Aquaphotomics and its extended water mirror concept...

1
Aquaphotomics and its extended water mirror concept explain why NIRS can measure low concentration aqueous solutions George Bázár, 1,2, * Zoltán Kovács, 1,3 Mariko Tanaka, 1 Roumiana Tsenkokva 1, * 1 Biomeasurement Technology Laboratory, Graduate School of Agriculture, Kobe University, Japan 2 Institute of Food and Agricultural Product Qualification, Faculty of Agricultural and Environmental Sciences, Kaposvár University, Hungary 3 Department of Physics and Control, Faculty of Food Science, Corvinus University of Budapest, Hungary * To whom correspondence should be addressed E-mail: [email protected] (G.B.); [email protected] (R.T.) INTRODUCTION Low concentration aqueous solutions low spectral intensity of solutes REFERENCES & ACKNOWLEDGEMENTS [1] Tsenkova, R., 2009. Journal of Near Infrared Spectroscopy, 17, 303-314. [2] Giangiacomo, R., 2006. Food Chemistry, 96, 371-379. [3] Tsenkova, R., 2010. Spectroscopy Europe, 22, 6-10. http://www.brooklyn.cuny.edu/bc/ahp/SDPS/SD.PS.water.html http://www1.lsbu.ac.uk/water/inos.html 10-100mM 1-10mM 0.1-1mM 0.02-0.1mM 1100-1800nm factor# 3 3 4 1.75±1.5 R 2 Cal 0.99±0.00 0.99±0.01 0.84±0.12 0.72±0.09 RMSEC 0.99±0.06 0.34±0.06 0.11±0.04 0.02±0.01 R 2 CV 0.99±0.00 0.97±0.02 0.48±0.27 -1.17±0.46 RMSECV 1.34±0.22 0.48±0.13 0.19±0.05 0.03±0.00 RPD CV 22.63 6.33 1.56 1.05 1300-1600nm factor# 3 3 3.75±0.5 1.75±1.5 R 2 Cal 0.99±0.00 0.99±0.01 0.80±0.15 0.68±0.09 RMSEC 0.99±0.07 0.34±0.07 0.13±0.07 0.02±0.01 R 2 CV 0.99±0.00 0.97±0.01 0.33±0.33 -0.91±0.70 RMSECV 1.34±0.23 0.48±0.12 0.22±0.05 0.03±0.00 RPD CV 22.59 6.37 1.37 1.07 10-100mM 1-10mM 0.1-1mM 0.02-0.1mM 1100-1800nm factor# 3 3 3 1 R 2 Cal 0.99 0.98 0.35 0.07 RMSEC 1.03 0.40 0.23 0.03 R 2 CV 0.99 0.98 0.24 -0.1 RMSECV 1.08 0.42 0.25 0.03 RPD CV 27.94 7.19 1.23 1.15 1300-1600nm factor# 3 3 3 1 R 2 Cal 0.99 0.98 0.33 0.05 RMSEC 1.04 0.40 0.23 0.03 R 2 CV 0.99 0.98 0.23 -0.04 RMSECV 1.08 0.42 0.25 0.03 RPD CV 28.04 7.13 1.22 1.15 The molecular changes of the water solvent caused by the solute can be detected by NIR spectroscopy and used for qualitative and quantitative analyses [1] Means and standard deviations of PLSR results on lactose concentration in the four replicate solutions (n = ca.30) at the different concentration ranges, applying different spectral regions PLSR results on lactose concentration at the different concentration ranges using the four replicate solutions together (n = ca.120), applying different spectral regions and ‘leave one replicate out’ cross-validation 1390-1425nm 1440-1475nm 1580-1615nm 1680-1700nm 1720-1755nm 1100 1350 1600 1800 Wavelengths [nm] 10 0 -10 -20 -30 10 -10 -30 50 0 -50 Regression power Regression power Regression power Structural changes of water in sugar solutions presented on aquagram BUT! Measuring low concentration of sugar in aqueous solutions based on the molecular changes of water [2] MATERIALS AND METHODS RESULTS 10-100mM 1-10mM 0.1-1mM - Analytical grade lactose diluted in Milli-Q water 0.02-0.1mM range, with 0.01mM steps 0.1-1mM range, with 0.1mM steps 1mM-10mM range, with 1mM steps 10mM-100mM range, with 10mM steps - Four independent replicates were prepared for each concentration range - FOSS XDS spectrometer (FOSS NIRSystems, Inc., Hoganas, Sweden) with RLA module and 1mm cuvette transmittance spectra (logT -1 ) at 1100- 1800nm with 0.5nm spectral step - R Project (www.r-project.org), The Unscrambler v9.7 (CAMO Software AS, Oslo, Norway), Pirouette 4.0 (Infometrics, Inc., Woodinville, WA, USA) Characteristic changes at water matrix coordinates describe the water spectral pattern of lactose. Visualization on aquagram [3]: 1344nm 1364nm 1412nm 1445nm 1475nm 1506nm 1590nm 1380nm 1400nm 1425nm 1440nm 1455nm A = Aquagram value a = absorbance after MSC μ = average of all spectra σ = SD of all spectra A= a− μ σ 0 0.5 1.0 1.5 Absorbance (water) 0 0.5 1.0 1.5 Absorbance (solutions) 0 0.125 0.25 1.5 Absorbance (lactose) -2 -1 0 1 2nd deriv. x10 -4 (water) 2 2 -2 -1 0 1 2nd deriv. x10 -4 (solutions) 2 2 -0.5 -0.25 0 0.25 0.5 0.75 1100 1300 1500 1700 1100 1300 1500 1700 1100 1300 1500 1700 1100 1300 1500 1700 1152 1412 1462 1344 1780 1140 1175 1208 1276 1302 1342 1366 1380 1450 1490 1536 1592 1616 1646 1696 1728 1764 1796 Raw spectra 2nd derivative spectra 2nd deriv. x10 -4 (lactose) water lactose Aqueous solutions of lactose (0.02-100mM) Signals of lactose can hardly be detected in low concentration solutions. Subtraction of water spectrum from those of the lactose solutions reveals the spectral differences of the different concentrations. Only water bands dominate Regression vectors of calibration models built on the subtracted spectra of water and lactose solutions show stability among replicate measurments, and accentuate the dominant water regions. 1-10mM 10-100mM CONCLUSIONS NIR technique coupled with aquaphotomics concept is useful method for quantification of the investigated carbohydrate solutes at millimolar level. Lactose causes gradual changes in the water molecular matrix, thus, performance of calibration on lactose concentration does not decrease when only the characteristic spectral region of water 1st overtone (1300-1600nm) is applied. In the contrary, it has been demonstrated in the present study that the absorption regions of water provide most useful information on the solutes in case of highly diluted aqueous solutions. Accordingly, the molecular changes of water caused by the solute can be traced and used for describing the amount of dissolved material. NIR technique and aquaphotomics based on the extended water mirror concept provide a quick and accurate alternative for classical analytical measurements even at very low (1-10mM) concentration levels. Financial support of Japan Society for Promotion of Science (G.B. JSPS P12409) is acknowledged. Dynamic changes of structure maker and structure breaker characteristics of sugar in water [2]. Breaker is dominant in low, while maker in high concentrations. H + ·(H 2 O) 4-8 OH-(H 2 O) 1,2,4 H + ·(H 2 O) 2 OH-(H 2 O) 1,4 O 2 -(H 2 O) 4 S0 Free -OH S0 OH-H bending H + ·(H 2 O) 2 S1 Deionized water OH-(H 2 O) 4,5 2ν 2 +ν 3 S2, S3 aν 1 +bν 3 (a+b)=2 H + ·(H 2 O) 2,6 H + ·(H 2 O) 2,6

Transcript of Aquaphotomics and its extended water mirror concept...

Page 1: Aquaphotomics and its extended water mirror concept ...aquaphotomics.com/wp-content/uploads/2016/07/nr2.pdfAquaphotomics and its extended water mirror concept explain why NIRS can

Aquaphotomics and its extended water mirror concept explain why NIRS can measure low concentration aqueous solutions

George Bázár,1,2,* Zoltán Kovács,1,3 Mariko Tanaka,1 Roumiana Tsenkokva 1,* 1 Biomeasurement Technology Laboratory, Graduate School of Agriculture, Kobe University, Japan

2 Institute of Food and Agricultural Product Qualification, Faculty of Agricultural and Environmental Sciences, Kaposvár University, Hungary 3 Department of Physics and Control, Faculty of Food Science, Corvinus University of Budapest, Hungary

* To whom correspondence should be addressed E-mail: [email protected] (G.B.); [email protected] (R.T.)

INTRODUCTION

Low concentration aqueous solutions low spectral intensity of solutes

REFERENCES & ACKNOWLEDGEMENTS [1] Tsenkova, R., 2009. Journal of Near Infrared Spectroscopy, 17, 303-314. [2] Giangiacomo, R., 2006. Food Chemistry, 96, 371-379. [3] Tsenkova, R., 2010. Spectroscopy Europe, 22, 6-10.

http://www.brooklyn.cuny.edu/bc/ahp/SDPS/SD.PS.water.html

http://www1.lsbu.ac.uk/water/inos.html

10-100mM 1-10mM 0.1-1mM 0.02-0.1mM

11

00

-18

00

nm

factor# 3 3 4 1.75±1.5

R2Cal 0.99±0.00 0.99±0.01 0.84±0.12 0.72±0.09

RMSEC 0.99±0.06 0.34±0.06 0.11±0.04 0.02±0.01

R2CV 0.99±0.00 0.97±0.02 0.48±0.27 -1.17±0.46

RMSECV 1.34±0.22 0.48±0.13 0.19±0.05 0.03±0.00

RPDCV 22.63 6.33 1.56 1.05

13

00

-16

00

nm

factor# 3 3 3.75±0.5 1.75±1.5

R2Cal 0.99±0.00 0.99±0.01 0.80±0.15 0.68±0.09

RMSEC 0.99±0.07 0.34±0.07 0.13±0.07 0.02±0.01

R2CV 0.99±0.00 0.97±0.01 0.33±0.33 -0.91±0.70

RMSECV 1.34±0.23 0.48±0.12 0.22±0.05 0.03±0.00

RPDCV 22.59 6.37 1.37 1.07

10-100mM 1-10mM 0.1-1mM 0.02-0.1mM

11

00

-18

00

nm

factor# 3 3 3 1

R2Cal 0.99 0.98 0.35 0.07

RMSEC 1.03 0.40 0.23 0.03

R2CV 0.99 0.98 0.24 -0.1

RMSECV 1.08 0.42 0.25 0.03

RPDCV 27.94 7.19 1.23 1.15

13

00

-16

00

nm

factor# 3 3 3 1

R2Cal 0.99 0.98 0.33 0.05

RMSEC 1.04 0.40 0.23 0.03

R2CV 0.99 0.98 0.23 -0.04

RMSECV 1.08 0.42 0.25 0.03

RPDCV 28.04 7.13 1.22 1.15

The molecular changes of the water solvent caused by the solute can be detected by NIR spectroscopy and used for qualitative and quantitative analyses [1]

Means and standard deviations of PLSR results on lactose concentration in the four replicate solutions (n = ca.30) at the

different concentration ranges, applying different spectral regions

PLSR results on lactose concentration at the different concentration ranges using the four replicate solutions together (n = ca.120), applying different spectral regions and ‘leave one replicate out’ cross-validation

13

90

-14

25

nm

14

40

-14

75

nm

15

80

-16

15

nm

16

80

-17

00

nm

17

20

-17

55

nm

1100 1350 1600 1800 Wavelengths [nm]

10

0

-10

-20

-30

10

-10

-30

50

0

-50 Reg

ress

ion

po

wer

R

egre

ssio

n p

ow

er

Reg

ress

ion

po

wer

Structural changes of water in sugar solutions presented on aquagram

BUT!

Measuring low concentration of sugar in aqueous solutions based on the molecular changes of water [2]

MATERIALS AND METHODS RESULTS

10-100mM

1-10mM

0.1-1mM

- Analytical grade lactose diluted in Milli-Q water 0.02-0.1mM range, with 0.01mM steps 0.1-1mM range, with 0.1mM steps 1mM-10mM range, with 1mM steps 10mM-100mM range, with 10mM steps - Four independent replicates were prepared for each concentration range - FOSS XDS spectrometer (FOSS NIRSystems, Inc., Hoganas, Sweden) with RLA module and 1mm cuvette transmittance spectra (logT-1) at 1100-1800nm with 0.5nm spectral step - R Project (www.r-project.org), The Unscrambler v9.7 (CAMO Software AS, Oslo, Norway), Pirouette 4.0 (Infometrics, Inc., Woodinville, WA, USA)

Characteristic changes at water matrix coordinates

describe the water spectral pattern of lactose.

Visualization on aquagram [3]:

1344nm

1364nm

1412nm

1445nm

1475nm

1506nm

1590nm

1380nm

1400nm

1425nm

1440nm

1455nm A = Aquagram value

a = absorbance after MSC μ = average of all spectra

σ = SD of all spectra

A =a − μ

σ

0

0.5

1.0

1.5

Ab

sorb

ance

(w

ater

)

0

0.5

1.0

1.5

Ab

sorb

ance

(so

luti

on

s)

0

0.125

0.25

1.5 Ab

sorb

ance (lacto

se)

-2

-1

0

1

2n

d d

eriv

. x1

0-4

(wat

er)

2

2

-2

-1

0

1 2

nd

der

iv. x

10

-4 (s

olu

tio

ns)

2

2

-0.5

-0.25

0

0.25

0.5

0.75

1100 1300 1500 1700

1100 1300 1500 1700 1100 1300 1500 1700

1100 1300 1500 1700

11

52

14

12

14

62

13

44

17

80

11

40

1

17

5

12

08

12

76

1

30

2

13

42

1

36

6

13

80

14

50

1

49

0

15

36

15

92

1

61

6

16

46

16

96

1

72

8

17

64

1

79

6

Raw spectra 2nd derivative spectra

2n

d d

eriv. x10

-4 (lacto

se)

water lactose

Aqueous solutions of lactose

(0.02-100mM)

Signals of lactose can hardly be detected in low concentration solutions.

Subtraction of water spectrum from those of the lactose solutions reveals the spectral differences of the different concentrations.

Only water bands dominate

Regression vectors of calibration models

built on the subtracted spectra of

water and lactose solutions show stability among

replicate measurments, and

accentuate the dominant water

regions.

1-10mM 10-100mM

CONCLUSIONS

NIR technique coupled with aquaphotomics concept is useful method for quantification of the investigated carbohydrate solutes at millimolar level. Lactose causes gradual changes in the water molecular matrix, thus, performance of calibration on lactose concentration does not decrease when only the characteristic spectral region of water 1st overtone (1300-1600nm) is applied. In the contrary, it has been demonstrated in the present study that the absorption regions of water provide most useful information on the solutes in case of highly diluted aqueous solutions. Accordingly, the molecular changes of water caused by the solute can be traced and used for describing the amount of dissolved material. NIR technique and aquaphotomics based on the extended water mirror concept provide a quick and accurate alternative for classical analytical measurements even at very low (1-10mM) concentration levels.

Financial support of Japan Society for Promotion of Science (G.B. JSPS P12409) is acknowledged.

Dynamic changes of structure maker and structure breaker characteristics of sugar in

water [2]. Breaker is dominant in low, while maker in high concentrations.

H+·(H2O)4-8

OH-(H2O)1,2,4

H+·(H2O)2

OH-(H2O)1,4

O2-(H2O)4

S0

Free -OH

S0

OH-H bending

H+·(H2O)2 S1

Deionized water

OH-(H2O)4,5

2ν2+ν3

S2, S3

aν1+bν3

(a+b)=2

H+·(H2O)2,6

H+·(H2O)2,6