Asian Journal of Physics - Special issue to be published.

Fiber Optic Chemical Sensor for Air Pollutant Measurement Design, Development and Applications

Ang Soo Seng, M. S. John and Anand Asundi School of Mechanical and Production Engineering Nanyang Technological University, Singapore 639798 Email: masundi@ntu.edu.sg

ABSTRACT
In this paper, the development of a sol gel encapsulated fiber optic chemical sensor to detect the presence and concentration of sulphur dioxide (SO2) and nitrogen dioxide (NO2), two major air pollutants are discussed. The sensing chemical, rhodamine 6G a fluorescence dye which absorbs light at 530nm and emits radiation at 560nm, is used to detect SO2 while a combination of sulfanilamide (SFA) and N, N Dimethyl-1-naphthalamine (DMNA), which absorb light at 450nm, is used to detect NO2. The sensing chemical used is exclusive to individual gases and responds to the specific gas. The chemical reaction between the gas and the sensing chemical changes the transmitted light intensity, which is detected by a spectrometer. The presence and concentration of the gas is determined by the change in light intensity and the rate of change in light intensity, respectively. Two probes are developed an intrinsic fiber optic chemical sensor made up of a sol gel encapsulated chemical coated optical fiber and an extrinsic fiber sensor which uses a sol gel encapsulated chemical coated glass plate as the sensing element. Experiments show that the intrinsic sensor is better in fluorescence quenching test while the extrinsic performs better in light absorption experiments.

-1-

Asian Journal of Physics - Special issue to be published.

1. INTRODUCTION
Sol-gel is a process that enables materials to mix on a molecular level from a sol into a colloidal gel. The moisture in the colloidal gel is removed and after drying the gel will become a porous solid. The porosity of the solid depends on the catalyst used in the mixing process, an acidic catalyst produce fast hydrolysis leading to weakly branched microporous structures while basic catalyst produce slow hydrolysis leading to highly branched structures with large pore size. This process is used to coat the liquid chemical onto a sample surface (optical fiber or glassplate) then after drying the solidified chemical will be immobilized on the sample surface. [1-3]

Entrapping a suitable chemical in the sol-gel forms a chemical sensor. The chemical reacts with the specific gas and changes the spectral characteristics of the transmitting light. In this paper intrinsic and extrinsic fiber optic chemical sensors are developed for detection and sensing. For the detection of sulphur dioxide gas, the principle of fluorescence quenching of rhodamine 6G entrapped in sol gel matrix is used. Similarly, the detection of nitrogen dioxide gas uses the principle of light absorption of sulfanilamide (SFA) and N, N Dimethyl-1-naphthalamine (DMNA), entrapped in sol gel matrix. Since the chemicals are gas specific multi-channel chemical sensing is also possible.

2. EXPERIMENTAL DETAILS

In light absorption experiments, the Lambert-Beer law is used to measure the intensity of absorption. If I0 is the intensity of radiation incident normally on an absorbing material layer of thickness b cm and concentration c then the intensity of the emergent beam I is, [4,5]

I = I 0 10 cb

(1)

-2-

Asian Journal of Physics - Special issue to be published.

I log 10 0 = cb = A I

(2)

where is the molar absorptivity and A is the absorbance of the sample in the beam. There is a linear relationship between the absorbance A and concentration c of a given solution if the optical path length and the wavelength of the light signal are kept constant.

Fluorescence quenching is the decrease in fluorescence intensity in the presence of a particular substance or gas. The dynamic quenching of fluorescence is described by the Stern-Volmer equation, [6]

F0 = 1 + k q 0 [Q ] = 1 + K D [Q ] F

(3)

KD = kq0

(4)

where F0 is the fluorescence intensity in the absence of quencher, F is the fluorescence intensity in the presence of quencher, kq is the bimolecular quenching constant, 0 is the lifetime of the fluorophore in the absence of quencher, [Q] is the concentration of quencher, KD is the Stern-Volmer quenching constant. A linear Stern-Volmer plot usually indicates that a single class of fluorophores is accessible to the quencher. If two classes of fluorophores are present and one is not accessible then the Stern-Volmer plot deviates from linearity toward the x-axis. 2.1 Preparation of Sensor

This process describes the preparation of the intrinsic sensor. A 200 m diameter chemically coated optical fiber sensor is prepared by first stripping away, from a 4 cm length, the plastic coating from -3-

Asian Journal of Physics - Special issue to be published. the optical fiber. The cladding of the optical fiber is then etched for 1 hour in 50% hydrofluoric acid to expose the core region. The fiber core is immersed in NHO3 solution for 5 minutes to enhance the chemical coating process. The rhodamine 6G solution is prepared by dissolving solid rhodamine 6G with ethanol into a 0.001M solution. The combination of sulfanilamide (SFA) and N, N Dimethyl-1naphthalamine (DMNA) solution is prepared by dissolving SFA and DMNA with acetic acid into a 0.01M solution. The sol gel mix is prepared by adding 12 ml of Tetraethyl orthosilicate (TEOS) to 2.8 ml of de-ionised water together with 1 ml of ethanol and stirred. After 20min, 1 ml of rhodamine 6G solution or sulfanilamide (SFA) and N, N Dimethyl-1-naphthalamine (DMNA) solution and 40 l of 37% hydrochloric acid are added and stirred continuously until mixture turns into a sol. The sol is aged at 24C for 40 hours for rhodamine 6G film or 10 hours for SFA-DMNA film until it becomes viscous. The prepared fiber core is dip-coated in the viscous solution. The coated fiber core is left to solidify for 3 days and become glassy. A schematic of the intrinsic Fiber Optic Chemical Sensor (FOCS) probe is shown in Figure 1(a).

(a)

(b)

Figure 1. (a) Schematic of the intrinsic FOCS probe and (b) extrinsic FOCS probe -4-

Asian Journal of Physics - Special issue to be published.

For ease of manufacture, a glass plate sensor coupled to a fiber optic illumination and receiving light probe was also developed. Since the fiber was used purely for delivery and receiving of the light so this configuration constitutes an extrinsic FOCS. The process for making this probe is the same as that described above except that a microscopic glass plate is coated with the chemical encapsulated sol-gel film. The schematic of this sensor is shown in Figure 1(b). Thus the first probe is referred to as an intrinsic Fiber Optic Chemical Sensor (FOCS) and the second is an extrinsic FOCS.

2.2 Experimental Setup

The schematic of the experimental setup is shown in Figure 2(a). A blue light emitting diode (wavelength 455nm) is used for sensing NO2 gas while a green laser diode (wavelength 532nm) is used to sense SO2 gas. Light is coupled into the 200m (core diameter) chemically prepared stepindex multi-mode optical fiber via a 40X objective lens. The probe is contained in a stainless steel gas chamber and the signal is detected using spectrometer. Figure 2(b) shows the typical spectrum obtained from the NO2 sensor at different times after exposure to the gas. The reduction in peak intensity is a measure of the absorbency and is used for later calculation.

-5-

Asian Journal of Physics - Special issue to be published.

NO2 concentration - 232.7ppm

2200

2000

1800

0 min 2 min 4 min 6 min 8 min 10 min

Intensity

1600

1400

1200

1000 430 440 450 460 470 480 490

Wavelength (nm)

(a)

(b)

Figure 2. (a) Schematic of experimental setup for intrinsic fiber optic chemical sensor (b) Typical spectrum from the NO2 sensor

3. RESULTS
3.1 Results for sensing NO2 using extrinsic fiber optic chemical sensor

The gas pressure in the chamber was varied from 0.06bar to 0.12bar. The following parameters are kept constant: the wavelength of the light source (455.88nm), the integration time which is the time interval for the spectrometer to register the data (110ms), the sensing chemical concentration (0.01Molar) and the data acquisition interval (60s). Figure 3 shows that the rate of decrease of light intensity increases as the NO2 gas concentration increases.

-6-

Asian Journal of Physics - Special issue to be published.

NO2 concentration - 181.7ppm
1.0

NO2 concentration - 205.7ppm wavelength at 455.8nm

wavelength at 455.8nm

1.0

0.9

0.8

Normalised Intensity

Normalised Intensity

Y1=A+B*X A 0.990 -2 B -4.96x10

0.9

Y1=A+B*X A 0.965 -2 B -5.81x10

0.8

0.7

0.7

0.6

Y2=A+B*X A 0.777 -3 B -7.01x10

0 2 4 6 8 10 12 14 16 18 20

Y2=A+B*X A 0.648 -3 B -2.74x10

0.6

0.5

0.5 0 2 4 6 8 10 12 14 16 18 20

Time (min) NO2 concentration - 232.7ppm

1.0

Time (min)

wavelength at 455.88nm

0.9

Normalised Intensity

0.8

Y1=A+B*X A 0.994 -2 B -7.42x10

0.7

0.6

Y2=A+B*X A 0.666 -3 B -9.95x10

0.5 0 2 4 6 8 10 12 14 16 18 20

Time (min)

Figure 3. Change in intensity as a function of time for different NO2 concentrations using the extrinsic FOCS

3.2 Results for sensing NO2 using intrinsic fiber optic chemical sensor

All parameters are the same as for the previous experiment, except that the integration time is set to 20 ms and the data acquisition interval is reduced to 30 s. Figure 4 again shows that the rate of decrease of light intensity increases as the NO2 gas concentration increases.

-7-

Asian Journal of Physics - Special issue to be published.

NO2 concentration - 181.7ppm

1.05

NO2 concentration - 205.7ppm

1.0

wavelength at 455.88nm
1.00

wavelength at 455.88nm

0.95

Normalised Intensity

0.90

Normalised Intensity

Y1=A+B*X A 0.97596 B -0.01094

0.9

0.8

Y1=A+B*X A 0.94009 B -0.02333 Y2=A+B*X A 0.68292 B -0.00286

0.85

0.80

Y2=A+B*X A 0.7908 B -0.00133

0.7

0.6 0.75

0.70 0 5 10 15 20 25 30 35 40 45

0.5 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Time (min)

Time (min)

1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0 2 4

NO2 concentration - 232.7ppm wavelength at 455.88nm

Normalised Intensity

Y1=A+B*X A 1.03059 B -0.04499

Y2=A+B*X A 0.56993 B -0.00467

10

12

14

16

18

20

Time (min)

Figure 4. Change in intensity as a function of time for different NO2 concentrations using the intrinsic FOCS

3.3 Results for sensing SO2 using extrinsic fiber optic chemical sensor

The fluorescence signal from rhodamine 6G in the plate fiber probe sensor is very weak for the extrinsic FOCS and thus the data is not reliable. The sensor was not used further although improvements to the sensitive are being investigated.

-8-

Asian Journal of Physics - Special issue to be published. 3.4 Results for sensing SO2 using intrinsic fiber optic chemical sensor

The gas pressure ranges from 0.1bar, 0.2bar, 0.3bar and 0.4bar. The following factors are kept constant, the wavelength of the light source at 548.59nm, the integration time at 1200ms, the chemical concentration at 0.001Molar, the time interval at 30s. Figure 5 shows that fluorescence quenching takes place as the SO2 concentration increases.

SO2 concentration - 207.0ppm

1.00 0.99 0.98

SO2 concentration - 280.2ppm

1.00 0.99 0.98

548.59nm

548.59nm

Normalised Intensity

Normalised Intensity

0.97 0.96 0.95 0.94 0.93 0.92 0 60

Y=A+B*X A 1.00184 -5 B -8.8322x10

0.97 0.96 0.95 0.94 0.93 0.92

Y=A+B*X A 1.00636 -4 B -1.5986x10

120

180

240

300

360

60

120

180

240

300

360

Time (sec)

Time (sec)

SO2 concentration - 342.1ppm

1.00 0.99 0.98

SO2 concentration - 395.3ppm

548.59nm

1.00 0.99 0.98

548.59nm

Normalised Intensity

Normalised Intensity

0.97 0.96 0.95 0.94 0.93 0.92 0 60 120

Y=A+B*X A 0.98831 -4 B -2.0658x10

0.97 0.96 0.95 0.94 0.93 0.92

Y=A+B*X A 1.00716 -4 B -2.5318x10

180

240

300

360

60

120

180

240

300

360

Time (sec)

Time (sec)

Figure 5. Fluorescence quenching as function of time for different SO2 concentrations using the intrinsic FOCS

-9-

Asian Journal of Physics - Special issue to be published.

4. DISCUSSION

4.1 NO2 experiment

Both the intrinsic and extrinsic FOCS showed a decrease in light intensity when exposed to NO2 gas. When the SFA and DMNA chemical coating reacts with the NO2 gas, the absorbance of the chemical increases leading to a reduction in light intensity reaching the detector.

Gradient
-5.0

Absorption gradient for intrinsic FOCS Gradient

-1.0 -1.5

Absorption gradient for extrinsic FOCS

-5.5

-2 Rate of change = -4.85x10

-2 Rate of change = -6.71x10

-2.0

-6.0

-2.5 -3.0

-6.5
-3.5

-7.0

-4.0 -4.5 -5.0

-7.5 180 190 200 210 220 230 240

180

190

200

210

220

230

240

NO2 concentration (ppm)

Gas concentration (ppm)

Figure 6. Absorption gradient for (a) intrinsic and (b) extrinsic NO2 FOCS

Figure 6 shows the light absorption gradient for the intrinsic and extrinsic NO2 FOCS. It is noticed that the gradient extrinsic FOCS is greater than that for intrinsic FOCS. Comparing the two gradients it is noticed that the extrinsic FOCS is 40% more sensitive than the intrinsic FOCS. The chemical coating process might be the cause for this as the flat surface of the glass plate is easier to coat than the cylindrical surface of the optical fiber. Further the coating on the glass plate can be made thicker than that on the cylindrical optical fiber. Since absorbance depends on the thickness of the chemical coating as well, the glass plate probe sensor is more effective in absorbing the light. There are several factors to consider when fabricating an effective absorption sensor like the concentration of the SFA and DMNA chemical coating, the thickness of the chemical coating, the surface roughness -10-

Asian Journal of Physics - Special issue to be published. of the chemical coating. The preparation of any chemicals, the chemical coating process and drying process are very important in order to fabricate a sensitive chemical coated sensor. Besides this, the distance between the probe and the chemical coating has to be optimized to achieve the best position.

4.2 SO2 experiment

When green light excites the rhodamine 6G dye, the chemical coating will fluorescence and emit light at the wavelength of 548.59nm. As the coating is exposed to SO2 gas, fluorescence quenching occurs and there is a drop in fluorescence quenched light intensity. Figure 7(a) plots the normalised intensity as a function of time for different concentration. It is seen that there is a reduction in intensity as a function of time and this gradient decreases with concentration.
Y=A+B*X for chemical coated fiber sensor
1.10 1.04

Stern-Volmer graph

1.02 1.00 0.98 0.96

Gradient B -2.53x10-4 at 395.3ppm Gradient B -2.07x10-2 at 342.1ppm Gradient B -1.60x10-4 at 280.2ppm Gradient B -8.83x10-5 at 207.0ppm Intensity ratio (F0/F)

1.09

1.08

Normalised Intensity

1.07

1.06

1.05

Y=A+B*X A 0.96199 -4 B 3.40x10

0.94 0.92

1.04

0.90 0 60 120 180 240 300 360 420

1.03 200 250 300

Concentration (ppm)
350 400

Time (sec)

Concentration (ppm)

Figure 7. (a) Quenched intensity as function of time for different concentrations (b) Stern-Volmer graph

-11-

Asian Journal of Physics - Special issue to be published. The Stern-Volmer graph is plotted by taking F0/F (time at 0s / time at 400s) against gas concentration and is shown in fig. 7(b). The quenching efficiency [6] is

kq = ko.

(5)

From Figure 7, the gradient of the lines gives the constant, KD= 3.4x10-4M-1. With 0 3.7ns [7], the bimolecular quenching constant, kq from the Stern-Volmer equation is

KD = kq0

Hence, kq = K D 3.4 x10 4 = = 9.12 x10 4 M-1s-1 9 0 3.7 x10

The diffusioncontrolled bimolecular rate constant (k0) can be found by solving the Smoluchowski equation [6] in order to calculate the quenching efficiency,

k0 =

4N (R f + Rq )(D f + Dq ) 1000

(6)

where Rf is the molecular radius of rhodamine 6G, Rq is the molecular radius of SO2, Df is the diffusion coefficient of rhodamine 6G, Dq is the diffusion coefficient of SO2 and N is the Avogadros number. The molecular radius is given as [8] r V
1 3

(7)

-12-

Asian Journal of Physics - Special issue to be published. where r is the molecular radius and V is the volume of the molecule. The molecular radius is approximately equal to the cube root of the volume of molecule. The diffusion coefficient of SO2 can be obtained from Stokes-Einstein equation [6],

Dq = kT / 6R

(8)

where k is the Boltzmann constant, is the solvent viscosity, R is the molecular radius and T is the temperature. Solving the diffusioncontrolled bimolecular rate constant gives k0 =3.45x105M-1s-1. Thus the quenching efficiency is as follows:

kq k0

9.12 x10 4 = 0.26 3.45 x10 5

Thus only 26% of the molecular collisions between the rhodamine 6G and SO2 are effective in fluorescence quenching. The reasons that might affect the quenching efficiency are the concentration of the rhodamine 6G chemical coating, the thickness of the chemical coating, the surface roughness of the chemical coating. The chemical coating process and drying process are also very important in improving the efficiency of the fluorescence quenching. This could be one of the reasons that the extrinsic FOCS is not sensitive enough. Furthermore for the intrinsic FOCS the fluorescence is integrated over the length of the fiber, while for the extrinsic FOCS, the fluorescence is confined to a small area and thus the signal is weaker.

5. CONCLUSION

Fiber Optic Chemical Sensors (FOCS) for detecting NO2 and SO2 have been designed, developed and tested. Two sensing configurations have also been investigated for each of these sensors: -13-

Asian Journal of Physics - Special issue to be published. (1) Chemical coated fiber sensor (intrinsic FOCS) (2) Glass plate probe sensor (extrinsic FOCS). The extrinsic FOCS is found to be better suited for absorption experiments as the thickness of the coating can be large as well more uniform. However for fluorescence measurements, since the intensity is weak, the intrinsic FOCS, with a larger sensing area, gives a stronger signal than the extrinsic FOCS. The preparation of the chemical coating is also crucial in the design of an effective gas sensor. Miscalculation in the chemical proportion or incorrect mixing of chemicals will inhibit the sol from setting. Besides this, the concentration of the sensing chemicals also affects the sensitivity of the gas sensor.

6. REFERENCES

1.

C. McDonagh, B. D. MacCraith, A. K. McEvoy, Tailoring of Sol Gel Films for Optical Sensing of Oxygen in Gas and Aqueous Phase, Anal. Chem. 1998, 70, pp45-50

2.

R.W. Jones, Fundamental Principles of Sol-Gel Technology, Institute of Metals (1988) Lisa C. Klein, Sol-Gel optics: Processing and Applications, Kluwer Academic Publishers (1994)

3.

4.

C. Burgess and A. Knowles, Standards in Absorption Spectroscopy, Chapman and Hall D. A. Krohn, Fiber optic sensors - Fundamentals and Application, Instrument Society of American (1988)

5.

6. 7.

Lakowicz Joseph R., Principles of Fluorescence Spectroscopy, Plenum Press (1983) Lifetime decay of Rhodamine 6G with two-photon excitation, http://www.mi.infm.it/~biolab/tpe/fh6g.htm

8.

Robert H. Perry and Don W. Green, Perrys Engineers Handbook (7th edition), Mcgraw Hill, pp2-230, pp5-50

-14-