Universal HPLC detector for hydrophilic organic compounds by means

10 hours ago - Most quantifications are achieved by comparison of the signals obtained with the sample to those from a standard. Thus, the purity and ...
9 downloads 4 Views 998KB Size
Subscriber access provided by UNIVERSITY OF LEEDS

Article

Universal HPLC detector for hydrophilic organic compounds by means of total organic carbon detection Shin-Ichi Ohira, Kyosuke Kaneda, Toru Matsuzaki, Shuta Mori, Masanobu Mori, and Kei Toda Anal. Chem., Just Accepted Manuscript • Publication Date (Web): 07 May 2018 Downloaded from http://pubs.acs.org on May 7, 2018

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 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 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.

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 16 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

Universal HPLC detector for hydrophilic organic compounds by means of total organic carbon detection

Shin-Ichi Ohira,†* Kyosuke Kaneda,† Toru Matsuzaki,† Shuta Mori,† Masanobu Mori‡ and Kei Toda†



Department of Chemistry, Kumamoto University, 2-39-1 Kurokami, Kumamoto, 860-8555, Japan.



Faculty of Science and Technology, Kochi University, 2-5-1 Akebono-cho, 780-8520, Japan.

Corresponding Author *E-mail: [email protected]. Fax: +81-96-342-3384

1

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

ABSTRACT:

Most quantifications are achieved by comparison of the signals obtained with the sample to those from a standard. Thus, the purity and stability of the standard are key in chemical analysis. Furthermore, if an analyte standard cannot be obtained, quantification cannot be achieved, even if the chemical structures are identified by a qualification method (e.g. high-resolution mass spectrometry). Herein, we describe a universal and analyte standard-free detector for aqueous-eluent-based high-performance liquid chromatography. This universal carbon detector (UCD) was developed based on total organic carbon detection. Separated analytes were oxidized in-line and converted to carbon dioxide (CO2). Generated CO2 was transferred into the gas phase and collected into ultrapure water, which was followed by conductivity detection. The system can be applied as a HPLC detector that does not use an organic solvent as an eluent. The system can be calibrated with a primary standard of sodium bicarbonate for organic compounds. The universality and quantification were evaluated with organic compounds, including organic acids, sugars, and amino acids. Furthermore, the system was successfully applied to evaluation of the purity of formaldehyde in formalin solution, and determination of sugars in juices. The results show the universal carbon detector has good universality and can quantify many kinds of organic compounds with a single standard such as sodium bicarbonate.

Keywords: high-performance liquid chromatography, universal detector, analyte standard-free analysis, organic compounds

2

ACS Paragon Plus Environment

Page 2 of 16

Page 3 of 16 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

Q

uantitative determination plays a key role in many scientific fields, and instruments for chemical analysis are highly

developed. Quantification generally depends on comparison of the signals obtained for the sample with those of analyte standards with known compositions and concentrations. There are more than 133 million compounds registered on the Chemical Abstracts Service1, but only 8 million of these are commercially available2. In other words, analyte standards are not readily available for many compounds. Furthermore, for reliable measurements, analytical standards need to have high purity and stability. Purity is also important not only for analytical standard compounds but also in a variety of chemistry research areas. Currently, purity can be determined by titration3, melting point4, elemental analysis, differential scanning calorimetry5, and nuclear magnetic resonance spectroscopy (NMR)6. Quantitative NMR is used in a number of applications and is typically used with chromatographic separation to ensure that contaminants do not remain in an NMR sample7. If a high-performance liquid chromatography (HPLC) detector could be used for quantification without analyte standards, quantification of the compounds, for which standards are unavailable or even unknown, could be achieved with suitable qualification methods such as high-resolution mass spectrometry. Qualification of unknown compounds using high-resolution mass spectrometry techniques, such as time-of-flight8 and Orbitrap9, is well established. These mass spectrometric methods can obtain exact analyte masses with resolving powers of > 20,000 full width at half maximum, and compounds can be identified with computation of the exact atomic masses10. These mass spectrometry usages are growing, and these qualification methods are also important in omics analysis11. The quantification of reaction intermediates12 and isomers such as anomer13 are also important in chemical synthesis.. In most cases, it is difficult to obtain well-certified analytical standards for each compound. HPLC is a well-known analytical method that provides effective separation and is used in many areas of chemistry and biology research. Detector selection for HPLC is important for achieving selective and accurate analysis. The most widely used detector is the ultraviolet (UV) detector, which traditionally uses a wavelength of 254 nm generated from a low-pressure mercury lamp. A UV detector can detect many kinds of organic and inorganic compounds with relatively high sensitivity. However, the sensitivity of UV detection depends strongly on the physical and chemical properties of the analytes, and sugars, for example, cannot be detected with sufficient sensitivity. The absorption of light by chemical compounds is affected by the solution composition, pH, and temperature14. Based on the chemical and physical properties of the analytes, many kinds of detectors such as fluorometric, electrochemical, and chemiluminescence detectors have also been used with HPLC. The most suitable detector is selected based on many kinds of aspects, such as sensitivity, selectivity and costs. The “universal detector” concept has been developed from two main approaches. The first of these uses the counts of particulates number formed from analytes by evaporation of the eluent. The formed particulates are detected by an evaporative light scattering detector (ELSD) 17,18,19

(CAD)

15,16

or electrically as charged aerosol using a Corona charged aerosol detector

. These methods are well established, and the instruments are commercially available. However, the responses

are strongly affected by the eluent composition because a high volume fraction of organic solvent leads to high transport efficiencies on the nebulizer19. The effect of eluent composition has been modified using an additional counter flow to maintain a constant volume fraction of an eluent and obtain an almost equal response for the analyte19. The universality of Corona CAD response and response prediction have been discussed18,19. However, some problems remain including the difficulty of detecting volatile compounds, and variation in sensitivity for different compounds (relative standard deviation of around 10% or less). The other approach uses a flame ionization detector (FID) typically used for gas chromatography, which produces a 3

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

Page 4 of 16

response that is linear to the carbon numbers for analytes with similar chemical structures. Split/heating 20 and nebulizer/spray chamber21,22 interfaces have been used between the HPLC and FID. These methods have been used to determine the ethanol content in drinks20 and the carboxylic and amino acid contents22. Universal detection can also be achieved with conversion of the analyte to CO2 for the determination of organic compounds. More than three decades ago, Gloor et al. reported the combination of HPLC and dissolved organic carbon detection for a detection23. The column eluent was introduced into a heated chamber for catalytic conversion of organic carbon into CO2, which was detected by an infrared absorption-based CO2 analyzer. The detector was reported to be universal but quantification with a single calibration curve was not shown for any analyte23. Recently, use of this approach has been reported for determination of the fractions in natural organic matter. Huber et al. reported on an organic carbon and nitrogen detector with chromatographic separation to evaluate the fractions of compounds in natural organic matter24. This method could not quantify specific compounds, and only gave the amounts of compounds in each fraction, which depended on size-exclusion separation. The aim of the present study was to develop a universal and analyte standard-free detector for HPLC with aqueous (non-organic solvent contained) eluent. This was performed with an integrated system for chemical conversion of organic compounds to CO2 and as well as the detection of CO2. The developed universal carbon detector (UCD) was applied successfully to the standardization of formaldehyde, which is unstable and vaporizes from aqueous solution, and sugar, an important metabolite that is difficult to detect with a conventional UV detector. Even though simultaneous qualification with other instruments is required, many kinds of organic compounds can be quantified with the present system using sodium bicarbonate as a calibration standard.

MATERIALS AND METHODS Materials. All commercially available reagents were of reagent grade, except for two special grade formalin solutions used as samples. All reagents were obtained from www.nacalai.co.jp or www.wako-chem.co.jp. Ultrapure water (UPW) for use in all experiments was produced using a Simplicity UV water purification system (www.merckmillipore.com). HPLC-UCD system. A schematic diagram of the detection system integrated with a conventional chromatograph named HPLC-UCD is shown in Figure 1. The chromatograph consisted of a dual plunger pump (KP-21-01, www.flom.co.jp), injector (Cheminert® C2-1006, www.vici.com), separation column, and detector (UV-2070 or RI-2031, www.jasco.co.jp). The UV and RI detector was used to obtain reference results for comparison with the UCD. Sodium persulfate was used as an oxidizer and mixed with the separation column effluent after the UV or RI detector. An acid solution was mixed with the oxidizing solution before it was mixed with the eluent. Next, the mixture was passed through a heated reactor (90 °C, stainless steel tube, 0.5 mm i.d., length 106 cm) and a UV reactor consisting of a quartz glass coil (1.5 mm i.d. × 3.2 mm o.d. × 565 mm length, www.theglassplant.com) filled with glass beads (ø 0.1 mm). A mercury lamp (81-1025-01, ozone generates, length 25.4 mm, www.bhkinc.com) was inserted into the quartz coil. The mixture was then passed through a nebulizer for effective transfer of generated CO2 into the gas phase. The gas phase was then passed through a gas diffusion scrubber, which collected CO2 gas molecules into a continuously flowing absorbing solution25. The gas and acceptor solution were flowed with counter-current flow for effective gas collection. The absorbing solution in the present study was UPW, which passed through an in-line purification column filled with mixed bed resin (Dowex Monosphere MR-450UPW, www.sigmaaldrich.com) before being introduced into the gas diffusion scrubber. The separation conditions (Table S1), nebulizer (Figure S1), and gas diffusion scrubber (Figure S2) details are given in the Supporting Information. 4

ACS Paragon Plus Environment

Page 5 of 16 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

Figure 1. Schematic diagram of the HPLC-UCD system. P1, P2, P3, P4, pumps; SC, separation column; UVD, ultraviolet (UV) detector; RI, refractive index detector; HR, heated reactor; UVR, UV reactor; N/GLS, nebulizer/gas liquid separator; UPW, ultrapure water; PC, purification column; GDS, gas diffusion scrubber; and CD, conductivity detector. Sample pretreatment. Various brands of prepackaged orange juice and apple juice were obtained from markets in Kumamoto, Japan and pretreated for the determination of sugars following the separation column manufacturer’s procedures26. For the orange juice, each sample was filtered with filter paper and then a 0.45-µm syringe filter. The effluent was introduced directly into a mini-column filled with a hydroxide form anion exchange resin, and then again filtered with the 0.45-µm syringe filter. The effluent was then introduced directly into the separation system for RI detection and further ×100 dilution for UCD. For the apple juice, a 5 g sample of each juice was placed in a beaker, followed by addition of 300 mL of UPW and neutralization with 10% (w/v) aqueous NaOH cooled with ice. The solution was ultrasonicated for 30 min, then filtered and diluted with 50 mL of UPW. The solution was further filtered through a 0.45-µm syringe filter and introduced directly into the separation system for RI detection and further ×100 dilution for UCD.

RESULTS AND DISCUSSION Selection of the carbon dioxide detection method. In the present study, universal carbon detection was achieved by oxidation of organic analytes to CO2, followed by vaporization of CO2 and continuous CO2 detection. Initially, the methods for continuous CO2 vaporization and sensitive detection were studied. Stable and quantitative vaporization was required for the UCD. Preliminary experiments suggested that a nebulizer provided the best transfer of generated CO2 into the gas phase. Effective nebulizing typically needs a high gas flow rate / liquid flow rate ratio27. In the present study, a micro-nebulizer (Figure S1), which required 10–50 mL/min of nebulizing gas for a solution flow of 0.80 mL/min, was developed. For CO2 detection, infrared (IR) spectroscopy, methanization followed by flame ionization detection (FID), and conductivity detection following CO2 collection into an absorbing solution were considered. Typically, gas phase IR detection requires a relatively large sample volume to replace a long path length cell with large cell volume. Furthermore, because of this large volume, the system cannot measure with high time resolution, which is required for use as a HPLC detector. Conversion of CO or CO2 into CH4 by reduced nickel catalysis under a stream of hydrogen gas followed by FID is 5

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

a highly sensitive method to determine CO228,29. However, the presence of water in the gas phase may poison the catalyst. Collection into an absorbing solution with conductivity detection has also been applied to continuously monitor CO2 concentrations30. Taking into consideration the deviation of the conductivity response of the solution in the collector, sufficient time resolution may be obtained31. However, the response will be noisy, and the sensitivity may not be suitable. In the present study, continuous and sensitive CO2 detection was achieved with gas diffusion scrubber collection followed by conductivity detection. A gas diffusion scrubber has been used for atmospheric gas analysis in earlier studies32,33. The gas diffusion scrubber collects gas molecules into a scrubbing solution via a gas permeable membrane. Several gas diffusion scrubber structures have been reviewed25. Briefly, in the present study, a porous membrane tube with a small diameter was used to contain the scrubbing solution and sample gas in its inner and outer compartments, respectively (Figure S2 in Supporting Information). The concentration of gas molecules on the surface of the porous membrane tube will be equal to zero if the tube and scrubbing solution can function as a perfect sink for the target gas molecules. This generates a concentration gradient, the gradient accelerates the gas molecules to reach the membrane surface, where they are collected from the gas to liquid phase. Counter-current flow was used to aid in effective gas collection. In our preliminary study, LiOH, Ba(OH)2, and UPW were tested as scrubbing solutions with conductivity detection. It was expected that an alkaline scrubbing solution would be more effective for CO2 collection, and produce a negative peak for the conductivity response. Lithium salt was selected based on its large hydrated radii, which result in low equivalent conductance. When CO2 was collected into an alkaline solution, alkaline metal and carbonate ions remained after CO2 collection. The negative response with CO2 is based on the conversion from hydroxide to carbonate ions in absorbing solution. The lithium ions produced a relatively larger signal to noise ratio than other alkaline metal ions. Barium salts were expected to produce a larger negative response than lithium salts because the formation of BaCO3 would reduce the conductivity to close to zero. However, alkaline scrubbing solutions showed high background responses and were not suitable for highly sensitive detection of CO2. With Ba(OH)2, the pores of the porous membranes were filled with formed BaCO3 and the collection efficiencies decreased over time. By contrast, UPW showed a stable baseline for the conductivity response even though the response was not linear over large concentration ranges. The additional purification column filled with mix-bed ion exchange resin placed just before the inlet of diffusion scrubber was effective, and the baseline noise decreased to 1/5 of what it was obtained without the column. The responses obtained with LiOH and UPW as scrubbing solutions for 10 mM NaHCO3 standard solution were compared (Figure S3 in Supporting Information), and the UPW gave a signal-to-noise ratio that was 20 times larger than LiOH. A highly humid gas is not suitable for gas diffusion scrubber collection because it condenses on the membrane tube and affects the absorption of CO2. In addition, the response time would dramatically decrease because of water condensation in the diffusion scrubber. The gas phase introduced into the gas diffusion scrubber can be saturated with water vapor by the nebulizer in the present system. Therefore, additional dry nitrogen gas flow was mixed with the effluent gas from the nebulizer to decrease and keep humidity constant. The parameters for the UCD system, such as the flow rates of the gas absorbing solution, additional nitrogen gas and nebulizer gas were further optimized on the basis of the sensitivity for NaHCO3 standard and peak half widths. The results are shown in Figures S4–S6. The optimized values were 0.2, 45, and 25 mL/min for the absorbing solution, additional nitrogen gas, and nebulizer gas flow rates, respectively. Universal detection of organic compounds. The oxidation procedure for the conversion of analyte organic compounds into CO2 is a key for universal carbon detection. Several preliminary experiments suggested that wet oxidation 6

ACS Paragon Plus Environment

Page 6 of 16

Page 7 of 16 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 with persulfate, heating (∆) and UV radiation made it possible to convert organic compounds to CO2 quantitatively. This reaction is widely used for total organic carbon (TOC) determination34. The newly developed UCD was initially evaluated with flow injection analysis. Aliquots (5 µL) of the standard solutions were injected into the carrier solution (3 mM HClO4 at 0.7 mL/min) and mixed with the oxidizer solution (80 mM K2S2O8 at 0.1 mL/min). The oxidation efficiency was evaluated with potassium hydrogen phthalate (KHP, eight carbon atoms), p-benzoquinone (p-BQ, six carbon atoms), and sucrose (12 carbon atoms). These compounds were used for the evaluation of the TOC analyzer using the United States Pharmacopeia (USP)35. They were selected because KHP, p-BQ, and sucrose are primary standards that form strong chemical bonds and have high carbon ratios. These results are shown in Figure 2. The concentration of carbonate, which was used as a primary standard, was linear up to approximately 50 mM with R2 = 0.997. The responses for all of the organic compounds increased linearly up to 15 mmol of carbon per liter (mmol C/L). Furthermore, the obtained peak areas for the compounds were as same as the peak area for carbonate. This suggested that the organic analytes were quantitatively oxidized into CO2 under the present conditions. The quantitatively convertible analyte concentration, which also depends on the carbon number of the analyte molecule, were on the order of several millimoles per liter. The range for the quantitative conversion can be expanded by changing the amounts of sample and oxidizer. In the present study, the sample volume and oxidizer concentration were fixed at 5 µL and 80 mM, respectively, while the oxidizer flow rate was with 1/7th of an eluent flow rate of. The dynamic ranges can be adjusted using these parameters, but the samples can be diluted if the concentrations are higher than the detectable range in the present study. The present system can detect 0.03 mM carbonate in a 5-µL sample, which corresponds to 1.8 ng of carbon. Commercial instruments for determination of TOC can detect 0.5 µg of carbon per liter in a 20-mL sample36. The sensitivity of the present system is similar to that for commercial TOC instruments. The universality of the present detection system was further studied by determination of carbonate, carboxylates, sugars, and amino acids. The results, which are shown as relative peak areas, are presented in Figure 3. The standard solutions were prepared on the basis of carbon concentrations. The relative standard deviations between compounds were 5.6, 3.3, 2.4% for 0.5, 1.0, 2.0 mmol C/L, respectively. The results clearly show that the method can be used for quantification of the tested compounds with carbonate as the standard compound. Nitrogen and sulfur contained in organic analytes such as amino acids are converted to nitrate and sulfate, respectively, by oxidation. With the present system, these did not greatly interfere with carbon detection because nitrate and sulfate were not transferred into the gas phase nor were they collected into the absorbing solution via the gas permeable membrane. In the present system, the response increased linearly with the carbon number in a molecule, that is, the sensitivity depended on the carbon number of the compound. The limits of detection based on a signal-to-noise ratio of three are given in Table S2 in the Supporting Information.

7

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

Figure 2. Calibration curves for carbonate and organic compounds. Experimental conditions: flow injection analysis; carrier, 3 mM HClO4 (0.7 mL/min); oxidizer, 80 mM K2S2O8 (0.1 mL/min); additional nitrogen gas flow rate, 45 mL/min; nebulizer gas flow rate, 25 mL/min; absorbing solution, UPW (0.3 mL/min); and sample injection volume, 5 µL. All of the analyses were performed in triplicate.

8

ACS Paragon Plus Environment

Page 8 of 16

Page 9 of 16 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

Figure 3. Universal responses to many kinds of organic compounds. Experimental conditions: flow injection analysis; and sample concentrations, 0.5, 1.0, and 2.0 mmol C/L. The other conditions were the same as in Figure 2. Analyte standard free HPLC detector. The UCD system has been applied as a HPLC detector. Many methods of HPLC have used organic solvents as eluent. Organic solvents eluent cannot be introduced to the present UCD because the organic solvents also generate CO2. However, many kinds of compounds such as amino acids, carboxylates, sugars can be separated with aqueous solution as eluents. Separation with aqueous eluents has also been widely studied from the viewpoint of environmental friendly chemistry37. Furthermore, 2D-HPLC38 will help to separate organic solvents in eluent and analytes before introducing to UCD. In the present study, aqueous eluent was used for evaluation of the performance of the present UCD as a HPLC detector. For the first performance test, linear carboxylates were determined with an ion exclusion separation column using 3 mM HClO4 as the eluent. For comparison purposes, a conventional UV detector (210 nm) was also connected in series. The chromatograms obtained with the UCD and UV detector for 5 mM mixtures are shown in Figure 4a. These detectors were connected in series as the UV-UCD. The peaks with UCD was broader than UV detector responses. The half peak widths obtained with UCD and UV detectors were 0.93±0.16 and 0.55±0.22 min, respectively. One of the reasons for this was the dispersion during the reactions for oxidation. The mixture of column effluent (0.7 mL/min) and oxidizer solutions (0.1 mL/min) was passing through the heated reactor (0.21 mL, 0. 25 mm i.d.) and quartz coil (1.5 mm i.d., volume 1.0 mL). Filling the coil with glass beads improved the dispersion to 2/3 of peak width while maintaining oxidation efficiency but the dispersion remained. The other reason is the time requirement procedure for CO2 desorption from and absorption into aqueous phases. The generated CO2 was desorbed from solution to gas phases in the nebulizer. The solution was heated previously for the reaction and the solutions were well nebulized to form small droplets. These are highly effective for releasing CO2 into the N2 gas flow. However, this equilibrium procedure is not 9

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

Page 10 of 16

effectively fast. In addition, the CO2 collection via porous membrane was not fast because hydration and dehydration of CO2 are not fast processes39. However, sufficient separation was still achieved as shown in the chromatogram (Figure 4a). The calibration curves are shown in Figure 4b and 4c for the UV detector and UCD, respectively. The responses for carboxylates with carbon numbers between 1 and 5 were successfully determined with the UCD. The relationship between the carbon number (nC) and peak area for 5 mM carboxylates can be expressed as follows: UCD peak area, V· s = (1.45 ± 0.02) × nC + (0.08 ± 0.09), R2 = 0.999

(1)

The peak area increased linearly with the carbon number. These results suggest that the response, or sensitivity, increases with increasing carbon number. The detection sensitivity could possibly be improved by derivatizing the analytes into compounds with larger carbon numbers. The calibration curves for UV detection did not agree with each other because the molar absorptivity depends on the chemical structures and matrix conditions such as solution pH. In contrast, the UCD responses agreed well when the molar concentration of carbons in the analyte were plotted on the x-axis. The responses agreed with the carbonate responses. Validation of the responses was less than < 10%. The calibration equation for carbonate was as follows: UCD peak area, V· s = (1.48 ± 0.03) × Ccarbonate + (0.20 ± 0.18), R2 = 0.999

(2)

The average slope for linear carboxylates, with carbon numbers from 1 to 5 and concentrations ranging from 0.5 to 2 mM, was 1.49 ± 0.09. This agrees with the slope of the calibration curve of carbonate. These results are clear proof that HPLC-UCD can quantify many compounds with a single calibration curve obtained with carbonate.

10

ACS Paragon Plus Environment

Page 11 of 16 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

Figure 4. HPLC–UV and -UCD for determination of carboxylates. The separation was conducted with ion exclusion chromatography, and all other conditions were the same as those in Figure 2. Standardization of HCHO in formalin. The purity of a compound is important for many kinds of chemistry research and chemical industries. Chromatography can separate major and minor compounds. However, typically, chromatography 11

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

Page 12 of 16

cannot be used for quantification without standard compounds. To determine the purities of compounds, a standard with a well-defined concentration is required. Purity determinations are conducted with methods such as NMR. The importance of purity evaluation and application of quantitative NMR (qNMR) for the evaluation are widely known, especially in the areas of pharmaceuticals, natural materials, and food analysis40. In the present study, the HCHO concentration in formaldehyde was determined without a HCHO standard using the developed HPLC-UCD system. Formalin is a ~37% HCHO aqueous solution that contains methanol to prevent polymerization of HCHO41. The HCHO concentration in this formalin is not stable because the solubility of HCHO gas in water depends on temperature, pressure, and the Cannizzaro reaction, in which two HCHO molecules react to generate methanol and formate. The concentrations of HCHO in water and air are monitored daily for environmental monitoring at work places and residences. This monitoring uses instrumental analysis and requires a standard solution, which is standardized by iodide titration42. The present HPLC-UCD system can simply quantify HCHO after separation with methanol and without the need for preparation of a HCHO standard. The obtained chromatogram and results are summarized in Figure 5. Separation of HCHO and methanol was achieved with an ion exclusion chromatography column (RSpak KC-811, www.shodex.com) and purified water as the eluent. Because quantitative oxidation was achieved under acidic conditions, the additional acid (24 mM HClO4, flow rate 0.1 mL/min) was mixed with the oxidizer. In-line mixing was compared with the use of the acid and oxidizer pre-mixed solution, and we found that the stability of the oxidizer decreased dramatically under the mixture with acid43. Thus, the acid was added into the oxidizer in-line just before mixing with column effluent. If the primary standard, NaHCO3, were eluted separately with HCHO and methanol, the sample could be analyzed along with the standard. In the present study, NaHCO3 and methanol were eluted with the same retention time. Thus, the system was calibrated separately with the NaHCO3 solution. The results obtained with HPLC-UCD for HCHO were compared with the results from iodide titration and certified values for ultrapure reagents (samples 1 and 2 only). The results for methanol were compared with only the certified values. The ratio of the average HCHO concentrations from HPLC-UCD, the iodide titration, and certified values was 1.00 ± 0.015:0.985 ± 0.006:1.00. The ratio of the methanol results from HPLC-UCD and the certified values was 1.01 ± 0.024:1.00. The results from HPLC-UCD agreed with those obtained by the other methods. Standardization of the reagents was successfully achieved with HPLC-UCD with the primary standard (NaHCO3).

12

ACS Paragon Plus Environment

Page 13 of 16 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 Figure 5. Purity evaluation of HCHO in a formalin solution. Experimental conditions: separation mode, ion exclusion chromatography column; eluent, UPW; and acid, 24 mM HClO4 at 0.1 mL/min. All other conditions were the same as those in Figure 2.

Determination of sugar in juices. Sugar is an important energy source and metabolite. However, sugars cannot be detected with conventional UV-visible detectors because its photoabsorption is poor or absent. Conventional methods for sugar detection use a refractive index (RI) detector after HPLC separation44. Slight changes in the RI of the eluent are caused by analyte elution, temperature, and eluent composition. The poor sensitivity of the RI method is not suitable. Recently developed universal detectors, such as the ELSD and Corona CAD, have achieved highly-sensitive sugar detection. However, these methods also detect inorganic compounds and desalting of samples is strongly recommended for accurate analysis. In the present study, the sugars in various brands of juices were separated by ligand exchange chromatography and then detected with the UCD. Four sugars (sorbitol, fructose, glucose, and sucrose) were separated with UPW as the eluent and quantified with the UCD. The calibration curves for these compounds agreed with the calibration curve obtained with the carbonate standard up to 12 mmol C/L. Orange and apple juice samples were pretreated and diluted (100 times), and then analyzed with the present UCD with the carbonate standard (UCD(CO32−)) for quantification. The conventional RI detector was used for comparison purposes. The detection limits obtained with the RI detector were 57, 75, 81, and 190 mg/L for sucrose, glucose, fructose, and sorbitol, respectively. In contrast, the present UCD showed 4.5, 4.6, 4.9 and 6.0 mg/L for sucrose, glucose, fructose, and sorbitol, respectively. Sensitivity that was one order better was obtained with present UCD. The RI results agreed with those from UCD(CO32−) with slopes of 1.02 (R2 = 0.998) (Figure 6). These results demonstrated the ability of UCD to detect organic analytes with a primary standard compound.

13

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

Page 14 of 16

Figure 6. Results for sugars in juices. RI, refractive index detector; and UCD(CO32−), the UCD calibrated with carbonate for all analytes.

CONCLUSIONS In summary, the newly developed UCD provides universal detection for organic compounds with aqueous eluent based HPLC separation. The UCD response depends on the carbon number of the analyte. The system can be calibrated using a single pure and stable primary standard for subsequent analysis of organic compounds. The system worked effectively for the quantification of the unstable compounds such as HCHO.

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

Notes The authors declare no competing financial interest. ACKNOWLEDGMENTS This work was supported by a Grant-in-Aid for Young Scientists (B) from the Japan Society for the Promotion of Science (JSPS) KAKENHI (Grant No. 23750089).

14

ACS Paragon Plus Environment

Page 15 of 16 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

References (1) Chemical abstract, www.cas.org Last accessed October 19, 2017. (2) http://www.sigmaaldrich.com/chemistry/chemistry-services/aldrich-market-select.html (3) Gong, N.; Liu, S.; Xu, W.; Shi Y.; Du, G.; Lu, Y. Anal. Methods, 2013, 5, 784–788. (4) Aston, J.G.; Fink, H.L., Tooke, J.W., Cines, M.R. Anal. Chem., 1947, 19, 218–221. (5) Plato, C; Glasgow, A.R. Anal. Chem., 1969, 41, 330–336. (6) Pauli, G.F.; Chen, S.-N.; Simmler, C.; Lankin, D.C.; Gödecke, T.; Jaki, B.U.; Friesen, J.B.; McAlpine, J.B.; Napolitano, J.G. J. Med. Chem. 2014, 57, 9220–9231. (7) The Japanese Pharmacopoeia 17th ed., Japanese ministry of health, labor, and welfare, p.137 (2016). (8) Guale, F.; Shahreza, S.; Walterscheid, J.P.; Chen, H.-H.; Arndt, C.; Kelly, A.T.; Mozayani, A. J. Anal. Toxic., 2013, 37, 17–24. (9) Krauss, M.; Singer, H.; Hollender, J. Anal. Bioanal. Chem., 2010, 397, 943–951. (10) He, M; Wu, H.; Nie, J.; Yan, P.; Yang, Z.-Y.; Pei, R. J. Pharm. Biomed. Anal., 2017, 146, 37–47. (11) Aebersold, R.; Mann, M. Nature, 2016, 357, 347–355. (12) Oonishi, J.; Watanabe, H.; Okuda, J. Bunseki Kagaku, 1996, 45, 589–594. (13) Baker, J.O.; Himmel, M.E. J. Chromatogr. A, 1986, 357, 161–181. (14) Christian, G.D.; Dasgupta, P.K.; Schug, K.A. Analytical Chemistry 7th ed.; Wiley: New York, 2013, Chapter 16. (15) Charlesworth, J.M. Anal. Chem., 1978, 50, 1414–1420. (16) Lucena, R.; Cárdenas, S.; Valcárcel, M. Anal. Bioanal. Chem., 2007, 388, 1663–1672. (17) Gamache, P.H.; McCarthy, R.S.; Freeto S.M.; Asa, D.J.; Woodcock, M.J.; Laws, K.; Cole, R.O. LCGC Europe, 2005, 18, 345–354. (18) Hutchinson, J.P.; Li, J.; Farrell, W.; Groeber, E.; Szucs, R.; Dicinoski, G.; Haddad, P.R. J. Chromatogr. A, 2010, 1217, 7418–7427. (19) Górecki, T; Lynen, F.; Szucs, R.; Sandra, P. Anal. Chem., 2006, 78, 3186–3192. (20) Yarita, T.; Nakajima, R.; Otsuka, S.; Ihara, T.; Takatsu, A.; Shibukawa, M. J. Chromatogr. A, 2002, 976, 387–391. (21) Young, E.; Smith, R.M.; Sharp, B.L.; Bone, J.R. J. Chromatogr. A, 2012, 1236, 16–20. (22) Young, E.; Smith, R.M.; Sharp, B.L.; Bone, J.R. J. Chromatogr. A, 2012, 1236, 21–27. (23) Gloor, R.; Leidner, H. Anal. Chem., 1979, 51, 645–647. (24) Huber, S.A.; Balz, A.; Abert, M.; Pronk, W. Water Res., 2011, 45, 879–885. (25) Toda, K. Anal. Sci., 2004, 20, 19–27. (26) Shodex application notes for SC1011 separation column: www.shodex.com (27) Murillo, M.; Carrión, N.; Colmenares, J.; Romero, J.; Alvarado, G.; Ríos, R.; Angulo, F. Quim. Nova, 2008, 31, 1315– 1318. (28) Porter, K.; Volman, D.H., Anal. Chem., 1962, 34, 748–749. (29) Welss, R.F., J. Chromatogr. Sci., 1981, 19, 611–616. (30) Toda, K.; Li, J.; Dasgupta, P.K., Anal. Chem., 2006, 78, 7284–7291. (31) Ohira, S.; Wanigasekara, E.; Rudkevich, D.M.; Dasgupta, P.K. Talanta, 2009, 77, 1814–1820. (32) Toda, K.; Ohira, S.; Tanaka, T.; Nishimura, T. Environ. Sci. Technol., 2004, 38, 1529–1536. (33) Ohira, S.; Azad, M.A.K.; Kuraoka, R.; Tanaka, T.; Mori, K.; Toda, K. Bunseki Kagaku, 2006, 55, 109–115. (34) Urbansky, E.T. J. Environ. Monit., 2001, 3, 102–112. (35) United State Pharmacopeia Chapter total organic carbon (36) Shimadzu Corp., Specifications of TOC-V Series TOC Analyzers, http://www.shimadzu.com/an/toc/lab/tocv-spec.html. (37) Smith, R.M.; Burgess, R.J. Anal. Comm., 1996, 33, 327–329. (38) Stoll, D.R.; Carr, P.W. Anal. Chem., 2017, 89, 519–531. (39) Stumm, W.; Morgan, J.J. Aquatic Chemistry, 2nd ed.; Willey: New York, 1981; pp. 210–214. (40) Pauli, G.F.; Chen, S.-N.; Simmler, C.; Lankin, D.C.; Gödecke, T.; Jaki, B.U.; Friesen, J.B.; McAlpine, J.B.; Napolitano, J.G. J. Med. Chem., 2014, 57, 9220–9231. (41) Bobkova, L.P.; Korsakov, V.S.; Romanov, L.M.; Yenikolopyan, N.S. Polymer Sci. U.S.S.R., 1963, 5, 1785–1789. (42) JIS K 8872:2008, “Formaldehyde solution (Reagent)”. (43) Pottasium persulfate material safety data sheet, datasheets.scbt.com/sc-203362.pdf. (44) Kelebek, H.; Selli, S.; Canbas, A.; Cabaroglu, T. Microchem. J., 2009, 91, 187–192.

15

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

For TOC Only

16

ACS Paragon Plus Environment

Page 16 of 16