A Direct Bicarbonate Detection Method Based on a Near ... - MDPI

sensors Article

A Direct Bicarbonate Detection Method Based on a Near-Concentric Cavity-Enhanced Raman Spectroscopy System Dewang Yang, Jinjia Guo *, Chunhao Liu, Qingsheng Liu and Ronger Zheng College of Information Science & Engineering, Ocean University of China, Qingdao 266000, China; [email protected] (D.Y.); [email protected] (C.L.); [email protected] (Q.L.); [email protected] (R.Z.) * Correspondence: [email protected]; Tel.: +86-532-6678-1211 Received: 18 October 2017; Accepted: 29 November 2017; Published: 1 December 2017

Abstract: Raman spectroscopy has great potential as a tool in a variety of hydrothermal science applications. However, its low sensitivity has limited its use in common sea areas. In this paper, we develop a near-concentric cavity-enhanced Raman spectroscopy system to directly detect bicarbonate in seawater for the first time. With the aid of this near-concentric cavity-enhanced Raman spectroscopy system, a significant enhancement in HCO3 − detection has been achieved. The obtained limit of detection (LOD) is determined to be 0.37 mmol/L—much lower than the typical concentration of HCO3 − in seawater. By introducing a specially developed data processing scheme, the weak HCO3 − signal is extracted from the strong sulfate signal background, hence a quantitative analysis with R2 of 0.951 is made possible. Based on the spectra taken from deep sea seawater sampling, the concentration of HCO3 − has been determined to be 1.91 mmol/L, with a relative error of 2.1% from the reported value (1.95 mmol/L) of seawater in the ocean. It is expected that the near-concentric cavity-enhanced Raman spectroscopy system could be developed and used for in-situ ocean observation in the near future. Keywords: laser Raman spectroscopy; bicarbonate; direct detection; ocean application

1. Introduction The carbon cycle in the ocean is a dynamic component of the global carbon budget, but the diverse sources and carbon sinks as well as their complex interactions in the ocean remain poorly understood [1,2]. The dissolved inorganic carbon (DIC) in the ocean is the sum of HCO3 − , CO3 2− , H2 CO3 and CO2 [3]. The concentration of DIC is about 2.30 mmol/L, with HCO3 (about 1.95 mmol/L) making up the majority of it [4]. Therefore, bicarbonate is a key factor in carbon cycle research. The modern methods for oceanic carbonate system measurement are mainly involved in total carbon and CO2 partial pressure (pCO2 ) investigations [5,6]. However, the direct detection of HCO3 − and CO3 2− in seawater it is still a huge challenge due to their low concentrations. Raman spectroscopy is potentially capable of overcoming this challenge on the condition of a significant ensitivity enhancement [7]. In 2004, scientists reported the first attempt to detect HCO3 − in water solutions directly by using Raman spectroscopy [8], with a limit of detection (LOD) of 7.5 mmol/L after an integration time of 20 min. They had also tried to enhance the HCO3 − signal with an introduced liquid core waveguide [9], eventually obtaining an amplification ratio of 7.8× in the peak intensity. All those efforts still do not make the direct Raman detection of HCO3 − in seawater practical due to the fact its concentration is as low as 2 mmol/L. The strong background signal resulting from SO4 2− , which has a concentration of ~28 mmol/L in seawater [10], makes direct detection of HCO3 − even more difficult. With a similar waveguide-based enhancement mechanism, the multi-pass cavity concept has been Sensors 2017, 17, 2784; doi:10.3390/s17122784

www.mdpi.com/journal/sensors

Sensors 2017, 17, 2784

2 of 6

Sensors 2017, 17, 2784

2 of 6

presented as an effective approach in gas Raman detection with desirable enhancement [11–14]. Takingsuch suchan an approach approachin in direct direct seawater seawater Raman Raman detection detection is is the the motivation motivation for for the the work work presented Taking presented in this article. A near-concentric cavity-enhanced Raman spectroscopy system was specially built for in this article. A near-concentric cavity-enhanced Raman spectroscopy system was specially built −−Raman detection in seawater liquid sample detection. With the aid of this system, direct HCO 3 for liquid sample detection. With the aid of this system, direct HCO3 Raman detection in seawater becomes readily readily feasible feasible as as expected. expected. becomes 2. Experiment Setup

As shown shownininFigure Figure 1, the near-concentric cavity-enhanced spectroscopy is As 1, the near-concentric cavity-enhanced RamanRaman spectroscopy system issystem adapted adapted on the basis of a formerly reported design [14]. In order to detect liquid samples, a sample on the basis of a formerly reported design [14]. In order to detect liquid samples, a sample cell of cell ofsilica fusedwith silica mm × 10 × 40 mmsize inner size and wall 1 mm wall thickness used and fused 10with mm 10 × 10 mm × mm 40 mm inner and 1 mm thickness is used is and placed placed at the of center of the cavity. this system, a diode-pumped, frequency-doubled Nd:YAG CW at the center the cavity. In this In system, a diode-pumped, frequency-doubled Nd:YAG CW laser lasernm) (532with nm) with a power ofmW 300 mW is used as Raman the Raman excitation laser source. polarization (532 a power of 300 is used as the excitation laser source. TheThe polarization of ◦ through of the laser beam is rotated throughpassing passinga ahalf halfwave waveplate plate(HWP), (HWP),and andthen thenthe the laser laser beam beam the laser beam is rotated byby 9090° couple of lens lens (a (a planoconvex planoconvex lens with a focal focal is compressed and focused into the chamber by a couple length of of 100 100 mm mm and and aa planoconcave planoconcavelens lenswith withaafocal focallength lengthof of− −75 length 75 mm). The near concentric cavity is composed of two identical spherical mirrors M1 with 25.4 mm diameter and 25 mm focal length. length. The reflectivity of these two mirrors is over 99% from 500 to 700 nm. These two mirrors are spaced The reflectivity of these two mirrors is over 99% from 500 to 700 nm. These two mirrors are spaced face ◦ to create face to at face at a distance of mm, 104.2and mm, the mirror is clockwise rotated about 0.04° to to face a distance of 104.2 theand mirror M1-2 isM1-2 clockwise rotated about 0.04 thecreate near the near concentric cavity reflection mode.ofBecause of the strong absorption of laser water, the laser beam concentric cavity reflection mode. Because the strong absorption of water, the beam number in number 18, which is number less than the sample cavity of a gas cell. The the cavityinisthe 18, cavity which is less than the inthe the number cavity ofin a gas cell. The sample scattering signal scattering signaloffrom theis center of cavity is collected doublet by an achromatic (f = 30 mm) from the center cavity collected by an achromatic lens L1 (f =doublet 30 mm)lens withL1a diameter of withmm. a diameter 25.4 mm. mirror M2 on (f =the 12.5 mm), placed on the side beyond the 25.4 A mirrorofM2 (f = 12.5Amm), placed opposite side beyond theopposite sample chamber, is used sample chamber, is used to enhance of thethe collection of the lens L2 = 30 mm) to enhance the collection efficiency signal. efficiency Another lens L2signal. (f = 30Another mm) is used to (f couple the is used signal to couple signal into thebundle delivery bundle μm ×with 100aμm, = 0.22), Raman intothe theRaman delivery optical fiber (19optical µm × fiber 100 µm, NA (19 = 0.22), longNA pass filter with a on long its front to get rid of Rayleigh scattering placed itspass frontfilter to getplaced rid of on Rayleigh scattering from the collected signal. from Fromthe the collected other endsignal. of the Fromthe thesignal otherisend the fiber, signal is then(Acton coupled into aPrinceton spectrometer (Acton SP2500i, fiber, thenofcoupled intothe a spectrometer SP2500i, Instruments, Trenton, Princeton Instruments, Trenton, NJ, USA), with 1200 g/mm grating 10 μm entrance slit width. NJ, USA), with 1200 g/mm grating and 10 µm entrance slit width. Theand Raman spectra are recorded by Raman spectra are recorded by a CCDInstruments, detector (SPEC-10: Princeton aThe CCD detector (SPEC-10: 400B, Princeton Trenton,400B, NJ, USA) with Instruments, a 1340 × 400 Trenton, imaging ◦ C. × −1 NJ, USA) with × 400 imaging array and 20 resulted μm pixel size operating at −40 °C.cm The array and 20 µma×1340 20 µm pixel size operating at −20 40μm The spectral range is 800~2000 −1 . To get resulted spectral range is 800~2000 cma−1relative with a resolution of 2 cm−1signal . To getand a relative desirable Raman with a resolution of 2 cm desirable Raman prevent the CCD from signal and prevent CCD from overexposure, is accumulated five with an overexposure, each the spectrum is accumulated five each timesspectrum with an integration time of 10times s, so the total integration time of 10 ss,for so one the total acquisition time is 50 s for sample. Fivesample measurements are acquisition is 50 sample. Five measurements areone taken for each to evaluate taken for each sample to evaluate the repeatability. the repeatability. reducer Nd:YAG CW Laser 532nm

M1-1 HWP

L2

L1 M2 LPF

Fiber

M1-2 Sample Cell 50000

20% CH4+20% C2 H6

Spectrograph PI SP2500i

intensity a.u.

40000 30000 20000 10000 0 2600

CCD PI OMA IV

2800

3000

Raman shift cm

3200

-1

Figure 1. Schematic of the near-concentric cavity-enhanced Raman spectroscopy system for liquid Figure 1. Schematic of the near-concentric cavity-enhanced Raman spectroscopy system for liquid sample detection. sample detection.

3. Results The reported LOD for SO42− by using a liquid core optical fiber Raman system is about 1.5 mmol/L [15]. As a comparison, we prepared 15 Na2SO4 solutions (0.10 mmol/L–2.00 mmol/L) to evaluate the LOD of the NC-CERS system. The LOD is defined as the ratio of three times the noise

Sensors 2017, 17, 2784

3 of 6

3. Results The reported LOD for SO4 2− by using a liquid core optical fiber Raman system is about 1.5 mmol/L [15]. As a comparison, we prepared 15 Na2 SO4 solutions (0.10 mmol/L–2.00 mmol/L) to Sensors 2017, 17, 2784 3 of 6 evaluate the LOD of the NC-CERS system. The LOD is defined as the ratio of three times the noise intensity (σ) to(σ) the of calculation curve. offour foursolutions solutions shown in Figure Sensors 2017, 17, 2784 3 of2a, 6 2a, intensity to slope the slope of calculation curve.The The spectra spectra of areare shown in Figure 2− can 2− and the Raman signal of SO be observed clearly in the spectrum of the 0.10 mmol/L solution. and the Raman signal of4SO4 can be observed clearly in the spectrum of the 0.10 mmol/L solution. 2which − which (σ) to the slope ofsystem calculation The spectra of four solutions shown intimes Figure 2a,lower Given this, the NC-CERS a curve. LOD less than 0.10mmol/L mmol/L for is 15 lower Givenintensity this, the NC-CERS system hashas a LOD less than 0.10 forSO SO42−4are is 15 times and the Raman signal of SO 42− can be observed clearly in the spectrum of the 0.10 mmol/L solution. thanreported that reported before. linear relationshipsbetween between concentrations and peak intensity of the than that before. TheThe linear relationships concentrations and peak intensity of the Given this, the NC-CERS system has a LOD less than 0.10 mmol/L for SO42− which is 15 times lower 42− are shown in Figure 2b and the response of the NC-CERS system for liquid detection has good SO4 2−SO are shown in Figure 2b and the response of the NC-CERS system for liquid detection has good than that reported Thethe linear relationships between concentrations and peak intensity of the 2 = before. linearity 0.996 over whole range. linearity with with R2 =R0.996 over the whole range. 2− Normalized Intensity (a.u.) (a.u.) Normalized Intensity

SO4 are shown in Figure 2b and the response of the NC-CERS system for liquid detection has good linearity with R2 = 0.996 over the whole range. (b)

(a) 4

SO24

CSO4 (mmol/L) 2-

(a) 981 cm-1

2CSO4 4 SO4 2 981 cm-1

2-

2 0

0.10 0.50 (mmol/L) 1.00 0.10 2.00 0.50 1.00 2.00

4

(b)

SO24 981 cm-1 SO24

4 2

981 cm-1

2 0 950 1000 1050 1100 0.0

900 0 Raman shift (cm-1)

R2=0.996 1.0 1.5 2.0

0.5

Concentration (mmol/L)

0 950 1000 1050 1100 0.0

900

R2=0.996

0.5

1.0

1.5

2.0

-1

Concentration (mmol/L) Raman (cmsolution ) FigureFigure 2. Raman spectra of aof range of Na 2.00mmol/L). mmol/L). (a) Four spectra 2. Raman spectra a range ofshift Na 2SO (0.10 mmol/L mmol/L toto2.00 (a) Four spectra 2 SO 4 4 solution(0.10 of typical concentration. (b) The linear relationship between concentrationsand andpeak peak intensities intensities of of typical concentration. (b) The linear relationship between concentrations of the Figure the SO2. 42−Raman signal. spectra of a range of Na2SO4 solution (0.10 mmol/L to 2.00 mmol/L). (a) Four spectra SO4 2− signal.

of typical concentration. (b) The linear relationship between concentrations and peak intensities of the 42− signal. ToSO evaluate its performance for HCO3−, we prepared nine NaHCO3 solutions with concentrations To evaluate its performance for HCO − , we prepared nine NaHCO solutions with concentrations 3 3 ranging from 0.40 mmol/L to 4.00 mmol/L. The corresponding concentrations of HCO3− are from

To evaluate its performance for HCOby 3−, we prepared ninebalance NaHCO 3 solutions with ranging 0.40tommol/L to 4.00 mmol/L. The corresponding concentrations ofconcentrations HCO3 − are from 0.37from mmol/L 3.79 mmol/L calculated using the carbon mode. All of the concentrations − are from ranging 0.40 mmol/L to 4.00the mmol/L. The the corresponding concentrations 0.37 mmol/L 3.79 mmol/L by using carbon balance mode. Allof the 3concentrations of HCOfrom 3−to are corrected bycalculated using equilibrium mode of carbon components inofHCO water. Figure 3a − 0.37 mmol/L to 3.79 of mmol/L by using the carbon mode. thein concentrations of HCO by the concentrations. equilibrium mode ofbalance carbon components water. signal Figure 3a shows thecorrected spectra fiveusing ofcalculated these The peak intensity of All the of HCO 3− Raman 3 are of HCO 3− are corrected by using the equilibrium mode of carbon components inbewater. Figure 3a − Raman increases as a function of these the concentration, and the Raman of HCO canHCO distinguished in shows the spectra of five of concentrations. The peaksignal intensity of3−the signal 3 shows the spectra of five of these concentrations. The peak intensity ofNC-CERS the HCO 3− Raman signal − − can the solution with 0.37 NaHCO 3. It isand proved the LOD of the system for HCO 3 increases as a function ofmmol/L the concentration, the that Raman signal of HCO be distinguished 3 increases as a 0.37 function of thewhich concentration, and the than Raman signal of HCO3− in canseawater. be distinguished in is less than mmol/L, is much lower its concentration The linear in the solution with 0.37 mmol/L NaHCO3 . It is proved that the LOD of the NC-CERS system for the solution with 0.37 mmol/L NaHCOand 3. It is proved that the LOD of the− NC-CERS system for HCO3− relationship between concentrations peak intensities of the HCO3 is shown in Figure 3b and it − is less than HCOis 0.37 mmol/L, which is much lower than its concentration in seawater. The linear 3 less than 0.37 mmol/L, which is much lower than its concentration in seawater. The linear shows good linearity with R2 = 0.978. − is shown in Figure 3b and it relationship between concentrations ofthe theHCO HCO relationship between concentrationsand andpeak peakintensities intensities of 3− 3is shown in Figure 3b and it 2 shows goodgood linearity with R R=2 =0.978. shows linearity with 0.978. 10 10 -2

-2

Normalized IntensityIntensity (a.u.) (a.u.) Normalized

1.2

CHCO3 (mmol/L)

(a)

-

0.37 1.16

HCO-3

10 0.8-2 1021 cm-1

2.0 (b) Calibration curve of HCO-3 1.6

10-2

2

1.2 R =0.978 2.0 (b) Calibration curve 3.06 0.8 0.37 of HCO-3 1.6 HCO3 3.79 1.16 0.0 1021 cm-1 0.8 0.4 R2=0.978 1.2 2.31 3.06 0.4900 1000 1100 1200 0.8 0 1 2 3 4 -1 3.79 Raman shift (cm ) 10-1Concentration (mmol/L) 0.0 0.4 (c) 0.9 (d) C CHCO3 (mmol/L) C 2-=28.00 mmol/L SO4 (a)

-

(c)

0.4 0.8

0.0 0.4900 0.0 900

SO2981 cm-1 24 SO4 981 cm-1

0.4 1.2

900 0.8

2.31 CHCO3 (mmol/L)

1000 SO4 1100 1200 (mmol/L) C 0.00 1.93 mmol/L CSO4 =28.00 3.79 (mmol/L) C

RamanHCO3 shift (cm-1) -

2-

2-

0

1 2 28.00 mmol/L

10 0.6-1Concentration 0.9 (d) C C

HCO3-

SO42-

0.3 28.00 mmol/L HCO3 HCO-3 1021 cm-1 0.6 HCO-3 1021 cm-1 0.00 0.0 1000 1.931100 1200 1000 1025 0.3 3.79 -1 -

Raman HCO shift-3 1021 (cmcm)-1 1000

1100

1200

3

-

4 0.00

(mmol/L) 1.93 3.79 (mmol/L) 0.00 1.93 3.79

1050

Raman shift (cm-1)

HCO-3 1021 cm-1

0.0 1000

1025

1050

Figure 3. The detection abilityRaman of theshift NC-CERS system for HCO 3−.-1(a) The spectra of 5 NaHCO3 (cm-1) Raman shift (cm ) aqueous solutions with different concentrations. (b) The linear relationship between concentrations − signal. Figurepeak 3. The detection of 3the NC-CERS system for HCO3−.of (a) The mixed spectrasolutions of 5 NaHCO 3 (c) The Raman intensities of ability the HCO Figureand 3. The detection ability of the NC-CERS system forspectra HCO3 − . three (a) The spectra of 5 which NaHCO3 aqueous solutions with different concentrations. (b) The linear relationship between concentrations 4 concentration (28.00 mmol/L) and different NaHCO3 concentrations. (d) The contains fixedwith NaSO aqueous solutions different concentrations. (b) The linear relationship between concentrations and − The Raman spectra of three mixed solutions which and peakspectra intensities of the HCO mixed (c) solutions. detailed (1000–1050 cm−1)3ofsignal. peak intensities of the HCO3 − signal. (c) The Raman spectra of three mixed solutions which contains contains fixed NaSO4 concentration (28.00 mmol/L) and different NaHCO3 concentrations. (d) The fixeddetailed NaSO4 spectra concentration (28.00 and different NaHCO3 concentrations. (d) The detailed −1)mmol/L) of mixedseawater solutions. (1000–1050 cmsimulated Furthermore, we prepared with fixed a NaSO4 concentration (28.00 mmol/L)

spectra (1000–1050 cm−13 )concentrations of mixed solutions. and different NaHCO (0.00, 2.00, 4.00 mmol/L) to evaluate its ability to quantify Furthermore, we prepared simulated seawater with fixed a NaSO4 concentration (28.00 mmol/L) HCO 3− in seawater. The corresponding concentrations of HCO3− are 0.00, 1.93 and 3.79 mmol/L. The and different NaHCO3 concentrations (0.00, 2.00, 4.00 mmol/L) to evaluate its ability to quantify HCO3− in seawater. The corresponding concentrations of HCO3− are 0.00, 1.93 and 3.79 mmol/L. The

Sensors 2017, 17, 2784

4 of 6

Furthermore, we prepared simulated seawater with fixed a NaSO4 concentration (28.00 mmol/L) and − in different NaHCO Sensors 2017, 17, 27843 concentrations (0.00, 2.00, 4.00 mmol/L) to evaluate its ability to quantify HCO3 4 of 6 seawater. The corresponding concentrations of HCO3 − are 0.00, 1.93 and 3.79 mmol/L. The corresponding spectra are shown in Figure and the detailed aredetailed shown in Figureare 3d. shown Compared with the corresponding spectra are 3c, shown in Figure 3c,spectra and the spectra in Figure 3d. 2− , the intensity of the − signal is much weaker. 2− − intensity of SO HCO Even so, the Raman signal of Compared with of the HCO3 signal is much weaker. Even so, the 4 the intensity of SO4 , the intensity 3 − HCO be detected the spectrum of the solutionof with mmol/L NaHCO Thismmol/L shows Raman signal of HCOin 3− can be detected in mixed the spectrum the1.93 mixed solution with3 .1.93 3 can − detection in seawater, but − the ability of the NC-CERS system for direct HCO the strong and adjacent NaHCO3. This shows the ability of the NC-CERS3 system for direct HCO3 detection in seawater, but − signal of seawater has 2− − SO impact the quantitative analysis the4 2strong and adjacent SO4 ansignal ofon seawater has an impact on of theHCO quantitative analysis of HCO3−. 3 . − In extraction method forfor HCO Inorder ordertotoreduce reducethis thisimpact, impact,we wedeveloped developedananadaptive adaptivesignal signal extraction method HCO 3 3− and shown in inFigure Figure4a–c. 4a–c.The The spectrum is obtained from a mixed solution andthe the process process is shown spectrum is obtained from a mixed solution with with 28.00 2− −1) and 2− 28.00 mmol/L 1.93 mmol/L spectrum, Raman mmol/L Na2SONa 4 and 1.93 mmol/L NaHCO3NaHCO . In this spectrum, Raman the peaks of SOpeaks 4 (981of cmSO 2 SO 4 and 3 . In this the 4 − − 1 − 1 − −1)HCO (981 and−1)Hcan (1640 cm recognized, ) can be clearly peak cm 1 ) H2Ocm (1640) cm clearly whilerecognized, the peak of while HCO3−the (1021 cmof is almost submerged 2 O be 3 (1021 isinalmost submergedGiven in thethis, background. Given this, the original spectrum is normalized according the background. the original spectrum is normalized according to the peak intensity of to peak intensity of H2 O reduce thefluctuation impact of laser power as shown in Figure 4a. Hthe 2O to reduce the impact of to laser power as shown in fluctuation Figure 4a. The next step is baseline The next stepWe is baseline correction.four We baseline have compared baseline results with four Gaussian, methods correction. have compared fitting four results with fitting four methods (Voigt, (Voigt, Gaussian, and bi-exponential fitting methods), and the polynomial functionthe shows Lorentzian and Lorentzian bi-exponential fitting methods), and the polynomial function shows best the best performance amongThe them. The baseline corrected shown a blue line in4b. Figure 4b. performance among them. baseline corrected result isresult shownisas a blueasline in Figure The last − signal extraction from the neighboring strong 2− signal. The right − 2− The last step is the weak HCO SO step is the weak HCO3 signal 3 extraction from the neighboring strong SO4 signal. 4 The right wing of 2− signal is regarded as the baseline of wing of42−the SO4is the HCO signal is afitted by using the SO signal regarded as the baseline of the HCO3− signal and 3is−fitted byand using bi-exponential −1. We again afunction bi-exponential region ofcm 1000–1050 cm−1 .tried We again four baseline fitting methods in thefunction region in of the 1000–1050 four tried baseline fitting methods (Voigt, (Voigt, Gaussian, Lorentzian and bi-exponential fitting methods), the polynomial function showed Gaussian, Lorentzian and bi-exponential fitting methods), and and the polynomial function showed the the performance among them. The fitted baseline is shown as thedotted greenline, dotted and the bestbest performance among them. The fitted baseline is shown as the green andline, the corrected corrected as circle the black dots4c. in The Figure 4c. The peak HCO3 −and is spectrumspectrum is shown is as shown the black dotscircle in Figure Raman peakRaman of HCO 3− is of extracted extracted fitted by using a Gaussian function which is shown red line fitted by and using a Gaussian function which is shown as the red lineasinthe Figure 4c. in Figure 4c. 10-3

0.6

8 4

0.3

Intensity (a.u.)

0

HCO3- 1021 cm-1

0.0

1000 1050

1.0

(b) SO2981 cm-1 4

0.5

1200

850

(a)

Normalized data

0.5 0.0

SO24 981 cm-1

1000

-1

H2O 1640 cm

1500

Raman shift (cm-1)

2000

(d)

CSO4 28.00 mmol/L 2-

CHCO3 (mmol/L)

8 HCO3

-

0.00 0.97 1.93 3.79 seawater

1021 cm-1

4 0

Baseline A Corrected spectrum

0.0

1.0

12

Intensity (a.u.)

Gaussian fitting HCO3-

Baseline B

1000

Peak area (10-1 a.u.)

10-1 0.9 (c)

1020

1040

1060

Raman shift (cm-1) 2

(e)

HCO3- + SO42- solutions linear fitting

1.1

1

R2=0.951

0.8

1.91 mmol/L

0 0

1.43

2

4

Concentration (mmol/L)

− . (a) Figure4.4.The The signal extraction method for quantitative HCO 3−.normalized (a) The normalized Figure signal extraction method for quantitative analysisanalysis of HCO3of The spectrum 2 SO 4 and 1.93 mmol/L NaHCO 3. (b) Baseline spectrum of mixed solution with 28.00 mmol/L Na of mixed solution with 28.00 mmol/L Na2 SO4 and 1.93 mmol/L NaHCO3 . (b) Baseline correction. correction. Theline redrepresents dotted line the baseline fitted byfunction, a polynomial function, the The red dotted therepresents baseline curve fitted bycurve a polynomial and the blue lineand is the − . The green 3−. The green dotted blue linecorrected is the baseline curve. (c) The signal extraction for HCO baseline curve. corrected (c) The signal extraction method for HCOmethod dotted line represents 3 linebaseline represents the by baseline B fitted by a double exponential function. extracted signal is shown the B fitted a double exponential function. The extracted signalThe is shown as black circle dots − Raman signals − as black circle dots and fitted by using a Gaussian function (red line). (d) Fitted HCO 3 and fitted by using a Gaussian function (red line). (d) Fitted HCO3 Raman signals of 1000 m depth of 1000 m depth and four mixed solutions with fixed 28.00 mmol/L Na2concentrations SO4 and different seawater and fourseawater mixed solutions with fixed 28.00 mmol/L Na2 SO different of 4 and − . The −. The pink point is the result of a blind concentrations of NaHCO 3 . (e) The calibration curve of HCO 3 NaHCO . (e) The calibration curve of HCO pink point is the result of a blind sample, and the blue 3 3 sample, and the of blue point is the result of a seawater sample. point is the result a seawater sample.

Based on onthis thisdata dataprocessing processingmethod, method,the theRaman Ramansignals signalsofofHCO HCO−3− in in simulated simulatedseawater seawater Based 3 solutionsare areextracted extractedand andshown shownininFigure Figure4d. 4d.The Thesignal signalof ofaadeep deepseawater seawatersample sampleisisalso alsoshown shown solutions as the cyan line in Figure 4d. The linear relationship between concentrations and peak intensities of as the cyan line in Figure 4d. The linear relationship between concentrations and peak intensities − 2 fitted HCO3 is shown in Figure 4e and it shows high linearity with R = 0.951 over the range. We chose a mixed solution with 28.00 mmol/L NaSO4 and 1.50 mmol/L NaHCO3 as a blind sample. The test result is 1.35 mmol/L with 5.0% relative error, which is shown as the pink point in Figure 4e. Furthermore, we tested deep seawater, knowing the concentration of HCO3− is 1.91 mmol/L which is shown as a blue point in Figure 4e. Compared with the reported HCO3− concentration value of 1.95 mmol/L in seawater [4], the relative error was about 2.1%.

Sensors 2017, 17, 2784

5 of 6

of fitted HCO3 − is shown in Figure 4e and it shows high linearity with R2 = 0.951 over the range. We chose a mixed solution with 28.00 mmol/L NaSO4 and 1.50 mmol/L NaHCO3 as a blind sample. The test result is 1.35 mmol/L with 5.0% relative error, which is shown as the pink point in Figure 4e. Furthermore, we tested deep seawater, knowing the concentration of HCO3 − is 1.91 mmol/L which is shown as a blue point in Figure 4e. Compared with the reported HCO3 − concentration value of 1.95 mmol/L in seawater [4], the relative error was about 2.1%. 4. Conclusions In this paper, we developed a near-concentric cavity-enhanced Raman spectroscopy system for liquid sample detection based on a near-concentric cavity and directly detected bicarbonate in seawater for the first time. Systematic experiments have been carried out for system evaluation. As shown by the NC-CERS system tests, the LOD for SO4 2− is less than 0.10 mmol/L, which is 15 times lower than the result reported before. The LOD for HCO3 is less than 0.37 mmol/L, which is much lower than its concentration in seawater (1.95 mmol/L). The detection performance of the NC-CERS system for HCO3 − in seawater is evaluated by analyzing mixed solutions of fixed 28.00 mmol/L NaSO4 and different NaHCO3 concentrations. The Raman signal of 1.93 mmol/L HCO3 − in mixed solution can be clearly detected, which shows the ability of direct HCO3 − detection in seawater. To further realize the quantitative analysis for HCO3 − , a specially developed data processing scheme is introduced. With this data processing method, the weak HCO3 − signal is extracted from the neighboring strong SO4 2− signal, hence a quantitative analysis with R2 = 0.951 is made possible. Based on this calibration curve, the HCO3 − concentrations of a blind sample and a seawater sample are calculated. The calculated values are 1.43 mmol/L and 1.91 mmol/L with 4.9% and 2.1% relative error, respectively. It is hoped that this NC-CERS system could be fully developed for field ocean observations in the near future. CO3 2− with a concentration down to 1/10 of the HCO3 − in seawater will be the next detection object. Acknowledgments: This work was supported by National Natural Science Foundation of China (61575181), Fundamental Research Funds for the Central Universities (201562008), National Key Research and Development Program of China (2016YFC0302101), the Strategic Priority Research Program, CAS (XDA11040301). The author (DewangYang) would like to appreciate Zhao Luo and Tengyue Wang’s language revision. Author Contributions: Jinjia Guo and Ronger Zheng initiated and supervised the research project. Dewang Yang and Jinjia Guo proposed the idea, conceived and designed the experiments. Dewang Yang, Jinjia Guo, Chunhao Liu and Qingsheng Liu performed the experiments. Dewang Yang processed the data and wrote the initial draft. Ronger Zheng revised the manuscript and figures. The final revisions and writing were done by Jinjia Guo. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4.

5.

6. 7.

Bauer, J.E.; Cai, W.J.; Raymond, P.A.; Bianchi, T.S.; Hopkinson, C.S.; Regnier, P.A. The changing carbon cycle of the coastal ocean. Nature 2013, 504, 61–70. [CrossRef] [PubMed] Pilson, M.E. An Introduction to the Chemistry of the Sea; Cambridge University Press: Cambridge, UK, 2012; pp. 114–117. Doney, S.C.; Fabry, V.J.; Feely, R.A.; Kleypasl, J.A. Ocean acidification: The other CO2 problem. Annu. Rev. Mar. Sci. 2009, 1, 169–192. [CrossRef] [PubMed] Dunk, R.M.; Peltzer, E.T.; Walz, P.M.; Brewer, P.G. Seeing a deep ocean CO2 enrichment experiment in a new light: Laser Raman detection of dissolved CO2 in seawater. Environ. Sci. Technol. 2005, 39, 9630–9636. [CrossRef] [PubMed] Bates, N.R.; Best, M.H.P.; Neely, K.; Garley, R.; Dickson, A.G.; Johnson, R.J. Detecting anthropogenic carbon dioxide uptake and ocean acidification in the North Atlantic Ocean. Biogeosci. Discuss. 2012, 9, 2509–2522. [CrossRef] Millero, F.J. The Marine Inorganic Carbon Cycle. Chem. Rev. 2007, 107, 308–341. [CrossRef] [PubMed] Pasteris, J.D.; Wopenka, B.; Freeman, J.J.; Brewer, P.G.; White, S.N.; Peltzer, E.T.; Malby, G.E. Raman spectroscopy in the deep ocean: Successes and challenges. Appl. Spectrosc. 2004, 58, 195A–208A. [CrossRef] [PubMed]

Sensors 2017, 17, 2784

8. 9.

10.

11. 12. 13. 14. 15.

6 of 6

Brewer, P.G.; Dunk, R.M.; White, S.N.; Peltzer, E.T.; Bowie, B.; Walz, P.M. First Attempts at Direct Raman Detection of the Oceanic Carbonate System; American Geophysical Union: Washington, DC, USA, 2004. Scholl, J.; Brewer, P.; Peltzer, E.; Walz, P. Liquid-Core Waveguide for Enhanced Laser Raman Spectroscopy in Oceanic Applications. Available online: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.598. 451&rep=rep1&type=pdf (accessed on 1 December 2017). Brewer, P.G.; Malby, G.; Pasteris, J.D.; White, S.N.; Peltzer, E.T.; Wopenka, B.; Freeman, J.; Brown, M.O. Development of a laser Raman spectrometer for deep-ocean science. Deep-Sea Res. Part I 2004, 51, 739–753. [CrossRef] Hippler, M. Cavity-enhanced Raman spectroscopy of natural gas with optical feedback cw-diode lasers. Anal. Chem. 2015, 87, 7803–7809. [CrossRef] [PubMed] Li, X.; Xia, Y.; Zhan, L.; Huang, J. Near-confocal cavity-enhanced Raman spectroscopy for multitrace-gas detection. Opt. Lett. 2008, 33, 2143–2145. [CrossRef] [PubMed] Taylor, D.J.; Glugla, M.; Penzhorn, R.D. Enhanced Raman sensitivity using an actively stabilized external resonator. Rev. Sci. Instrum. 2001, 72, 1970–1976. [CrossRef] Yang, D.; Guo, J.; Liu, Q.; Luo, Z.; Yan, J.; Zheng, R. Highly sensitive Raman system for dissolved gas analysis in water. Appl. Opt. 2016, 55, 7744–7748. [CrossRef] [PubMed] Du, Z.; Chen, J.; Ye, W.; Guo, J.; Zhang, X.; Zheng, R. Investigation of two novel approaches for detection of sulfate ion and methane dissolved in sediment pore water using Raman spectroscopy. Sensors 2015, 15, 12377–12388. [CrossRef] [PubMed] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).