemicals are involved in the chemical processes associated with this apparatus. Sources of physical risk include the moving parts and elevated temperature of the equilibrator (110 °C) and the elemental analyzer (1400 °C) and of the mass spectrometer. Compressed gases, vacuum (mass spectrometer), and electric supply to the apparatus constitute an additional working risk. Special attention is drawn to the carrier (He) and reference gases (H2, CO). The carbon monoxide reference gas should be stored securely in a well-ventilated area, preferably away from the location of work in a vented store. The laboratory should contain appropriate CO and low-oxygen detectors to alert users in the event of leakage from a cylinder or buildup of CO during operation of the pyrolysis unit, which should be vented externally if possible. All compressed gases should be treated with care, used, and stored according to supplier safety guidelines.



Table 1. Mean and Precision Data for Standard Materials Run for the Triple-Isotope Testa mean IAEA-CH7 Sigma IAEA-C3 Pinus longaeva Juniperus procera Sigma IAEA-C3 Pinus longaeva Juniperus procera Sigma IAEA-C3 Pinus longaeva Juniperus procera

Hydrogen −100.60 −99.16 −62.94 −99.25 −2.89 Carbon −23.89 −24.45 −20.52 −22.39 Oxygen 27.30 33.11 37.06 38.24

standard deviation

N

1.87 2.52 1.46 1.18 3.14

10 10 10 9b 10

0.07 0.10 0.08 0.06

10 10 9b 10

0.15 0.19 0.10 0.11

10 10 9b 10

a

No oxygen- or carbon-isotope data are presented for IAEA-CH7 as carbon monoxide is not produced during pyrolysis of polyethylene. bN = 9 for Pinus longaeva cellulose as one sample did not fall properly from the autosampler.

The second experiment evaluated system performance in a typical plant-physiological or palaeoclimatological application using a series of 50 triplicate measurements of α-cellulose prepared from a single Larix decidua Mill. tree growing in the Kamniško-Savinjske Alpe, Slovenia. This material was prepared from an absolutely dated tree ring series and had previously been analyzed (single measurements) by standard methods (combustion (δ13C), pyrolysis (δ18O), and nitration with combustion (δ2H)).21 Figure 1 confirms the efficacy of the triple-isotope approach and the capability of the system to reproduce results generated using conventional methods. Common interannual trends were preserved throughout the series, with mean annual precision calculated from triplicate measurement of each of the 50 rings analyzed, well within acceptable limits (±0.07‰ δ13C, ± 0.20‰ δ18O, and ±1.28‰ δ2H). No significant differences were found between the results of the standard methods and the new triple-isotope method described here (2-tailed paired t test, p = 0.05, N = 50).



CONCLUSIONS The development of a triple-isotope method represents a significant improvement over existing methods. The system performs comparably with standard methods in terms of analytical precision and, importantly, requires less than a third of the sample material needed for single determinations using standard techniques. Analysis time is comparable with that of a single isotope determination by continuous flow. As the samples do not require nitration prior to analysis, significant efficiencies arise in the development of a multiple-isotope time series. A further benefit of running three isotopic measurements on a single common sample is the increased statistical confidence in the mean value resulting from replicate analyses in cases where the sample material is limited. If the tripleisotope approach is routinely adopted, we anticipate that the added information provided by the hydrogen-isotope data from α-cellulose samples will greatly facilitate modeling of hydrogenisotope fractionation in plants and lead to a more complete

RESULTS

Values of analytical precision for the reference materials analyzed across the experiment (σn−1, n = 10, run in two blocks of 5) all fell within the levels of precision typically reported for the three isotope ratios determined using standard methods (±0.15‰ δ13C, ± 0.30‰ δ18O, and ±3.0‰ δ2H) with the exception of the Juniperus hydrogen, which narrowly exceeded this threshold (±3.14‰). The Juniperus sample was prepared from multiple blocks of tree rings and probably exhibits a greater level of sample heterogeneity (Table 1). 378

dx.doi.org/10.1021/ac502557x | Anal. Chem. 2015, 87, 376−380

Analytical Chemistry



Technical Note

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This study was supported by UK NERC Grants NE/G019673/ 1, NE/I022809/1, NE/I022981/1, NE/I022833/1, and NE/ I023104/1, iTREE (SNSF Sinergia project 136295), and the Climate Change Consortium of Wales (C3W). We thank our colleagues in the EU “Millennium” Project (Roderick Bale, Rhodri Griffiths, Högne Jungner, Danny McCarroll, Hanspeter Moret, Peter Nyfeler, Eloni Sonninen, and Giles Young) for support and Dan Charman for “gentle encouragement”.



Figure 1. Comparison of the isotope series developed from α-cellulose prepared from a single Larix decidua Mill. tree growing in the Julian Alps, Slovenia, using the triple-isotope method reported here (colored lines) and the results obtained on the same sample material by Leuenberger and Filot20 using standard methods (black circles). (Top panel) δ13C, blue circles; (Middle panel) δ18O, green circles; (Lower panel) δ2H, red circles. No significant differences were found between the results of the standard methods and the new triple-isotope method described here (2-tailed paired t-test, p < 0.05).

understanding of carbon-, oxygen-, and hydrogen-isotopic variability in the Earth system.



REFERENCES

(1) Farquhar, G. D.; O’Leary, M. H.; Berry, J. A. Aust. J. Plant Physiol. 1982, 9, 121−137. (2) Hill, S. A.; Waterhouse, J. S.; Field, E. M.; Switsur, V. R.; Ap Rees, T. Plant, Cell Environ. 1995, 18, 931−936. (3) Waterhouse, J. S.; Switsur, V. R.; Barker, A. C.; Carter, A. H. C.; Hemming, D. L.; Loader, N. J.; Robertson, I. Quat. Sci. Rev. 2004, 23, 803−810. (4) Loader, N. J.; Young, G. H. F.; Grudd, H.; McCarroll, D. Quat. Sci. Rev. 2013, 62, 97−113. (5) Treydte, K. S.; Frank, D.; Esper, J.; Andreu, L.; Bednarz, Z.; Berninger, F.; Boettger, T.; D’Alessandro, C. M.; Etien, N.; Filot, M.; Grabner, M.; Guillemin, M. T.; Gutierrez, E.; Haupt, M.; Helle, G.; Hilasvuori, E.; Jungner, H.; Kalela-Brundin, M.; Krapiec, M.; Leuenberger, M.; Loader, N. J.; Masson-Delmotte, V.; Pazdur, A.; Pawelczyk, S.; Pierre, M.; Planells, O.; Pukiene, R.; Reynolds-Henne, C. E.; Rinne, K. T.; Saracino, A.; Saurer, M.; Sonninen, E.; Stievenard, M.; Switsur, V. R.; Szczepanek, M.; Szychowska-Krapiec, E.; Todaro, L.; Waterhouse, J. S.; Weigl, M.; Schleser, G. H. Geophys. Res. Lett. 2007, 34, L24302. (6) Saurer, M.; Spahni, R.; Frank, D. C.; Joos, F.; Leuenberger, M.; Loader, N. J.; McCarroll, D.; Gagen, M.; Poulter, B.; Siegwolf, R. T. W.; Andreu-Hayles, L.; Boettger, T.; Dorado Liñań , I.; Fairchild, I. J.; Friedrich, M.; Gutierrez, E.; Haupt, M.; Hilasvuori, E.; Heinrich, I.; Helle, G.; Grudd, H.; Jalkanen, R.; Levanič, T.; Linderholm, H. W.; Robertson, I.; Sonninen, E.; Treydte, K.; Waterhouse, J. S.; Woodley, E. J.; Wynn, P. M.; Young, G. H. F. Global Change Biology 2014, 20 (12), 3700−3712 DOI: 10.1111/gcb.12717. (7) Frank, D. C.; Poulter, B.; Saurer, M.; Esper, J.; Huntingford, C.; Helle, G.; Treydte, K.; Zimmermann, N. E.; Schleser, G. H.; Ahlström, A.; Ciais, P.; Friedlingstein, P.; Levis, S.; Lomas, M.; Sitch, S.; Viovy, N.; Andreu-Hayles, L.; Bednarz, Z.; Berninger, F.; Boettger, T.; D’Alessandro, C. M.; Daux, V.; Filot, M.; Grabner, M.; Gutierrez, E.; Haupt, M.; Hilasvuori, E.; Jungner, H.; Kalela-Brundin, M.; Krapiec, M.; Leuenberger, M.; Loader, N. J.; Marah, H.; Masson-Delmotte, J.; Pazdur, A.; Pawelczyk, S.; Pierre, M.; Planells, O.; Pukiene, R.; Reynolds-Henne, C. E.; Rinne, K. T.; Saracino, A.; Sonninen, E.; Stievenard, M.; Switsur, V. R.; Szczepanek, M.; Szychowska-Krapiec, E.; Todaro, L.; Waterhouse, J. S.; Weigl, M. Nat. Clim. Change, submitted for publication. (8) Daley, T. J.; Thomas, E. R.; Holmes, J. A.; Street-Perrott, F. A.; Chapman, M. R.; Tindall, J. C.; Valdes, P. J.; Loader, N. J.; Marshall, J. D.; Wolff, E. W.; Hopley, P. J.; Atkinson, T.; Barber, K. E.; Fisher, E. H.; Robertson, I.; Hughes, P. D. M.; Roberts, C. N. Global Planet. Change 2011, 79, 288−302. (9) Sturm, K.; Hoffmann, G.; Langmann, B.; Stichler, W. Hydrol. Processes 2006, 19, DOI: 10.1002/hyp.5979. (10) Tindall, J. C.; Valdes, P. J.; Sime, L. C. J. Geophys. Res. 2009, 114, D04111. (11) Roden, J. S.; Lin, G.; Ehleringer, J. R. Geochim. Cosmochim. Acta 2000, 64, 21−35.

ASSOCIATED CONTENT

S Supporting Information *

Information outlining the sample materials used for evaluation of the method and the experimental setup. This material is available free of charge via the Internet at http://pubs.acs.org. 379

dx.doi.org/10.1021/ac502557x | Anal. Chem. 2015, 87, 376−380

Analytical Chemistry

Technical Note

(12) Yakir, D. Plant, Cell Environ. 1992, 15, 1005−1020. (13) McCarroll, D.; Loader, N. J. Quat. Sci. Rev. 2004, 23, 771−881. (14) Ramesh, R.; Bhattacharya, S. K.; Gopalan, K. Earth Planet. Sci. Lett. 1986, 79, 66−74. (15) Schimmellman, A. Anal. Chem. 1991, 63, 2456−2459. (16) Filot, M. S.; Leuenberger, M.; Pazdur, A.; Boettger, T. Rapid Commun. Mass Spectrom. 2006, 20, 3337−3344. (17) Young, G. H. F.; Loader, N. J.; McCarroll, D. Palaeogeogr., Palaeoclimatol., Palaeoecol. 2011, 300, 23−28. (18) Woodley, E. J.; Loader, N. J.; McCarroll, D.; Young, G. H. F.; Robertson, I.; Heaton, T. H. E.; Gagen, M. H.; Warham, J. O. Rapid Commun. Mass Spectrom. 2012, 26, 109−114. (19) Boettger, T.; Haupt, M.; Knoller, K.; Weise, S. M.; Waterhouse, J. S.; Rinne, K. T.; Loader, N. J.; Sonninen, E.; Jungner, H.; MassonDelmotte, V.; Stievenard, M.; Guillemin, M. T.; Pierre, M.; Pazdur, A.; Leuenberger, M.; Filot, M.; Saurer, M.; Reynolds, C. E.; Helle, G.; Schleser, G. H. Anal. Chem. 2007, 79, 4603−4612. (20) Leuenberger, M. C.; Filot, M. S. Rapid Commun. Mass Spectrom. 2007, 21, 1587−1598. (21) Hafner, P.; Robertson, I.; McCarroll, D.; Loader, N. J.; Gagen, M.; Bale, R. J.; Jungner, H.; Sonninen, E.; Hilasvuori, E.; Levanič, T. Trees 2011, 25, 1141−1154. (22) Loader, N. J.; Buhay, W. M. Rapid Commun. Mass Spectrom. 1999, 13, 1828−1832. (23) Waterhouse, J. S.; Cheng, S. Y.; Juchelka, D.; Loader, N. J.; McCarroll, D.; Switsur, V. R.; Gautam, L. Geochim. Cosmochim. Acta 2013, 112, 178−191.

380

dx.doi.org/10.1021/ac502557x | Anal. Chem. 2015, 87, 376−380