Fan Fiction

Measurement of glass transition in native wheat flour by dynamic mechanical thermal analysis (DMTA)

Measurement of glass transition in native wheat flour by dynamic mechanical thermal analysis (DMTA)
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  *Correspondent: SPI – Sociedade Portuguesa de InovaçãoEdf. Les Palaces, R. Júlio Dinis, 242 – S.208, 4050 Porto,Portugal. Fax: +351 2 6099164; email: Measurement of glass transition in native wheat flourby dynamic mechanical thermal analysis (DMTA) Pedro M. Pereira 1 & Jorge C. Oliveira 2 * 1Escola Superior de Biotecnologia, Universidade Católica Portuguesa, R. Dr. António Bernardino de Almeida, Porto,Portugal2Inter-University Institute of Macau, NAPE, Lote 18, R. Londres, Edf. Tak Ip #P, Macau (Received 13 December 1998; Accepted in revised form 15 May 1999) Summary This work describes a method to study glass transition on native starch powders, basedon dynamical mechanical thermal analysis using compression tests, and was applied towheat flour (13.5% water content). This method will allow the determination of Tg innative (unprocessed) starchy materials, with minimal disturbance of the natural struc-tures. The influence of the test conditions (heating rate, frequency and strain) on the glasstransition measurements was determined using factorial designs. The values of Tg deter-mined as the maxima of the energy dissipation (peaks in E  ) of native flour and of freeze-dried pre-gelatinized flour were not statistically different (around 64  C). The heating ratedid not affect the measurements in the range tested (0.25 to 1  Cmin  1 ). An interactiveeffect of the strain amplitude and the frequency was detected. The significance of thisinteraction can be caused by differences in mechanical energy dissipation, which wouldindicate that not only temperature but also the total energy input may affect this transi-tion. Slight effects of phase separation between gluten and starch were found on nativeflour. Keywords Phase transitions, starch, viscoelastic properties. Introduction Low moisture, amorphous or semi-crystalline bio-materials, such as flours, behave as meta-stablesystems, where molecular mobility plays a majorrole on the kinetics of (desirable or undesirable)chemical and biochemical reactions. The state of the matrix components and its structure is there-fore a key factor in the definition of the mechani-cal properties, stability or reactivity of a givensystem. Changes in molecular mobility and viscos-ity that take place relatively close to the state tran-sition known as glass transition affect the physicalstability and processability of the amorphous frac-tions of foods and the rate of chemical reactions(del Pilar-Buera et al  ., 1995). The glass transitionof biopolymers is, thus, a key parameter for defin-ing the mechanical and storage properties of foods(Kalichevsky & Blanshard, 1992).The glass transition of biological and food mate-rials is a dynamic process influenced by tempera-ture, time and composition of the matrix, wherebyan amorphous matrix changes from a glassymechanical solid (capable of supporting its ownweight against flow due to the force of gravity),where molecular mobility is restricted to vibration,to a soft rubbery state, where re-crystallization,microbial growth, enzymatic activity and generaldegradation reactions may occur (Slade & Levine, International Journal of Food Science and Technology 2000, 35 , 183–192 © 2000 Blackwell Science Ltd 183  1991). The transition from a glassy to a rubberystate and the accompanying loss of stiffness, hard-ness, or strength are traditionally described by arelationship between the magnitudes of these para-meters and temperature, moisture content, orwater activity (Peleg, 1994). A material in glassystate is generally regarded as stable, while rubberystate is necessary for processability. The glass transition, as a second order transition(a change of state but not of phase – Roos, 1995),takes place over a range of temperatures, being themidpoint of the range determined by differentialscanning calorimetry (DSC) conventionally taken asthe single reference point, named Tg. However,whether Tg should be the midpoint, the onset or theendset of glass transition is a matter of discussion(Oliveira et al  ., 1999). The importance of the glasstransition temperature range as a reference parame-ter for an increasing number of applications in foodscience and engineering involving stability and reac-tion kinetics in amorphous matrixes is very clear(Roos, 1995). The measurement of Tg for any givensubstance is crucial for the application of these con-cepts in product and process design and optimiza-tion. Available methods rely on the changes of themacroscopic material properties (such as heatcapacity, thermal expansion coefficient, viscosity,elasticity, etc) associated to the changes of molecu-lar mobility around the glass transition (Roos,1995).In dynamic mechanical thermal analysis (DMTA)three major changes are detected when the vis-coelastic properties of the materials are analyzedagainst temperature: a significant drop in the stor-age modulus (E  ), which is constant in the glassystate; a peak in the loss modulus (E  ), related to theresonance frequency of the molecular motions(Allen, 1993); and a peak in the viscoelasticity ratio(tan  ), which is also frequency-dependent. Thesechanges have been correlated with, respectively, theonset, midpoint and endset of the transition asdetected by DSC (MacInnes, 1993; Roos, 1995).Moreover, DMTA is a frequency-response analysis,that is, the parameters measured will show a fre-quency dependence which is a mirror image of thetime-dependence of the state transition. While rais-ing problems on how to compare Tg measurementsbetween DMTA and DSC, this offers new possibil-ities to investigate glass transition.Wheat flour is one of the most used food rawmaterials world-wide, presenting a very complexstructure partially crystalline, partially amorphous,comprising a number of components that canundergo a glass transition process. Most of the rel-evant components have been thoroughly studiedindividually, as isolated substances (wheat starch –Biliaderis et al  ., 1986; Zeleznak & Hoseney, 1987;amylopectin – Kalichevsky et al  ., 1992; Kalichevsky& Blanshard, 1993; gluten – Hoseney et al  ., 1986;Kalichevsky et al  ., 1992; glutenin – Kalichevsky et al  ., 1992; gliadin – Graaf et al  ., 1993) or as bina-ry mixtures of the pure substances (amylose andgluten – Kalichevsky & Blanshard, 1992). Althoughthis approach enables a more precise insight intothe behaviour of each component, the effect of thecomplexity of the whole structure on glass transi-tion cannot be accurately inferred from model sys-tems (e.g. the granular structure of flour, withcrystalline and amorphous regions, the interactionsbetween components that cannot be reproduced onartificial mixtures, the intra-granular structure). It isalso noted that in most methods used to determineTg in flours (DSC for example), either the samplepreparation process can be disrupting in terms of the native structure and component distribution,which can potentially affect the measured value or,at low moisture contents, the transitions are notclearly defined and the methods do not have a verygood accuracy.The objective of this work was the developmentof a glass transition measurement method thatcould be applied to native flours, based on DMTA.A study of the influence of the test parameters onthe Tg values determined and of the influence of thenative structure (e.g. crystallinity) on the glass tran-sition was also required to understand and validatethe method. Materials and methods Principles of DMTA The Rheometrics Solids Analyser RSAII(Rheometrics, Inc , Piscataway, NJ, USA) machinewas used to compress the sample between twoplates. The lower plate was moved by the actuatorat a specified frequency for a specified length, thuscausing an oscillatory strain input. The top platemoved to adjust a force sensor, thus measuring theresulting sinusoidal stress. DMTA of native flour Pedro M. Pereira & Jorge C. Oliveira  International Journal of Food Science and Technology 2000, 35 , 183–192 184 © 2000 Blackwell Science Ltd  After some time the output stress frequency isthe same as the input strain frequency, but thewave is deviated by the so-called phase angle,rotating between 0  for an elastic solid to 90  fora pure viscous liquid. From this phase angle andthe ratio of the amplitudes of the input and outputsinusoids, the relationship between stress andstrain is fully quantified by the following relation-ships:Input strain wave:     0 sin(  t)Output stress wave:     0 sin(  t   )Complex elasticmodulus:E*   0 /  0   E    iE  Storage modulus:E    E* cos  Loss modulus:E    E* sin  This is a typical frequency-response analysis, wide-ly used in process dynamic studies of chemicalengineering (Luyben, 1990).E  quantifies the energy that is stored by thematerial during the compression movement of theplates and is released back during expansion andis known as storage modulus. E  quantifies theenergy that is dissipated, that is, is supplied duringcompression but lost as thermal energy or used inmolecular rearrangements and is known as lossmodulus.These relationships apply only if the materialexhibits linear viscoelasticity, which implies thatE  , E  and tan  must be independent of the strainamplitude. This usually requires that measure-ments can only be made at sufficiently low strainamplitudes. According to the frequency-responsetheory, E* is also the value of the transfer functionof the relationship between stress and strain in the Laplace domain, obtained by replacing theLaplace variable by  i.In DMTA, measurements can be continuouslytaken while the temperature is changed inside themeasuring chamber. Native flour (NF) samples Commercial wheat flour was used (FarinhaExtreme 75, MOANOR, Porto, Portugal). Theflour was equilibrated at 50% relative humidity at25  C for 2 weeks prior to measurement, corre-sponding to a final moisture content of 13.5g of water/100g of flour (determined gravimetrically byoven drying, norm ISO 1666 1973, method II). Pre-gelatinized flour (PGF) samples The same commercial wheat flour (MOANOR,Porto, Portugal) was gelatinized by preparing anaqueous solution of 10g of flour / 100g of water andheating at 80  C for sufficient time to ensure fullgelatinization (this was verified by DSC – Shimadzu,Japan). The resulting gel was cooled to  28  C (inan air freezer) and freeze-dried at 1 psi and 20  C(ARMFIELD FT33, Armfield Limited, Hampshire,England). The resulting solid (moisture content 5 gof water/100g of solids) was then ground and equi-librated at 50% relative humidity at 25  C for 2weeks achieving a final moisture content of 14.2 g of water/100g of solids (determined gravimetrically byoven drying, norm ISO 1666 1973, method II). Sample preparation Cylindrical pellets (   13 mm, h  18 mm) wereobtained by pressing the powders at 2 bar forapproximately 5 min in a pellet press (SPECAC,Porto, Portugal Model 15.011). The pellets wereimmediately wrapped in aluminium foil to protectfrom moisture exchange. DMTA mechanical spectra measurement Prior to the measurements, a full cycle at all fre-quencies and strain amplitudes, but at room tem-perature, was performed in each sample, to smoothout the effect of residual compressibility in the pel-lets. Samples were then tested using a linear tem-perature increase, in the compression mode, withcylindrical plate tools (plate diameter: 13 mm). Thetests were conducted from ambient temperature upto 100  C. Water loss during testing was determinedby weighing the initial and final samples.The values of temperature, storage modulus (E  ),loss modulus (E  ), viscoelastic ratio (tan  ) and thevariation of sample length (  L) were recorded dur-ing the tests by the equipment software (Rhios, ver-sion 4.44 , Rheometrics, Inc , Piscataway, NJ, USA).It was not possible to repeat runs with the samesample due to significant degradation at high tem-peratures. Replicates were performed with differentsamples. Moisture loss during testing Without protective means, samples would dehy- DMTA of native flour Pedro M. Pereira & Jorge C. Oliveira  International Journal of Food Science and Technology 2000, 35 , 183–192 185 © 2000 Blackwell Science Ltd  drate during measurement and therefore theywere wrapped. Polyethylene film and aluminiumfoil were tested, comparing the weight loss duringtesting with the one observed with unwrappedpellets.Different wrappings were also tested: (1) com-pletely wrapped samples, being wrapped on thetop, bottom and side; (2) partially wrapped sam-ples, only on the side, leaving top and bottom indirect contact with the DMTA tool plates.The samples were tested in the DMTA cham-ber with a temperature ramp default test, fromambient temperature to 90  C, with a strainamplitude of 0.0013 and a frequency of 1Hz. Heat transfer efficiency measurements In situ measurements of the sample temperaturewere made by a thermocouple (Fe-CuNi (Iron-Constantan) wire,) placed as close to the centre of the sample as possible, with the thermocouple endbeing insulated with PTFE to prevent heat con-duction through the wire. Another thermocouplewas placed inside the DMTA chamber to moni-tor chamber temperature. Both thermocoupleswere connected to a temperature acquisition sys-tem. Tests were then performed at different heat-ing rates (0.5 and 1  Cmin  1 ). The frequency andstrain amplitude were 1 Hz and 0.0013, respec-tively. Measuring conditions To assess the influence of the test parameters onthe DMTA results, native flour samples were test-ed using different combinations of heating rate,frequency and strain amplitude according to areplicated full factorial design at two levels.Table1 shows the values of the levels considered.The more significant effects were then detailed bymeans of a full factorial design at four levels, withthree replicates, for the two input factors strainamplitude and frequency, as shown in Table2.The heating rate was then constant and set to1  Cmin  1 .In order to assess the influence of the nativestructure of flour, pre-gelatinized samples weretested at five different frequencies (0.1, 0.2, 0.5,0.8 and 1 Hz), with a strain amplitude of 0.0004and at a heating rate of 1  Cmin  1 . Results and discussion Moisture loss The glass transition temperature is very sensitiveto water content (Zeleznak & Hoseney, 1987) DMTA of native flour Pedro M. Pereira & Jorge C. Oliveira  International Journal of Food Science and Technology 2000, 35 , 183–192 186 © 2000 Blackwell Science Ltd ExperimentHeating rate (  Cmin  1 )Frequency (Hz)Strain amplitude(  1000) 11142111.3310.24410.21.350.251460.2511.370.250.2480.250.21.3 Table 2 Experimental design for four level factorial designfor two factors ExperimentFrequency (Hz)Strain amplitude(*1000) 10.2420.2730.21040.21350.5460.5770.51080.51390.84100.87110.810120.813131414171511016113 Table 1 Experimental design forthe two level factorial design forthree factors  and though the starch was compressed in pelletsthat only had the sides exposed to the air in thechamber, the moisture loss during testing was toohigh to allow for accurate measurements, asshown in Fig.1. However, when wrapping thesamples completely in aluminium foil, moistureloss was prevented. This would not be achieved if wrapping only the sides (partial wrapping) or if using polyethylene film instead of aluminium foil,as can be seen in Fig.1.It must be noted that aluminium foil does notgo through significant changes in the range of temperatures in question (ambient to 90  C) andas the measurements are based on variations of the values of E  and E  (that is, relative changesrather than absolute values), this procedure isacceptable. Thermal lag Ideally, the thermal lag between chamber and sam-ple temperature should be negligible or constantfor all test conditions, as otherwise the heatingrates established in the experimental design mightbe slightly different from the real ones, as the ther-mal lag would vary somewhat with the heatingrate. Moreover, the temperature gradient in thesample itself (from surface to centre temperature)might be too high and affect the desired accuracyof the readings. Figure2 shows the thermal lag interms of the real pellet temperature versus thechamber temperature readings. It can be seen thatwith full aluminium foil wrapping, the thermal lag is approximately constant, around 2  C, up to70  C for the maximum heating rate used –1  Cmin  1 . This is not the case of the unwrappedsamples, which may be due also to the effect of thewater evaporation. Mechanical spectra The mechanical spectra of the native flour (NF)samples showed a clear change of the viscoelasticproperties according to the expected glass transi-tion effects: a drop of E  around 55  C, a peak of  DMTA of native flour Pedro M. Pereira & Jorge C. Oliveira  International Journal of Food Science and Technology 2000, 35 , 183–192 187 © 2000 Blackwell Science Ltd Figure1 Moisture loss during testing. A – unwrappedsample; B – sample partially wrapped in aluminium foil;C – sample totally wrapped in aluminium foil; D –sample partially wrapped in polyethilene film; E – sampletotally wrapped in polyethilene film. Figure2 Thermal lag duringtesting for unwrapped (  ) andwrapped (  ) samples inaluminium foil.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks