Prediction of absorption coefficients by pulsed laser induced photoacoustic measurements

Prediction of absorption coefficients by pulsed laser induced photoacoustic measurements

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 127 (2014) 85–90 Contents lists available at ScienceDirect Spectrochimica Acta P...

952KB Sizes 3 Downloads 111 Views

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 127 (2014) 85–90

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa

Prediction of absorption coefficients by pulsed laser induced photoacoustic measurements Mallika Priya a, B.S. Satish Rao b, Satadru Ray c, K.K. Mahato a,⇑ a

Biophysics Unit, School of Life Sciences (SOLS), Manipal University, Manipal 576104, India Division of Radiobiology and Toxicology, SOLS, Manipal University, Manipal 576104, India c Department of Surgical Oncology, Kasturba Medical College, Manipal, Manipal University, Manipal 576104, India b

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 Pulsed laser induced photoacoustic

spectroscopy setup was developed.  A method for predicating absorption

coefficients using the setup is proposed.  Absorption coefficient determined for tryptophan concentrations & serum samples.  Outcomes were cross-validated with spectrophotometric measurements.

a r t i c l e

i n f o

Article history: Received 5 August 2013 Received in revised form 4 December 2013 Accepted 9 February 2014 Available online 22 February 2014 Keywords: Photoacoustic spectroscopy PAS setup Absorption coefficient Tryptophan Serum Spectrophotometry

a b s t r a c t In the current study, a pulsed laser induced photoacoustic spectroscopy setup was designed and developed, aiming its application in clinical diagnostics. The setup was optimized with carbon black samples in water and with various tryptophan concentrations at 281 nm excitations. The sensitivity of the setup was estimated by determining minimum detectable concentration of tryptophan in water at the same excitation, and was found to be 0.035 mM. The photoacoustic experiments were also performed with various tryptophan concentrations at 281 nm excitation for predicting optical absorption coefficients in them and for comparing the outcomes with the spectrophotometrically-determined absorption coefficients for the same samples. Absorption coefficients for a few serum samples, obtained from some healthy female volunteers, were also determined through photoacoustic and spectrophotometric measurements at the same excitations, which showed good agreement between them, indicating its clinical implications. Ó 2014 Elsevier B.V. All rights reserved.

Introduction Photoacoustic spectroscopy, has come a long way and established itself as an upcoming technology in the biomedical field because of its simplicity, highly sensitive and easy to handle ⇑ Corresponding author. Tel.: +91 820 2922425; fax: +91 820 2571919. E-mail address: [email protected] (K.K. Mahato). http://dx.doi.org/10.1016/j.saa.2014.02.021 1386-1425/Ó 2014 Elsevier B.V. All rights reserved.

qualities. The technique is widely being used in a variety of research areas, ranging from remote sensing for military purposes to detection of cancer, from trace element analysis in different materials to breath analysis. There is almost no field of science left where it has not been tested and found useful [1–9]. The principle behind this technique is that, when a modulated/pulsed light of specific wavelength is absorbed by a constituent of a sample, the constituent gets excited and upon de-excitation, the whole/part

86

M. Priya et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 127 (2014) 85–90

of the energy dissipates in the form of heat as non-radiative relaxations. These periodic heat fluctuations in the sample result in pressure wave or acoustic wave generation, which is then, detected using suitable pressure detectors (microphone or piezo-electric transducers). Over decades, the technique has witnessed a lot of new developments, leading to its application in tomography and imaging either in isolation or in combination with other biomedical tools such as ultrasound, sonography and microscopy [7,10– 12]. The translational significance of this technique is majorly observed in the fields of biology, biotechnology and medicine for imaging of tissues and cells, detection of tumors in circulating cells, monitoring of body vasculature etc. [13–16]. There are also reports in the literature referred under this study that show development of new algorithms and statistical tools for analyzing photoacoustic data and extracting hidden information in them [17–20]. In photoacoustic spectroscopy, the resulting signal displays absorbed energy in terms of temporal amplitudes and phases, largely dependent on factors, such as, sample’s physical properties (optical, acoustic, thermal), incident light fluence, photoacoustic cell dimensions and non-radiative relaxation processes [20–23]. The signal is the map of the amount of optical energy absorbed by the sample, and hence, it can be correlated to the product of optical absorption coefficient of the sample, and the incident light fluence [15,17,19,23]. In recent years, the usefulness of photoacoustic spectroscopy in determining optical absorption coefficients in various specimens, including biological samples, has been verified satisfactorily. The method followed for this purpose is either iterative fitting of the photon diffusion equation, employing point spread function (PSF) of the photoacoustic measurements [24] or the use of temporal amplitude and phase information of the photoacoustic signal along with the acoustic velocity within the sample [23–26]. There have been reports on work done on optical absorption coefficients for different seed genotypes (wheat, maize) [27,28] by photoacoustic measurements. Attempts have also been made to derive acoustic speed, thermal acoustic transformation coefficients in liquid samples, using photoacoustic spectroscopy [13–25,28–31]. The main idea behind the present study is to establish a facility based on pulsed laser-induced photoacoustic spectroscopy to evaluate biological variations in tissues, body fluids (serum and saliva) etc., subject to disease initiation. With this intension, we have established the setup through step-by-step development, achieving a very good sensitivity of minimum detectable concentration of 0.035 mM with tryptophan. The system has then been used successfully to determine the absorption coefficients for various tryptophan concentrations as well as for serum samples. The outcomes have later been compared with the corresponding spectrophotometric results.

described previously [32]. The cell houses a sample holder and a detector (PZT in this case) coupled together. The sample holder is an arrangement for holding the cuvette containing sample with mechanical support provided by two rectangular stainless steel blocks. In the setup, in one of the blocks, an appropriate groove was carved for partial fixing of the cuvette. The PZT transducer was housed in the second block with Teflon isolation in such a way that nearly 1 mm of the PZT was always protruding beyond its surface. The PZT casting was made up of stainless steel to minimize any spurious signals generated from stray light absorption at the cell walls. The Teflon envelope containing PZT and the metal disc was then mounted onto the cylindrical metal base with threaded outer wall to fit in with the PZT housing. The metal casing enclosing the PZT cylinder minimizes the electrical pickups. Further, to reduce the acoustic reflection back into the transducer, the metal casing was soldered with lead material. The PZT (Model PIC 181, length 10 mm, diameter 5 mm, PI Ceramics, Germany) detector was then connected to the pre-amplifier using a BNC connector, which was further coupled to a Cathode Ray Oscilloscope (Tektronix, TDS 5034B) for signal processing and recording. The laser light from the Nd-YAG laser pumped dye laser was focussed onto the PA cell containing the sample in a quartz cuvette, held in between the stainless steel blocks of the cell using a 5 cm focussing lens. The block diagram of the experimental set up is shown in Fig. 1. The photoacoustic signal generated in the sample upon laser excitation was detected by the PZT detector which, upon further amplification using the pre-amplifier, was recorded on an oscilloscope. The oscilloscope recorded the photoacoustic data in time domain. Determination of absorption coefficients by photoacoustic measurements The experimental setup used to record photoacoustic spectra in the present study was first optimized using carbon black samples dissolved in Milli Q water and recording the corresponding photoacoustic signal at 281 nm excitation. Subsequently, detection limit of the system was evaluated using tryptophan concentrations. It is an established notion that absorption coefficient of a sample derived through photoacoustic measurements is proportional to its signal amplitudes [15–17,19]. When the logarithmic profile of the first temporal amplitude in time-resolved photoacoustic measurements for a sample under study is plotted as a function of its rise-time, the slope (S1) of the best fit line to the plot represents product of acoustic velocity (m) and the absorption coefficient () for the sample [23–26,35] as shown in Eq. (1) below. Further, the acoustic velocity in a particular sample can be derived by varying the excitation-detector separation and by recording the

Methodology Experimental setup The basic components of the experimental setup consisted of excitation source, photoacoustic (PA) cell, pre-amplifier and signal processing unit. The excitation source used was a combination of Nd-YAG laser (LM1278 LPY707G-10, LITRON Lasers, UK) and a dye laser system (PULSARE Pro, FINE ADJUSTMENT, Germany) with a frequency doubling option in it. The second harmonic (532 nm) output beam from the Nd-YAG laser was used to pump the dye laser containing Rhodamin 6G dye to provide lasing in the wavelength range of 545–580 nm. The dye laser output was then frequency doubled to get the required wavelength of 281 nm used in the present study. The next component, the PA cell which is the heart of the setup was designed and fabricated in-house as

Fig. 1. Experimental block diagram of the photoacoustic spectroscopy setup.

M. Priya et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 127 (2014) 85–90

corresponding change in the first temporal amplitude of the photoacoustic signal and by plotting it for different excitationdetector separations. The reciprocal of the slope (S2) obtained from this plot provides acoustic velocity (m) in the sample as shown in Eq. (2) below [23,25,35]. This acoustic velocity can then be used to calculate the absorption coefficient by fitting it in Eq. (1).

S1 ¼ m 2

ð1Þ

S1 = slope of the plot of log (first temporal amplitude of the PA signal profile) versus rise time.

1=S2 ¼ m

ð2Þ

S2 = slope for the plot of delay time of the temporal amplitude versus excitation-detector separation. In the present study, in order to determine the absorption coefficients of various tryptophan concentrations (0.035 mM, 0.07 mM, 0.14 mM, 0.28 mM, 0.57 mM), corresponding photoacoustic spectra were recorded at 281 nm excitations, maintaining the same experimental conditions for all the measurements at 20–22 °C. Further, these concentrations were subjected to spectrophotometric measurements of absorption coefficients at 281 nm excitation using Shimadzu Spectrophotometer UV-1800, and the values were used as reference. For measuring absorption coefficients in them, the study was then extended to clinical samples involving blood samples collected from five healthy female volunteers (aged 32– 55 years) with proper informed consents as per the institutional ethical guidelines. In each of the samples under study, around 500 ll serum specimen was used for recording of corresponding photoacoustic signals at 281 nm excitations. The experiments were then conducted in a similar manner as mentioned above for determining acoustic velocity in the serum samples, and these were attempts to determine absorption coefficients in them. The results were then compared with the respective absorption coefficient values of the samples obtained using UV spectrophotometer at the same excitation. In order to avoid any contamination in the samples under study during successive spectral recordings, the cuvette used to hold the samples was thoroughly washed twice using distilled water and then with alcohol, before loading any new sample in it. Further, to determine absorption coefficients of the serum samples exploiting UV spectrophotometer, total protein concentration in the samples was utilized, and the values were calculated by applying Bradford method [36,37] with Bradford Reagent (B6916 from SIGMA–ALDRICH) and appropriate BSA (bovine serum albumin) standards. The standards (5 ll) and the unknowns were taken in duplicates in a 96-well plate. Bradford reagent (250 ll) was then added and mixed well. Readings were then taken in ELISA plate reader (TECAN Infinite M200) at 595 nm using Magellan 6 software. The Bradford method measures protein contents in a sample based on the amino acid composition in the sample. The basis of this method relies on the binding of the Coomassie brilliant blue G-250 dye to proteins, which primarily bind themselves to the aromatic amino acid residues in a protein which includes tryptophan, tyrosine and phenylalanine.

87

Results The experimental setup used to record the photoacoustic signals in the present study was first optimized using carbon black samples dissolved in Milli Q water at 281 nm excitations. The carbon black in the present study was used as a test sample because it has strong absorption at all wavelengths, thus making the evaluation process simple. Further, to ensure Milli Q water does not contribute to photoacoustic signal generation in the above procedure, corresponding experiments were also conducted using Milli Q water alone. Attempts were also made to verify whether electrical/electronic components in the setup contribute towards photoacoustic signal generation; the corresponding experiments were also performed without laser excitations, simply blocking the laser light and recording the corresponding signals. These time-domain photoacoustic spectra of Milli Q water alone with and without laser excitations as well as of carbon black in Milli Q water with laser excitation and the corresponding FFT patterns in frequency-domain are shown in Fig. 2a–c respectively. Once the performance of the setup was successfully evaluated and found optimal, limits of detection by the setup was estimated using tryptophan concentrations, and was found to be 35 lM (0.035 mM). Subsequently, the setup was used for the recording of photoacoustic patterns involving tryptophan concentrations (0.035 mM, 0.07 mM, 0.14 mM, 0.28 mM, and 0.57 mM) to determine absorption coefficients for them. Further, absorbance values for each of these tryptophan concentrations were also recorded at 281 nm excitation (Fig. 3a) using an UV-spectrophotometer and the corresponding absorption coefficients were calculated applying Beer–Lambert’s law. In photoacoustic measurements, the acoustic velocities in the respective tryptophan concentrations (0.035–0.57 mM) were derived by changing the separation between the point of laser illumination in the sample and the detector position, and by recording the corresponding change in the first temporal amplitude of the photoacoustic signal. Fig. 3b represents the plot of delay-times of the first temporal amplitude in the PA signal versus excitation-detector separations. These experiments were repeated thrice and the corresponding acoustic velocities were calculated as mentioned above. The mean velocity in each concentration was determined as listed in Table 1. The acoustic velocities for these concentrations were found to be in the range

Spectral analysis In the present study, the time domain PA spectra for all the samples were converted into frequency domain using Fast Fourier Transform (FFT) tools of the MATLAB V7.10.0 R2010a algorithms [32–34]. For further data analysis and graphical presentations Microcal Origin 6.0 has been used.

Fig. 2. Average photoacoustic (PA) patterns (left) of (a) Milli Q water with laser excitation, (b) Mili Q water without laser excitation and (c) carbon black in Milli Q water with laser excitations and the corresponding frequency patterns (right).

88

M. Priya et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 127 (2014) 85–90

Fig. 3. Plot of (a) absorbance (Mean + SEM) at 281 nm for various tryptophan concentrations (spectrophotometer), (b) excitation-detector separation versus time-delay (Mean + SEM) in PA signal generation, (c) rise-time of the 1st PA amplitude versus log of its normalized intensity (inset: a typical normalized photoacoustic signal) and (d) photoacoustic and spectrophotometrically measured absorption coefficients for different tryptophan concentrations.

Table 1 The acoustic velocities and absorption coefficients for different tryptophan (Trp) concentrations. Trp conc. (mM)

Acoustic velocity by PAS ( 103 m/s) MEAN ± SEM

0.57 0.28 0.14 0.071 0.035

5.26 ± 0.44 4.73 ± 0.81 3.64 ± 0.36 3.01 ± 0.64 2.79 ± 0.20

of 2.79 ± 0.20  103 m/s to 5.27 ± 0.44  103 m/s. Subsequently, the absorption coefficients in each of the tryptophan concentrations were calculated from the corresponding semi-logarithmic plots for the first temporal amplitude in the respective photoacoustic spectra as a linear function of time as shown in Fig. 3c. A typical normalized photoacoustic spectrum, used to extract the first temporal amplitude for plotting it in semi-logarithmic scale, is shown in the inset of Fig. 3c. The values of absorption coefficients for various tryptophan concentrations obtained with photoacoustic measurements have shown a very close agreement with those of spectrophotometric values with a correlation coefficient of 0.96 as shown in Fig. 3d and also listed in Table 1. Further, paired T test performed on the data has also shown non-significant difference (p = 0.2110, a = 0.05) between the absorption coefficient values predicted by photoacoustic and spectrophotometric measurements. The experiments were also conducted with 5 serum samples from healthy female volunteers for determining absorption coefficients in them. The absorbance values for these samples were recorded using UV spectrophotometer as shown in Fig. 4a, and the corresponding absorption coefficients were determined. Further, to

Absorption coefficient (cm1) MEAN ± SEM PAS

Spectrophotometer

4.064 ± 0.31 4.49 ± 0.73 5.13 ± 0.57 6.00 ± 1.39 6.12 ± 0.47

4.08 ± 0.22 4.76 ± 0.05 4.99 ± 0.13 6.05 ± 0.30 6.30 ± 0.47

determine absorption coefficients by photoacoustic measurements, first the acoustic velocities in the respective samples were determined from the slope of the plot of time-delay versus excitation-detector separation (Fig. 4b). This was followed by determination of the corresponding absorption coefficients in a similar way as mentioned above. Comparison of the absorption coefficient values obtained with spectrophotometric and photoacoustic measurements is shown in Table 2. The paired T test performed on the data too has shown non-significant difference (p = 0.3753, a = 0.05) between the absorption coefficient values predicted by photoacoustic and spectrophotometric measurements.

Discussion In the present study, before carrying out the actual planned experiment, normal functioning of the experimental setup was evaluated using carbon black samples dissolved in Milli Q water, and by recording the corresponding photoacoustic signals at 281 nm pulsed laser excitations. The carbon black samples yielded

89

M. Priya et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 127 (2014) 85–90

Fig. 4. Plot of (a) mean absorbance at 281 nm excitation for 5 different serum samples obtained from healthy female volunteers using spectrophotometer, (b) excitationdetector separation versus time delay in photoacoustic signal generation for 5 different serum samples.

Table 2 Mean protein concentrations, acoustic velocities and average absorption coefficients for serum samples at 281 nm excitations. Sample

Mean protein concentration (mg/ml)

Acoustic velocity by PAS ( 103 m/s) MEAN ± SEM

S1 S2 S3 S4 S5

9.40 11.31 9.35 9.29 9.50

8.28 ± 0.32 6.65 ± 0.54 7.44 ± 0.45 10.59 ± 0.62 8.37 ± 0.50

a distinct photoacoustic pattern linking to a common frequency component at 0.480 MHz in all of the repeated measurements, demonstrating optimal setup performance. Further, when photoacoustic signals of Milli Q water alone with and without laser excitations were recorded, in both the cases, no signal was detected. These observations clearly demonstrated that the signal detected with carbon black samples in Milli Q water was solely from the carbon component alone, and there was no contribution of Milli Q water in the signal. This was further confirmed by the fact that the water did not have absorption at the excitation wavelength (281 nm) used in the present study. These initial experiments have clearly demonstrated that the setup is fully optimized, and responds only when there is absorption by the sample under study at the excitation wavelength used. Following optimization, the sensitivity of the setup was also evaluated using tryptophan concentrations in water, and was found to be 35 lM (0.035 mM), which is quite encouraging as compared to the earlier reported values with benzene (1.5 mg/l) and toluene (3 mg/l) [38]. Subsequently, the experiments for predicting absorption coefficients of various tryptophan concentrations were carried out. The photoacoustic signal amplitudes in all of these experiments were found to increase, with increasing tryptophan concentrations showing concentration dependency in the photoacoustic signal generation, and indicating that these signals are arising solely from the tryptophan samples. The acoustic velocity in tryptophan was also found to be increasing with concentration similar to the case of glucose concentrations reported earlier [23], with both tryptophan and glucose solutions being dissolved absorbers. The reason behind the change in the velocity in soluble absorbers with concentration may be due to the alterations in the elastic properties of the solution upon concentration variation in them. In the present study, the alterations in the elastic properties with concentration can be due to the fact that much larger tryptophan molecules (molecular weight 204 g/mol) displace lighter water molecules

Absorption coefficient (cm1) MEAN ± SEM PAS

Spectrophotometer

2.53 ± 0.09 2.48 ± 0.19 2.73 ± 0.16 2.45 ± 0.14 2.54 ± 0.16

2.69 ± 0.00 2.24 ± 0.01 2.78 ± 0.00 2.77 ± 0.00 2.73 ± 0.01

(molecular weight 18 g/mol) differentially as per the concentrations in the solutions. As already mentioned earlier, absorption coefficients for different concentrations of tryptophan were derived utilizing semi-logarithmic plots for the 1st temporal amplitudes in the corresponding photoacoustic patterns and were found to be maintaining inverse relation with concentration. This trend of acoustic velocity and absorption coefficient values in various concentrations of tryptophan (0.035 mM to 0.57 mM) at 281 nm excitation and at 20–22 °C temperature was in accordance with the previously reported results for soluble absorbers [23,25,37,39]. This could also be explained by the basic physical phenomenon that mechanical waves need a medium to travel. At a higher concentration, the sample becomes denser wherein acoustic waves travel faster as compared to the samples with lower concentrations. These observations can also be obtained using Beer lambert’s law, which states that the absorption coefficient () and concentration (c) of a solution are inversely proportional to each other; that is when the concentration of a solution decreases, the absorption coefficient in them increases and vice versa. The absorption coefficient values for various concentrations of tryptophan with photoacoustic measurements are in close agreement (correlation coefficient = 0.96) with the spectrophotometric measurements for the same, suggesting that photoacoustic spectroscopy is equally capable of determining actual absorption coefficients of different types of samples. The measurement error in the calculation of absorption coefficients for various concentrations of tryptophan was found to be less than 10%. Ensuring correct prediction of absorption coefficients in various concentrations of tryptophan, the system was further extended to clinical samples for predicting corresponding absorption coefficients in them. The absorption coefficient values for the clinical samples obtained with photoacoustic measurements too were in close agreement with the spectrophotometric outcomes for the same. The measurement error in this case was found to be less

90

M. Priya et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 127 (2014) 85–90

than 11.5%. The absorption coefficients for the samples under photoacoustic measurements were determined based on the amplitude and phase information of the photoacoustic signal as a function of wavelength. All the experiments in the present study were repeated thrice and mean values were used to determine the absorption coefficient in each. There have been reports in the literature showing application of time resolved photoacoustic measurements, in predicting absorption coefficients along with other parameters for clear liquid samples/dissolved absorbers [23,25,37], colloidal samples/mixtures [23,25,37], biological samples and also for layered samples like tissues [39–44] etc. However, that there are no reports in the literature for predicting absorption coefficients of clinical samples by photoacoustic measurements is probably the main strength of the present study, which can of course be strengthen further by including greater number of samples in the study and evaluating the system performance. Conclusion In summary, in the present study, a pulsed laser induced photoacoustic spectroscopy setup was designed and developed, which could correctly measure the absorption coefficients of the tryptophan concentrations and the serum samples of healthy female volunteers. As per our knowledge, this is probably the first attempt towards predicting absorption coefficient of clinical samples by photoacoustic measurements. The added advantage with the photoacoustic spectroscopy in predicting absorption coefficient of a sample or a mixture of samples is that it is based on direct absorption of radiation by the samples under study, unlike the spectrophotometric measurements which are based on indirect absorption by the samples. One of the biggest advantage of photoacoustic technique in comparison with the conventional techniques is that it is unaffected by the scattering and reflections losses that take place due to the nature of the samples. Therefore, this technique may be suitable for measuring absorption coefficients in biological samples. Acknowledgements Authors would like to acknowledge the patronage of Manipal University, Manipal; TIFAC-CORE in Pharmacogenomics at SOLS, Manipal University, and Vision Group of Science & Technology (VGST), Karnataka State. We would also like to thank Professor K. Satyamoothy, Director, SOLS, Manipal University, Manipal for his constant encouragement and support and Mr. Subhash Chandra for his technical assistance. One of the authors, Ms. Mallika Priya acknowledges the support of Manipal University in providing research fellowship under its structural Ph.D. program.

References [1] A. Rosencwaig, Photoacoustics and Photoacoustic Spectroscopy, Wiley and Sons, New York, 1980. [2] A. Rosencwaig, Annu. Rev. Biophys. Bioeng. 9 (1980) 31–54. [3] J.B. Kinney, R.H. Staley, Annu. Rev. Mater. Sci. 12 (1982) 295–321. [4] S.D. Kamath, S. Ray, K.K. Mahato, J. Biomed. Opt. 16 (2011) 1–10, 067001. [5] S.D. Kamath, R.A. Bhat, S. Ray, K.K. Mahato, Photomed. Laser Surg. 27 (2009) 325–335. [6] C.S. Sunandana, Phys. Stat. Sol. (a) 105 (1988) 11–43. [7] V.S. Chivukula, M.S. Shur, D. Ciplys, Phys. Stat. Sol. (a) 204 (2007) 3209–3236. [8] M.R. Bugs, R.K. Bortoleto-Bugs, M.L. Cornélio, R.J. Ward, Biochem. Biophys. Res. Commun. 353 (2007) 889–894. [9] A.J. Sadler, J.G. Horsch, E.Q. Lawson, D. Harmatz, D.T. Brandau, C.R. Middaugh, Anal. Biochem. 138 (1984) 44–51. [10] C. Li, L.V. Wang, Phys. Med. Biol. 54 (2009) R59–R97. [11] J. Yao, L.V. Wang, Contrast Media Mol. Imaging 6 (2011) 332–345. [12] K. Narayanan, S. Chandani, T. Ramakrishna, C.M. Rao, Biophys J. 72 (1997) 2365–2368. [13] L. Fass, Mol. Oncol. 2 (2008) 115–152. [14] Y. Zhang, H. Hong, W. Cai, Photoacoustic imaging, Cold Spring Harb. Protoc. 9 (2011) 1015–1025. [15] G. Xu, I.A. Dar, C. Tao, X. Liu, C.X. Deng, X. Wang, Appl. Phys. Lett. 101 (221102) (2012) 1–5. [16] R.P. Solano, F.I. Ramirez-Perez, J.A. Castorena-Gonzalez, E.A. Anell, G. Gutiérrez-Juárez, et al., AIP Adv. 2 (2012) 1–15, 011102. [17] J. Zalev, M.C. Kolios, Proc. SPIE 7899 (2011) 1–15, 78992M. [18] X.L. Dean-Ben, V. Ntziachristos, D. Razansky, Appl. Phys. Lett. 98 (2011) 1–3. [19] J.A. Swearingen, S.H. Holan, M.M. Feldman, J.A. Viator, J. Biomed. Opt. 15 (2010) 016019-1–016019-9. [20] M. Malin´ski, L. Bychto, S. Legowski, J. Szatkowski, J. Zakrzewski, Microelectron. J. 32 (2001) 903–910. [21] Z. Guo, S. Hu, L. Wang, Opt. Lett. 35 (2010) 2067–2069. [22] Y. Wang, R. Wang, Proc. SPIE 6856 (2008) 1–7, 685628. [23] Y. Shen, Z. Lu, S. Spiers, H. MacKenzie, H. Ashton, J. Hannigan, S. Freeborn, J. Lindberg, Appl. Opt. 39 (2000) 4007–4012. [24] F.J.M. Harren, J.J. Laarhoven, Photoacoustic Spectroscopy, The Optics Encyclopedia, 2007.. [25] Z. Zhao, R. Myllyla, Appl. Opt. 51 (2012) 1061–1066. [26] P.R. Crippa, A. Vecli, C. Viappiani, J. Photochem. Photobio. B: Biol. 24 (1994) 3–15. [27] H.A. Claudia, C.O. Alfredo, I. Rumen, A. Dominguez, A. Carballo, M. Ivan, R. Ricardo, Food Biophys. 64 (2011) 481–486. [28] Y. Wang, R. Wang, Phys. Med. Biol. 53 (2008) 6167–6177. [29] A. Mandelis, J. Appl. Phys. 54 (1983) 3404–3409. [30] C.G.A. Hoelen, F.F.M. de Mul, J. Acoust. Soc. Am. 106 (1999) 695–706. [31] Z. Li, H. Li, Z. Zeng, W. Xie, W.R. Chen, J. Biomed. Opt. 17 (2012) 1–6, 061216. [32] S.D. Kamath, V.B. Kartha, K.K. Mahato, Spectrochim. Acta A Mol. Biomol. Spectrosc. 70 (2008) 187–194. [33] R. Pratap, Getting Started with MATLAB: A Quick Introduction for Scientists and Engineers, Oxford University Press, USA, 2009. [34] http://www.mathworks.in. [35] Y. Shen, S. Spiers, H.A. MacKenzie, Anal. Sci. 17icpp (2001) s221–s222. [36] C.N. Pace, F. Vajdos, L. Fee, G. Grimsley, T. Gray, Protein Sci. 4 (1995) 2411– 2423. [37] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254. [38] V. Hanyecz, A. Mohacsi, S. Puskas, A. Vago, G. Szabo, Meas. Sci. Technol. 22 (2011) 125602. [39] Z. Zhao, Doctoral thesis Acta Universitatis Ouluensis, Series C, 2002, pp.169. [40] P.C. Beard, T.N. Mills, Phys. Med. Biol. 42 (1997) 177–198. [41] A. Mandelis, Y.C. Teng, B.S.H. Royce, J. Appl. Phys. 50 (1979) 7138–7146. [42] S. Telenkov, A. Mandelis, Rev. Sci. Instrum. 81 (2010) 1–7, 124901. [43] G.A. Holzapfel, G. Sommer, C.T. Gasser, P. Regitnig, Am. J. Physiol. Heart Circ. Physiol. 289 (2005) H2048–H2058. [44] J.R. Rajian, P.L. Carson, X. Wang, Opt. Express 17 (2009) 4879–4889.