./orrnld ofhod Er1girwring 30 (1YYh) zyxwvutsrqponmlkjihgfe I 555162 Copyright CZ lYY6 Elscvier Scicncc Lmlrcd Printed in Great Britain. All righta rcscrved 0x0-X774/06 $1.5.00+ (l.O(l Pll:SO260-8774(96)00017-9 Photopyroelectric (PPE) Measurement of Thermal Parameters in Food Products D. Dadarlat,” J. Gibkes,” D. Bicanic” & A. Pasca,” ” institute of Isotopic and Molecular Technology, POB 700, Cluj-Napoca 5, R34OC)Romania “Wageningen Agricultural University, Department of Agricultural Engineering and Physics. Agrotechnion, Bomenweg 4, 6703 HD Wageningen, The Netherlands (Received 20 April 1995; accepted I1 January 19%) zyxwvutsrqponmlkjihgfedcba ABSTR4 CT The stundard (SPPE) und inverse (IPPE) photopyroelectric configurations were combined in order to obtuin room temperature values of the thermal pummeters for vurious types of food products. The SPPE conjiguration (with thermalLy thick sample and sensor) allows the direct and absolute measurement qf the thermal diffusivity. The thermal efiusivity was obtained in the inverse configuration (with thermally thin and optically opaque sensor and thermally thick sample). The static (volume specific heat) and dynamic (thermal conductivity, diffusivity und effusiviv) thermal parameters for a variety of liquid und push’ products such as oils, margarine, butter; creums, musturd. ketchrip and juices are reported. Copyright 0 1996 Elsel’ier Science Limited INTRODUCTION In recent years, the photopyroelectric (PPE) method has proved to be a useful tool for optical and thermal characterization of materials (Bicanic, 1992). On the other hand, studies of thermophysical properties of foodstuffs represent the subject of intensive research effort carried out both in industry and academic institutions. Once known and their behavior understood, such properties can be used to benefit industrial manufacturers by improving specific applications, as well as to control the stability and the shelf-life of the product. Due to their composition and intrinsic inhomogeneity foods are generally difficult to analyze thermally. Recently the PPE method, in various configurations, was proposed as a new approach to thermal characterization of fatty acids and triglycerides (Dadarlat et al., 1995a, 1995b). Essentially, the PPE technique is concerned with the detection (via a pyroelectric sensor) of the heat developed in a sample exposed to modulated radiation (Mandelis & Zver, 1985; Chirtoc & Mihailescu, 1989). The result of a PPE 155 156 D. Dadarlat et al measurement, the PPE voltage, is a function of two or in some particular cases one of the sample’s related thermal parameters (Dadarlat et al., 1995a). It is well known that the four thermal parameters (volume specific heat, C, thermal conductivity, k, diffusivity, c(, and effusivity, e) are connected by two relationships and, consequently, only two are independent [see eqn (3)]. For the purpose of obtaining all four thermal parameters for a given sample, usually one measures a dynamic thermal parameter and uses literature values for the volume specific heat (in fact values of the density and mass specific heat). Unfortunately, for many foods, values for the volume specific heat are not available. For such products, a method capable of directly measuring two thermal parameters is necessary. In this paper it is proposed to combine two PPE configurations in order to measure the thermal diffusivity and effusivity and to use these values to calculate the remaining thermal parameters. In such a way, the method is self-consistent, being independent of other methods and/or literature values. The usability of the proposed approach is demonstrated with studies on food specimens for which the value of the volume specific heat is not known. THEORY Standard configuration In the standard configuration the radiation impinges on the front surface of the sample; the pyroelectric sensor glued to the sample’s rear side measures its temperature variation. Marinelli et al. (1992) demonstrated that, in the standard configuration, when the sample and the sensor are both thermally thick (the geometrical thickness, L, is much larger than the thermal diffusion length, ~1= (~x/o)]‘~, in the material), the phase of the complex PPE signal is given by: (1) where (I) is the angular chopping frequency of the radiation and the subscript ‘s’ refers to the sample. Equation (1) indicates that, for a calibrated cell (i.e. L, is precisely known), a frequency scan of the phase of the PPE signal provides the direct and absolute measurement of the sample’s thermal diffusivity. Inverse configuration In the inverse configuration, the modulated radiation is absorbed by the front surface of the pyroelectric sensor, the sample (in good thermal contact with the sensor’s rear side) acting as a heat sink. Dadarlat (Dadarlat & Frandas, 1993; Dadarlat et al., 1995) showed that, in this configuration, when the sensor is thermally thin (Lp <pp; .p3 referring to the pyroelectric sensor) and optically opaque, and the sample is thermally thick (L, $ pJ, the amplitude of the PPE signal is given by: Photopyroelectric measurement of thermul parumeters v=!x 157 (2) es where V,, is a calibration factor depending on the radiation intensity, on the characteristics of the pyroelectric sensor and the electrical circuit used for signal processing (Dadarlat et al., 1995a). Equation (2) indicates that, provided an appropriate calibration was completed, the direct measurement of the sample’s thermal effusivity is possible. The remaining two thermal parameters can be derived using the well known relationships: k = ,(,)‘I2 and C = e(x) “’ (3) EXPERIMENT The experimental set-up and the PPE cells, both for standard and inverse configurations, were presented elsewhere (Dadarlat et zyxwvutsrqponmlkjihgfedcbaZYXWVUTSR al., 1995a, 1995~). Only some details are given here. The pyroelectric sensors consisted of a 300-Lirn thick LiTa03 single crystal in the standard configuration and a 9-pm thick polyvinylidene difluoride (PVDF) foil in the inverse geometry. The signal from the detectors was processed with a SR850 model Stanford Research lock-in amplifier. The radiation source was a 31) mW Melles Griot diode laser, the radiation of which was electronically chopped by the internal oscillator of the lock-in. For the standard configuration a calibrated cell was designed (Fig. 1). The thickness of the cell was measured by two methods: (i) using a microscope; and (ii) by performing a frequency scan with water as reference sample (Fig. 2). Using the literature value of the thermal diffusivity of water (x = 1.42 x lo-’ m’is) in eqn (1) one obtains the thickness of the cell. Both methods lead to a thickness of 570 ( +5) itm, considered constant for all the investigated samples. Radiation GlilSS rings Fig. 1. Calibrated cell used in the standard configuration. D. Dadarlat et al. 158 The frequency range for the scan was 0.05-l Hz. Figure 2 indicates that (for water), starting at about 0.5 Hz, the conditions for a thermally thick regime for the sample and sensor are fulfilled. Due to the fact that the composition of the samples under investigation is based mainly on water or fats, we expect a similar frequency range for them. In the inverse configuration the thickness of the sample exceeded 3 mm. A chopping frequency of 0.1 Hz assures the validity of eqn (2) with an accuracy of about 2% (Dadarlat et al., 1995). The measurement was calibrated with water again (f&MU = 1600 W.s”2/m2.K). After each measurement the sensor was cleaned with water and chloroform; the procedure was repeated until the signal obtained with the empty sensor was always the same within 0.2%. In both configurations the typical signal to noise ratio was better than 1000. The investigated samples included: different types of oils (soya, sunflower, olive, pumpkin, motor oils), margarine, butter, peanut butter, apple juice, mustard, sour cream and ketchup. No data about the volume specific heat of these products were found in the literature. The following data about the origins and composition of the investigated samples are available: Soya zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA oil: produced by Albert Heijn, The Netherlands; name of product: Slaolie; typical composition: saturated fatty acids (C16:0, Cl&O) 14%, unsaturated (Cl&l) 21%, (C18:2) 56%, (C18:3) 8%. 3.0 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC 2.5 2.0 I3 -2.5 -3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 sqrt (f) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON (Hz’9 Fig. 2. Frequency scan performed (with water as sample) sample. to measure the thickness of the Photopyroelectric measurement qf’ thermal parameter? ; 1% oil: produced by Carbonell, Spain; composition: (Cl60 and Cl8:O) 43%, and C18:l) 2176, (C18:2 and C18:3) 6396, higher fatty acids trace level. Sunflower oil: produced by VNR (Reform), The Netherlands; main fatty acid composition: (C16:l and C18:l) 21%, (C18:2 and C18:3) h3%, saturated fatty acids 12%. Pumpkin oil: produced by Oljarna FRAM, Slovenia; name of product: GIMA, bucino olje; composition: not specified. Motor oil: produced by Adam Opel AG, Ruesselsheim, Germany; name of product: Motor Oil SAE 15 W-40, API-SGICD CCMC G4iPD2; composition: not specified on the package. Murgurine: produced by Albert Heijn, The Netherlands; name of product: Esona: composition: mainly C18:2 and C18:3 (45%). Butter: produced by van der Pol en Zonen BV, Wijk en Aalburg, The Netherlands; name of product: Echte Boter; composition: mostly saturated fatty acids. some amount of C18:l. Peunut huttcr: produced by Calve, The Netherlands; name of product: Calve Pindakaas; composition: peanuts, oil and fat of plant origin. Mustard: produced by Kuhne Netherlands: name of product: Franse Mustard: composition: water, mustard flour, vinegar, salt, sugar, seasonings. Ketchup: produced by Heinz, Elst, The Netherlands; name of product: IHeinz Super Hot Ketchup; composition: tomato puree, vinegar, sugar, glucose syrup, salt. seasonings. SOLIYcream: produced by Bubi, Dortmund, Germany; name of product: Saurc Sahne; composition: 10% fat; no other data available on the package. Orunge juice: produced by Riedel, Dranken Industrie, The Netherlands; composition: 100% orange juice. Olive (Cl&l RESULTS Typical results of the frequency scan measurements performed in the standard configuration for soya, olive and sunflower oils are displayed in Fig. 3. The slopes obtained in the standard confieuration and the PPE signals obtained in the inverse configuration are presented in Table 1. According to gqn (2), the sample’s thermal effusivity can be calculated from: e, = ewatcl.VLI’<IICI./V\ All four thermal parameters are presented (3) in Table 2. DISCUSSION The PPE technique in two configurations (SPPE and IPPE) was proposed in order . . to obtain room temperature values of static and dynamic thermal parameters tor some (semi-)liquid foodstuffs. The method is able to directly measure two dynamic thermal parameters (thermal diffusivity and effusivity); the remaining two thermal parameters can be easily derived. The method is self-consistent, being independent of results obtained by other calorimetric techniques. The only literature data used for calibration were the 160 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA D. Dadarlat et al. 3.0 1 2.5 2.0 A. u 0 L V \ 1.5- 4 % 1.0 - 5 'I E 0 _c a.5 - a 0.0 - l.'*. w n_ -0.5 - CL -1.0 -1.5 - -2.0 /, 0.0 0.2 0.4 1.2 sqrt(f)0~6(Hz@ Fig. 3. Frequency zyxwvutsrqponmlkjihgfe I,0 scan for: (a) soya; (b) olive; and (c) sunflower oils. thermal parameters of water (literature scatter for thermal parameters of water is low). Due to the very good signal to noise ratio, the typical accuracy of such a measurement is about 0.1%. Unfortunately, the final accuracy of the results is much lower. TABLE 1 The Slope in the Direct Configuration and the PPE Signal Amplitude Inverse Configuration for the Investigated Samples Sample Soya oil Olive oil Sunflower oil Pumpkin oil Motor oil Margarine Butter Peanut butter Mustard Sour cream Apple juice Ketchup Water IPPE amplitude (m V) Slope 3.662 3.632 3.382 3.638 4.154 3.111 3.213 3.332 1.477 1.532 1.463 1.548 1.410 - 2.695 - 2.654 - 2.665 - 2.627 - 2.506 - 2.959 - 3.253 -3.216 - 3.236 - 2.983 - 2.754 - 3.474 “Water was used for calibration. The precision residuals about mean explained (PRME). of the fit is also presented Obtained in the PRM E 100 100 100 100 99 99 99 100 100 100 99 100 by the percent of Photopyroelectric measurement 161 zyxwvuts of thermul purumefers TABLE 2 The Static and Dynamic Thermal Parameters (W .,sI::,m’.K) 1.40 I .45 143 1.62 I .48 I.11 0.06 WY9 0.97 I.14 I .?i 0~85 616 621 667 620 543 725 702 677 1527 1462 1542 1457 Olive oil Sunflower oil Pumpkin oil Motor oil Margarine Butter Peanut butter Mustard Sour cream Apple juice Ketchup Samples c’ x lo- (’ (J/m-‘ .K) zyxwvutsrqp 2 x 10’ (m’ /s) Sample Soya oil of the Investigated 1.64 I .hS I .7s I54 I .42 2.16 2.2’) 2.12 J.Y4 4.30 3.22 4.04 0.23 0.24 0.25 0.25 0.21 0.24 0.22 0.2 1 0.48 0.4’) 057 0.42 In the IPPE measurement, eqn (2) is valid only within 2% (Dadarlat et al., 1955). The reproducibility of the slope value (for the same material) in the standard configuration is about 2% (see Fig. 4 for margarine), giving about 4% accuracy for Y. This discrepancy is not due to the method, but probably sample inhomogeneity. We want to stress the fact that this inhomogeneity of real food samples is the main limitation on the final accuracy of the results. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP 7. ‘A ‘.;.. *:. *.*. *.*. 0.‘. *. *I) *. *. l . *’ 0;. . * -. ** . *. ‘* -1.0 b slopes: zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA a -3.032 \ a !Ib -2.959 -1.5 1 -2.00( 0.2 0.4 sqrt(f) 0.6 1.0 1.2 (Hz’%: Fig. 4. Frequency scans performed for margarine (two samples from the same Between the two successive runs the cell was dismounted and cleaned. pack). 162 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA D. Dadarlat et al. Some of the values obtained are typical for fats (oils, margarine, butters) and the others are not far from the values of the thermal parameters of water (for mustard, sour cream, apple juice, ketchup), denoting the high water content of these samples. Various oils exhibit similar values for their thermal parameters. Some difference was found in the thermal effusivity of butters and margarine, thus explaining the different taste sensation in the mouth for these two types of products. As a final conclusion, the method proposed in this paper is simple, fast and sensitive. Its main merit is the possibility of obtaining independently of other methods all four thermal parameters. In this paper the method was used with liquid and semi-liquid samples, but its use can be extended to solid food samples; in this case a coupling fluid must be used between the sample and the sensor. The method should be a good tool for reducing the present lack of thermal parameter data for many food products. ACKNOWLEDGEMENT The authors acknowledge the financial Scientific Research (NWO). support from the Dutch Organization for REFERENCES Bicanic, D. (ed.) (1992). Photoacoustic and Optical Sciences, Vol 69. Springer, Berlin. Photothermal Phenomena III, Springer Series in Chirtoc, M. & Mihailescu, G. (1989). Theory of the photopyroelectric method for investigation of optical and thermal materials properties. Phys. Rev., B40, 906. Dadarlat, D., Bicanic, D., Visser, H., Mercuri, F. & Frandas A. (1995a). Photopyroelectric method for determination of thermophysical parameters and detection of phase transitions in fatty acids and triglycerides. Part I: Principles, theory and instrumentational concepts. 1. Amer. Oil Chem. Sot., 72, 273. Dadarlat, D., Bicanic, D., Visser, H., Mercuri, F. & Frandas A. (199%). Photopyroelectric method for determination of thermophysical parameters and detection of phase transitions in fatty acids and triglycerides. Part II: Temperature dependence of thermophysical parameters. J. Amer. Oil Chem. Sot., 72, 281. Dadarlat, D. & Frandas, A. (1993). Inverse photopyroelectric detection method. Appf. Phys., A56, 235. Dadarlat, D., Visser, H. & Bicanic D. (199%). Improved inverse photopyroelectric cell for thermal effusivity measurements of food products. M eas. Sci. Technol., 6, 1215. Mandelis, A. & Zver, M. M. (1985). Theory of photopyroelectric spectroscopy of solids. J. Appl. Phys., 57, 442 1. Marinelli, M., Zammit, U., Mercuri, F. & Pizzoferrato R. (1992). High-resolution simultaneous photothermal measurements of thermal parameters at a phase transition with the pyroelectric technique. J. Appl. Phys., 72, 1906.