We developed a new illumination method called the simultaneous illumination method. This method does not require synchronization between light sources and sensor signals, which drastically simplifies the instrumentation. As a proof-of-concept, we applied this method to an oceanographic fluorometer. In principle, using this method, one can easily increase the number of characterized emission wavelengths by mounting optical sensors for as many emission wavelengths as needed. Our fluorometer has two emission-wavelength channels and twelve excitation wavelengths. The aim of this prototype is to demonstrate a viable in situ N-channel emission fluorometer with multiple wavelengths of excitation, which has not been previously realized.
© 2011 OSA
Fluorescence measurements are highly sensitive optical methods [1,2]. Therefore, they are frequently used for oceanographic studies [3–5]. Chlorophyll-a induces fluorescence near 680 nm when it is excited near 450 nm. However, it is difficult to estimate the phytoplankton concentration using fluorescence at those wavelengths because the Chlorophyll-a concentration is highly variable in phytoplankton from species to species and with differing physiological states . To overcome this problem, multi-wavelength excitation fluorometers, which measure fluorescence at one emission wavelength, i.e., 680 nm, have been developed by several companies and are available commercially. A synchronization circuit for the light source and sensor is necessary for these systems [7–9]. However, our method enables detection of not only Chlorophyll-a, but also other pigments present in phytoplankton. Therefore, our fluorometer has the potential to distinguish among different phytoplankton groups.
The fluorescence spectra of different types of matter are determined by their own excitation and emission wavelengths. Therefore, it is desirable to measure those wavelengths . In many laboratories, this is often achieved using spectral fluorometers. However, an in situ fluorometer for this purpose has not yet been realized because the conventional method for employing the excitation wavelength is to scan it electronically and detect the fluorescence signals. In this case, the signals must be synchronized with the excitation light source to distinguish the excitation wavelength. This requires complex electrical circuitry and mechanical systems because the system must both scan the emission and excitation wavelengths and synchronize the signals.
In this article, we present a new illumination method, the simultaneous illumination method (SIM) . Also, we demonstrate the simplicity and power of this method when it is applied to multi-excitation-emission fluorescence measurements in situ.
2. Principles of SIM
In the case of standard optical measurements, such as transmittance measurements , the output signal of a sample, , is proportional to the transmittance of a sample, , the intensity of the light source, , and the spectral response of the detector, . That is:
The transmittance of a sample can be obtained by dividing the output signal received by the detector with that of a reference signal. This will eliminate the dependence of the signal on both the light source intensity variation and the spectral response of the detector :
For the case of SIM, assume that the input light intensity has its amplitude modulated by a given waveform, i.e., the input light intensity varies with time. In this case, the output signal will have the same waveform as the input light intensity. For a time-varying amplitude input light waveform, the optical properties of a sample can be obtained in the same manner as above. That is:
Extending this concept, multi-wavelength lights, such as LEDs, can be used to simultaneously irradiate a sample with amplitude-modulated light at different modulation frequencies for each wavelength. The proposed technique is shown in Fig. 1 . An optical sensor detects the combined signal light composed of different wavelengths, which each have different and unique intensity modulation frequencies. This signal can be decomposed, and the individual amplitudes corresponding to each frequency of the input light can be obtained by using a frequency analysis technique, such as fast Fourier transform (FFT), because the frequencies are constant. Consequently, once the frequencies of the given wavelengths are fixed, the multi-wavelength optical properties can be determined with their amplitude via frequency analysis. In other words, we use the frequency as a marker for identifying the wavelength of the input light.
In general, square-wave modulation of the illumination is performed to eliminate ambient light. By contrast, the proposed technique uses sine-wave modulation and the ambient light does not affect the signal. Therefore, we can examine a large number of wavelengths with distinct frequencies and determine the quasi-spectral properties of the sample, because, in principle, higher harmonic waves do not appear in spectrum analysis. However, this is not true in the case of square modulation.
It is important to confirm that SIM is a reliable optical measurement method. It is known that the transmittance of a pair of linear polarizers varies as a function of the angular difference between their fast axes, i.e., a maximum while parallel and a minimum while perpendicular . We first measured the transmittance for various angular differences using the monochromatic light from an LED. These transmittance measurements were carried out at four wavelengths; 395, 471, 570, 623 nm. Then, we obtained the transmittances for the four wavelengths simultaneously using SIM for relative angles at which the transmittances had already been determined. As shown in Fig. 2 , the relationship between the transmittances measured using the former method and that obtained by SIM had a high one-to-one linearity over a wide dynamic range for the four wavelengths. Note that transmittances are normalized to one using the parallel position of the polarizers. As shown in the figure, the relationships between the amplitudes obtained through the FFT analysis and the transmittance have a high one-to-one linearity over a wide dynamic range.
3. Application of SIM to a two-channel emission fluorometer with multiple excitation wavelengths
We can place a fluorometer by inserting an optical filter between a sample and a photo-sensor (Fig. 1).
The advantages of SIM are: (1) Synchronization circuits between the light source and the signal from the sensor are unnecessary. This drastically simplifies the electronics. (2) The measurement time is independent of the number of emission wavelengths.
Advantage (1) is quite attractive for general fluorometer applications. Namely, we can easily increase the number of fluorescence measurement wavelengths by mounting as many optical sensors as needed. Advantage (2) is useful for making real-time measurements outdoors. A schematic of a submersible multi-channel emission fluorometer employing SIM is shown in Fig. 3 .
To demonstrate the effectiveness of SIM, we developed a two-channel emission fluorometer with twelve excitation wavelengths. A photograph of the two-channel emission fluorometer is shown in Fig. 4 . A cylindrical glass cell with an inner diameter of 14 mm was mounted at the center of the sample holder. The sample was excited from the bottom, and two fluorescence sensors that face each other were used to measure the fluorescence from the side of the cylindrical cell. The nine emission wavelengths of the fluorescence between 400 nm to 700 nm were measured using changing band path interference filters, which had a half band wavelength (HBW) of approximately 20 nm (Table 1 ). Figure 5 shows the spectral properties of the LEDs used as fluorescence excitation light sources. Table 2 shows the specifications of the LEDs and their corresponding modulation frequencies. The fluorescence signals were digitized using a 16-bit analog-to-digital converter, and then the frequency analysis was performed using FFT. The fluorescence measurement, which used two emission wavelengths and twelve excitation wavelengths, was performed by averaging ten FFT sets to improve the signal to noise ratio.
4. Data processing procedure
The data processing procedure is shown in Fig. 6 . Note that the sample is cultured Tetraselmis tetrathele (Prasinophyceae) phytoplankton. The upper panels in Fig. 6 show examples of the frequency spectrum of the output signals at three emission wavelengths (500, 600 and 675 nm), using twelve excitation wavelengths (395, 415, 430, 450, 465, 470, 485, 505, 525, 570, 600 and 620 nm). In the case of 500 nm emission, we can see fluorescence signals at 160, 190, 230, 280 and 310 Hz, which correspond to excitation wavelengths 395, 415, 430, 450 and 465 nm, respectively. The peaks at 350, 400, 435 and 450 Hz are caused by elastic scattering. This can be seen from the limited bandwidth of the LEDs and interference filters. However, for 600 nm emission, the fluorescence signals are much smaller than those at 500 nm. The large peaks appearing at higher frequencies, larger than 585 Hz, arise from elastic scattering. Furthermore, for 675 nm emission, except for a peak at 585 Hz, we only observed fluorescence peaks which are caused by Chlorophyll-a.
To characterize the complex relationship between the fluorescence properties of the excitation and emission wavelengths, the measured data are assembled and presented in a 3D graph (Fig. 6(d)).
Note that the contour maps presented in this article have not been calibrated with fluorescence units and are presented as output signals because the aim of this study is to demonstrate the power of applying SIM to fluorometers.
We performed measurements using seven species of cultured phytoplankton: Synechococcus sp. (Cyanophyceae), Porphyridium purpureum (Porphyridiophyceae), Chlorella kessleri (Trebouxiophyceae), T. tetrathele (Prasinophyceae), Nannochloropsis oculata (Eustigmatophyceae), Chaetoceros ceratosporum and Phaeodactylum tricornutum (Bacillariophyceae). The discrete wavelength dependences of the excitation–emission contours for each of the phytoplankton are presented in Fig. 7 .
All phytoplankton show a fluorescence peak near 700 nm when excited with a broad wavelength of light, between 400 and 450 nm. This is caused by Chlorophyll-a. Furthermore, note that C. kessleri (Fig. 7(c)) and P. tricornutum (Fig. 7(g)) have a large fluorescence in this region compared with other phytoplankton.
A small but clear fluorescence peak at 700 nm excited by 500 nm light was observed for P. purpureum (Fig. 7(b)), C. kessleri (Fig. 7(c)) and P. tricornutum (Fig. 7(g)). Synechococcus sp. (Fig. 7(a)) has a Phycocyanin fluorescence, i.e., excitation at 625 nm and emission at 700 nm. P. purpureum (Fig. 7(b)) contains Phycoerythrin, which has been detected at the emission wavelengths of 600 nm when excited at 525 nm. Note that the latest commercially available multi-wavelength excitation fluorometers cannot detect this pigment.
T. tetrathele (Fig. 7(d)), N. oculata (Fig. 7(e)) and C. ceratosporum (Fig. 7(f)) have rather simple and similar emission–excitation contour maps. However, Synechococcus sp. (Fig. 7(a)) and P. purpureum (Fig. 7(b)) have much more pronounced contour maps, which implies that they contain various pigments in their cells.
6. Discussion and conclusion
The output signal from the optical sensors based on the SIM technique has a high one-by-one linearity over a large dynamic range. Furthermore, it is applicable to optical instruments for obtaining discrete multi-wavelength optical properties in a single measurement. It can achieve this without complex electrical circuits because the light sources and optical sensors are totally independent. As an example, we developed a two-channel emission fluorometer with multi-wavelength excitation. Moreover, we are able to realize an N-channel emission fluorometer with multi-wavelength excitation by simply mounting fluorosensors without any additional electrical circuitry using SIM.
Because we use LEDs, which have a broad Gaussian spectrum, as excitation light sources, the output signals result from inelastic and elastic scattering by the phytoplankton if the difference between the excitation and emission wavelengths is small. In the present study, we handled fluorescence signals when the wavelength difference between them was larger than 25 nm. However, it is necessary to develop the correction algorithm to remove the influence of elastic scattering.
Note that our fluorometer based on SIM can detect fluorescence signals and elastic scattering simultaneously without signal saturation or underflow (Fig. 6(a)). This is because it has a large dynamic range and high linearity (see Fig. 2). This is another advantage over conventional methods.
When we apply SIM to optical measurements, the frequency of the modulated input light must be highly stable. Otherwise, a large error is obtained. Therefore, the LEDs are modulated using a direct digital synthesizer. The instability in the modulation frequency of the input light is less than 0.5 Hz.
The fluorometer developed in this study using SIM has no mechanical wavelength scanning system. Therefore, it is very robust and suitable for field use.
This research was supported by the Adaptable and Seamless Technology Transfer Program through target-driven R&D (AS2111381B), the Japan Science and Technology Agency and the Kurita Water and Environment Foundation (No. 19279). We would like to express our cordial thanks to Dr. Motoaki Kishino for encouraging this work and our gratitude to Fukada Salvage and Marine Works Co., Ltd., for cooperation in this experiment. Our thanks also go out to Dr. Niels Hojerslev for supplying valuable comments.
References and links
1. G. C. Papageorgiou, “2.Fluorescence of photosynthetic pigments in vitro and in vivo,” in Chlorophyll a Fluorescence: A Signature of Photosynthesis, G. C. Papageorgiou and Govindjee, ed. (Springer, 2004).
2. A. Grinvald, W. N. Ross, and I. Farber, “Simultaneous optical measurements of electrical activity from multiple sites on processes of cultured neurons,” Proc. Natl. Acad. Sci. U.S.A. 78(5), 3245–3249 (1981). [CrossRef] [PubMed]
3. C. S. Yentsch and D. W. Menzel, “A method for the determination of phytoplankton chlorophyll and phaeophytin by fluorescence,” Deep-Sea Res. Oceanogr. Abstr. 10(3), 221–231 (1963). [CrossRef]
4. N. K. Højerslev, “Bio-optical measurements in the Southwest Florida shelf ecosystem,” J. Cons. Cons. Int. Explor. Mer 42, 65–82 (1985).
5. N. K. Højerslev and T. Aarup, “Optical measurements on the Louisiana Shelf off the Mississippi River,” Estuar. Coast. Shelf Sci. 55(4), 599–611 (2002). [CrossRef]
6. P. G. Falkowski and J. A. Raven, Aquatic Photosynthesis (Blackwell Science, 1997)
7. BBE CO., Ltd., “Fluoro Probe,” http://www.bbe-moldaenke.de/chlorophyll/fluoroprobe/.
8. JFE Advantech CO, Ltd., “Multi-Exciter,” http://www.jfe-alec.co.jp/html/multi-exciter-e.htm.
9. L. Zhou, S. Ye, H. Chen, J. Pan, and J. Yang, “The design of a low-power self-contained multi-EX-EM array deep sea in situ organism measurement device,” The 2nd International Conference on Bioinformatics and Biomedical Engineering, ICBBE 2008, 1196–1199 (2008).
10. C. A. Parker, Photoluminescence of Solutions (Elsevier, 1968).
11. T. Oishi, A. Tanaka, S. Yano, H. Ebata, Y. Takahashi, H. Kondo, H. Tan, and R. Doerffer, “Development of simultaneous multi-wavelength excitation fluorometer,” in Ocean Optics XIX, CD-ROM (2008).
12. N. G. Jarlov, Marine Optics (Elsevier, 1976)
13. C. V. I. Melles Griot, “Fundamental Optics Guide,” http://www.cvimellesgriot.com/products/Documents/TechnicalGuide/fundamental-Optics.pdf.
15. M. Jonasz, and G. Fournier, Light Scattering by Particles in Water: Theoretical and Experimental Foundations (Elsevier, 2007).