This item is only available as the following downloads:
EXPERIMENTAL THER M AL ANALYSIS OF AQUEOUS Al 2 O 3 NANOFLUIDS BY KATHERINE MCALPINE A Thesis Submitted to the Division of Natural Science New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts Under the sponsorship of Dr. Mariana Sendova Sarasota, FL May, 2013
ii A c kn owl e d g e m e nt s Thank you Dr. Mariana Sendova for all of your support over the past year during this thesis project. Without your thoughtful guidance, I would not have learned nearly as much as I did about my project. Thank you Ralf Raud for your dedication and curiosity. Thanks Dr. Brian Hosterman for your help, especially with the liquid nitrogen orders.
iii Table of Contents A c kn owl e d g e m e nt s ................................ ................................ ................................ ............. ii Abstract ................................ ................................ ................................ .............................. vi Introduction ................................ ................................ ................................ ......................... 1 1. Nanofluid Overview ................................ ................................ ................................ ....... 3 1.1. Definition of Nanofluids ................................ ................................ ......................... 3 1.2. Preparation of Nanofluids ................................ ................................ ....................... 3 1.3. Thermal Properties ................................ ................................ ................................ .. 5 1.4. Phase Transitions: Freezing ................................ ................................ .................... 8 1.5. Differential Scanning Cal orimetry ................................ ................................ .......... 9 1.5.1. Measuring the Specific Heat Capacity of Water ................................ ............ 11 1.6. Applications of Nanofluids ................................ ................................ ................... 12 2. Theoretical Calculation ................................ ................................ ................................ 15 2.1. Calculating the Specific Heat Capacity ................................ ................................ 15 2.2. Calculat ing the Latent Heat ................................ ................................ ................... 16 2.3. Calculating the Melting Point ................................ ................................ ............... 16 3. Experimental Procedure ................................ ................................ ............................... 17 3.1. The Differential Scanning Calorimeter ................................ ................................ 17 3.2. Calibration ................................ ................................ ................................ ............. 19 3.3. Data Acquisition ................................ ................................ ................................ .... 19 4. Experimental Results and Analysis ................................ ................................ ............. 22 4.1. Heat Flow Curves ................................ ................................ ................................ .. 22 4.2. Speci fic Heat Capacity Calculations ................................ ................................ ..... 27 4.3. Additional Thermal Analysis ................................ ................................ ................ 30 5. Conclusions and Further Experiments ................................ ................................ .......... 39 Bibliography ................................ ................................ ................................ ..................... 41
iv Figures and Tables Table 1.3.1 Notations used in the Debye Model equation ...7 Figure 1.3.1 Theoretical heat capacity of germanium and silicon ...7 Figure 1.5.1 DSC cell schematic Figure 3.1.1 DSC picture Figure 3.1.2 Nanofluids picture Figure 3.1.3 DSC cell aerial picture Figure 3.3.1 Sample pans .......................................................................................20 Table 4.1.1 Sample mass measurements Figure 4.1.1 Normalized h eat flow for 2% samples Figure 4.1.2 Normalized heat flow for 5% sa mples Figure 4.1.3 Normalized heat flow for 10% samples Figure 4.1.4 Average heat flow Figure 4.1.5 Average heat flow where heat flow is approximately constant Figure 4.2.1 Specific heat ca pacity for samples in liquid state Figure 4.2.2 Specific heat capacity for samples in solid state Figure 4.2.3 Average specific heat capacity versus concentration in liquid state Figure 4.2.4 Average specific heat capacity ve rsus concentration in solid state Figure 4.2.5 Specific heat capacity of all samples in liquid state Table 4.3.1 Calculated thermal properties for 2% samples Table 4.3.2 Calculated thermal properties for 5% samples .. Table 4.3.3 Calculated thermal properties for 10% samples 32
v Table 4.3.4 Calculated thermal properties for water .. 32 Figure 4.3.1 T M versus concentration Figure 4.3.2 T F versus concentration Figure 4.3.3 Latent heat versus concentration Figure 4.3.4 FWHM versus concentration Table 4.3.5 Heat flow curve slopes Table 4.3.6 Thermal parameters for selected samples
vi EXPERIMENTAL THERMAL ANALYSIS OF AQUEOUS Al 2 O 3 NANOFLUIDS Katherine McAlpine New College of Florida ABSTRACT Thermal properties of nanofluids, including specific heat capacity and melting point, are measured experimentally u sing a differential scanning calorimeter (DSC). The nanofluids consist of Al 2 O 3 nanoparticles dispersed in de ionized water at three different volumetric concentrations: 2%, 5%, and 10%. The theory behind the measured thermal properties and the DSC is presented It was found that the alumina chemically reacted with the aluminum pans used to hold the samples This reaction makes the measurements from the DSC inconclusive. Dr. Mariana Sendova Division of Natural Science
1 Introduction W ater has a h igh thermal conductivity and high specific heat capacit y in comparison to other common liquids. Thus, water is widely used for applications involving heat transfer and storage. It has been shown that adding nanoparticles (NPs) to water can further increa se its specific heat capacity and thermal conductivity Nanoparticle s are particles between 1 100 nm in diameter. 1 The resulting mixture of NPs and water is an aqueous nanofluid. Nanoparticles stay suspended in water and do not settle because of their size. 2 To determine the thermal conductivity of a nanofluid, the specific heat capacity must be measured precisely. In this study, the specific heat capacity of water based nanofluids with Al 2 O 3 nanoparticles at three different volumetric concentrations (2%, 5%, 10%) was measured. Four other properties of the nanofluids were calculated: melting point, latent heat, the temperature at which the sample is entirely melted, and the full width at half maximum (FWHM) of the heat flow curve during the phase tra nsition from liquid to solid. Measurements were performed with a differential scanning calorimeter (DSC). A DSC functions by comparing a sample with unknown properties to a reference sample with known thermal properties. This method has proven to give accurate values for the specific heat capacity of water, within 1.5% of accepted values. 3 I t is proposed that the DSC will give an accurate and precise measurement of properties of the nanofluids. This work consists of five chapters. The first chapt er is a general review of nanofluids. It includes methods of preparation, theory which describes their thermal properties, and the theory behind a DSC used to study them. The second chapter details
2 the theory behind the calculations of the thermal proper ties measured in the work. The third chapter describes the experimental method involved in using the DSC. The fourth chapter reports the data acquired. The fifth chapter analyzes the data and reports the results of the calculations performed.
3 1. Nanof luid Overview 1.1 Definition of Nanofluids Nanofluids are colloids which consist of nanoparticles suspended in a base fluid. 2 A colloid is a mixture ( usually of two substances ) where one substance is suspended in the second. The suspended substance i s in the form of particles, exceeding molecular size. The base fluids are often water or other organic fluids such a s ethanol and ethylene glycol. Nanoparticles (NPs) are particles between 1 100 nm in diameter. 1 Unlike millimeter and micrometer sized particles, NPs are of the order of magnitude of the molecules of the fluid. Thus, they tend to be more stable. Consequences that come from mixing larger particles such as sedimentation, clogging, abrasion, and pressure drop are not applicable for colloids 2 Particles do not settle down in the liquid but rather they stay suspended for a long period of time. 1.2 Preparation of Nanofluids In addition to maintaining the chemistry of the base fluid, nanofluids must be prepared in a way to avoid the problems associated with larger particles listed in the previous section The two methods for nanofluid preparation are detailed in the following paragraphs. There are two single step methods T he nanoparticles are synthesized into the liquid, as opposed to syn thesis without involvement of the base fluid. The single step methods are classified as either a single step physical method or a single step chemical method. With the single step physical method, the nanoparticles are prepared by physical vapor depositi on (PVD). The single step chemical method involves preparation by a liquid chemical reaction. 2
4 The single step physical method was reported by Akoh et al. 5 and Eastman et al. 6 In the technique, solid bulk copper (Cu) is heated, and the resulting vapor, exposed to ethylene glycol (EG), condenses into nanoparticles. This process forms a nanofluid with an EG base fluid of 10 nm Cu pa rticles. A second method involves heating a copper or titanium metal electrode using arc sparking produces particles. The vapor condenses inside of a vacuum chamber which is filled with a liquid creating the nanofluid of CuO or TiO 2 particles. This process is referred to as Submerged Arc Nanoparticles Synthesis System (SANSS). 7, 8 The single step methods are advantageous b ecause problems associated with drying, storing, transporting, and dispersion of the nanoparticles are avoided. The particles also do not agglomerate. The disadvantage of the process is that it is only possible to create low vapor pressure flu ids. The t wo step method disperses synthesized nanoparticles into a base liquid. 2 The first step produces the nanoparticles, nanotubes, or nanofibers as dry particles by methods such as inert gas condensation, chemical vapor deposition, or mechanical alloying. The nanoparticles are then dispersed into the base fluid. This two step method has been used by Eastman et al. 6 and Lee et al. 9 when preparing Al 2 O 3 /water nanofluids. Hong 10 dispersed iron (Fe) nanocrystalline powder in EG by this method as well. With the two step method, it is more common for particle agglomeration to occur, which leads to settling, or clogging. Consequentially thermal properties of the nanofluid may be altered. In order to correct these issues, ultrasonic agitation, surfactant addition or pH control is used to disperse the particles. Industrial companies use the two step method because the method is cost effective. The technique is more commonly used to prepare non metallic nanofluids. 2
5 1.3. Thermal Properties Nanofluids are interes ting to scientists and engineers because of their thermal properties 2 In particular, nanofluids have an increased thermal conductivity 2 Thermal conductivity defines the property of a material to conduct heat. The thermal conductivity of a nanofluid is significantly greater than that of the base fluid, with 40% and 161% increases measured experimentally 11 This observation is attributed to the increase of the specific surface area of the nanoparticles, which increases the heat transfer surface and allo ws heat to flow faster between the particles and fluids 2 T hermal conductivity depends on the specific heat capacity of the fluid, which motivates an accurate calculation of the specific heat capacity 2 The specific heat capacity is defined as th e heat c apacity per unit mass. The heat capacity for a system at a constant pressure is defined as the partial derivative of the energy of the system with respect to temperature at a constant pressure 12 The quantities thermal conductivity and specif ic heat capacity c nf are relate d at a constant pressure, (1. 3 .1) where is thermal diffusivity is the density All specific heat capacity values reported are at a constan t pressure. Because thermal conductivity depends on specific heat capacity it is important that the specific heat capacity is known accurately and precisely. An expression for the specific heat capacity of a nanofluid can be derived using the principle of superposition An equation for the heat capacity of a mixture, C in thermal equilibrium is the sum of the heat capacity of each component of the mixture 5 : C = C 1 +C 2 +...+C N (1.3 .2)
6 The specific heat capacity, c, is defined as the heat capacity per unit mass, m and is defined a s (1. 3 .3) where the i th component has a volume v i and density i Dividing the numerator and denominator of Eq. 1.3.3 by the total volume yields: (1. 3 .4) Where the volume fraction of the i th component in the mixture is represented by and is the density of the mixture Applying Eq. 1. 3 .4 to the mixture of nan oparticles and water whose density is nf yields: 5 (1. 3 .5) where subscript f denotes fluid, and np denotes nanoparticle. In general, specific heat capacity depends on the degrees of freedom associated with a specimen Molecules have translational degrees of freedom which is associated with the kinetic energy of the specimen There are more degrees of freedom for molecules which can rotate and vibrate. The more deg rees of freedom a molecule has, t he higher the heat capacity. 12 Phonons are another way substances absorb energy and increase their heat capacity. Phonons are collective excitations in a periodic, elastic arrangement of molecules. The resulting heat capa city is described by the Debye model. The specific heat capacity at a constant volume C V ( ), is given by 13
7 (1.3.6) where notations in Eq. 1.3.6 are described in Table 1.3.1. N N umber of atoms in the specimen k B T Thermodynamic Temperature Angular frequency of the lattice mode x v Wave velocity V Volume x d Table 1.3.1 No tations used in 1.3.6. Figure 1.3.1 shows the heat capacity as a function of temperature of germanium and silicon as modeled by the Debye model. Figure 1.3.1: The heat capacity of germanium and silicon. 13
8 Under some conditions, some nanofluids have an en hancement in heat capacity that is not predicted by Eq. 1. 3 .5 Silica particles of 20 nm increased the heat capacity of the lithium potassium carbonate eutectic by 22% at a .01 mass fraction. 1 3 This unpredicted increase in heat capacity has also been rep orted from a measurement of a carbon nanotubes/liquid carbonate eutectic nanofluid, where the heat capacity was measured to be 9% 17% larger at .0001 .001 and .005 .01 mass fractions, respectively. 13 A eutectic mixture remains liquid at temperatures lower than the melting point of the individual constituents. 12 The increased specific heat capacity comes from a rearrangement around the solid particles in the eutectic mixture to a configuration which has a higher heat capacity. 13 The solid liquid interface may change the phonon vibration mode of the nanoparticle. This change could alter the heat capacity of the entire nanofluid 5 The simple theory that gives Eq. 1.3 .5 successfully models the specific heat capacity of most nanofluids under most conditions, but it does not take into account this additional way that a substance could absorb energy. 1. 4 Phase Transitions: Freezing The surface free energy between a nanoparticle and water is proportional to r 3 where r is the radius of the nanoparticle. Fo r nanofluids, t he radius of nanoparticles is small and on the order of magnitude of a water molecule Therefore, the wettability between the particle and fluid is high Wett ability is a measure of the contact angle between the solid and the liquid. A su bstance with high wettability has a contact angle of nearly 0 (Figure 1.4.1) 14
9 Figure 1.4.4: The rounded shapes represent a fluid and the line below them represents a solid. Low wettability is shown on the left, and high wettability is shown on the rig ht. Because the surface free energy between the nano particles and water is low, nanopart icles act as a nucleating agent. 15 Nucleation is the formation of small clusters of particles. 14 In general, when a substance freezes, a few molecules begin to form a solid crystal lattice. 16 The nanofluids increase the amount of nucleation sites, thus reducing the freezing point. This theory is supported experimentally. H. Xie et al. reported that adding TiO 2 to water decreases the melting temperature from 0C to 4.0C. 1.5 Differential Scanning Calorimetry Differential Scanning Calorimetry (DSC) measures the thermal properties of liquids and solids including specific heat capacity. 18 In this method, a sample with unknown properties is compared to a referen ce sample with known thermal properties The method employed in this experiment is called the three run method. It is referred to as the three run method because it consists of three heating experiments In each of these experiments heat flow (with unit s of power per unit mass) from an empty pan to a sample pan is measured as temperature is increased at a constant rate The first ru n measures heat flow to an empty pan used to hold samples. The second run measures he at flow to a sample of sapphire, with a known heat capacity The third run measures he at flow to the
10 sample, a nanofluid, in this study. Information from these three runs is used to calculate the specif ic heat capacity of the sample. 19 To measure the heat flow in any of the three runs, a re ference pan and sample pan are placed on separate conducting disk s as shown in Figure 1.5.1 The reference pan is always empty, and the sample pan contains either nothing (first run), sapphire (second run), or the sample (third run). The conducting disk s are the size of the base s of the pan s A surrounding furnace heats the disk s The disk s conduct heat to both the sample pan a nd the reference pan T hermocouple s measure s the temperature difference between the sample pan and the reference pan 20 A the rmo couple consists of two dissimilar metals in the form of wires with two junctions. A potential difference between the two junctions is created s ince the conductors are at two different temperatures (Seebeck effect) Charge carriers from the hotter cond uctor move toward the colder conductor. This creates a potential difference which is proportional to the temperature difference 20, 21 This relationship is used to determine the temperature differenc e Heat will flow between the sample and the reference a s they attempt to reach thermal equilibrium. flow equals the change in temperature ( measured with the thermocouple ) divided by the resistance of the disk. These m easurements are used to c alculate the specific heat capacity of a sample. 22
11 Figure 1.5.1: A schematic of the DSC cell. The advantage of using a DSC is that a sample and reference are subject to the same cell conditions. Power variations will affect both the sample and refere nce equally, and errors are thus minimized. The type of DSC used in the present study is classified as a turret type measuring system. Heat exchange occurs inside of small cylinders. This method has fast thermal responses, implying a large heating and c ooling rate. 1.5.1 Measuring the Specific Heat Capacity of Water Ac cepted values for the therm al properties of water are determined by the International Association for the Properties of Water and Steam (IAPWS). Their 1995 for the Thermodynamic Properties of s for the specific heat capacity of water. 3 The data in this formulation comes from studies by Sirota et al. 24 based on measurements taken from 1956 1970. Sirota et al. used a flow calorimeter to obtain values for the specific heat capacity of wate r. 24 Flow calorimeters make measurements on liquid samples which flow through
12 two tubes in series. The first is at the desired pressure a nd temperature and the second is at the sample pressure at room temperature This method is advantageous to study liquids because there is no vapor space in the tubes to contribute to heat capacity. Measurements for the specific heat capacity of water performed with a DSC are Lourenco et al. 22 reports values for specific heat capac ity for water within 1.5% of the values reported by Sirota et al 3 Six years prior to these measurements, Sampio and Nieto de Castro performed specific heat capacity measurements of water on the same DSC. Their values are also within 1.5% of the values obtained by Sirota et al. 3 This suggests that a DSC is a reliable apparatus to measure the s pecific heat capacity of water 1.6 Applications of Nanofluids thermophysical properties have t he potential to improve a variety of devices which rely on heat transfer. The list of technological areas which could use nanofluids for device improvement is large an d diverse, and includes the following: 2 5 Engine cooling. Engine transmission oil. Heating and cooling of buildings. Electronics cooling. Transformer cooling. Nuclear Systems cooling. Solar water heating. Refrigeration. High power lasers. Biomedical app lica tions Electronics often generate an enormous amount of heat. This is problematic and attempts to remove this high heat flux include air cooling, liquid cooling, and two phase
13 cooling. The high heat transfer coefficients of nanofluids make the use of nanofluids in electronics packaging a viable option to explore. In particular, adding nanodiamond to mineral oil in electrical power transformers can enhance the the rmal conductivity of the oil without making the oil electrically conductive. Retrofitting transformers with mineral oil with nanoparticles additives is a feasible and cheaper alternative to replacement. 2 6 Nanofluids can improve the performance of ch illers in air conditioning systems. Al 2 O 3 /Water nanofluids can be used as a potential new phase change material for the thermal energy storage of cooling systems. The fact that nanofluids freeze very rapidly implies a smaller running time for these types of refrigeration system s Nanofluids can also improve the domestic refrigerator. The addition of TiO 2 nanoparticles to lubricating mineral oils has shown to improve refrigerator efficiency because of a reduction in energy consumpti on. The nanofluids im prove the solubility of the refrigerant and the TiO 2 which allows for the compressor to be more uniformly and efficiently lubricated. 2 6 Nanofluids can be used to make vehicles more efficient by improving engine cooling and vehicle thermal management. Choi et al. 2 7 showed that nanofluids have the potential to be used as an engine coolant because of their higher thermal conductivity. The addition of nanoparticles to engine coolant increases the cooling rate for heavy duty engines. This fact allows for a reduction in size of coolant systems. Reducing the weight of a car will allow it to be more fuel efficient. Additionally, nanofluids have shown to improve engine transmission oil. Nanofluids were shown to have the lowest transmission temperatures when used to lubricate the exterior of a rotary blade coupling transmission of a four wheel drive vehicle. 26
14 Nanofluids have the capacity to improve solar water heaters. Tyagi et al. 2 8 investigated theoretically the possibility of using a nonconcentrating direct absorption solar radiation compared to pure water. Heat transfer enhancement in solar devices is a limitation of saving energy and making solar devices small er, so the use of nanofluid s in this context is promising. 2 6 There are many opportunities for nanofluids to improve a variety of devices w hich depend on heat exchange. In principle, t heir novel thermal properties and in particular, their increased therma l conductivity are the properties which make them suitable replacements for other heat exchange agents and methods However, most works detail theoretical applicatio ns and few applications have been realized in practice. There are still obstacles to emp loying nanofluids in heat exchange devices. The production cost and the long term stability of nanoparticle dispersion in base fluids are limiting factors. 2 6 Nonetheless, nanofluids are a new technology. Therefore there is still much room for explorati on
15 2. Theoretical Calculation 2.1 Calculating the Specific Heat Capacity As stated in S ection 1.3, specific heat capacity is defined as the temperature derivative of internal en ergy U per unit mass m: (2.1.1 ) Since U(T) = U(T(t)), Eq. 2.1.1 is rewritten : (2.1 .2) The DSC compares the heat flow of sapphire, denoted by the subscript s, to the heat flow of a nanofluid. A proportionalit y arises between the specific heat capacity, and heat flow of substances heated at the same rate: ( 2.1 .3) In Eq. 2.1.3 is the heat flow measured at a part icular temperature for sapphire minus the heat flow measured at the same temperature for the empty reference pan. Similarly, is the difference between the heat flow measured at a particular temperature for the nanofluid and the empty r eference pan. This equation can be rearranged to solve for c nf and thus a formula for the heat capacity of a nanofluid arises. ( 2.1.4 ) In the experiment, must be subtracted from the resulting equation to account for the differ ence in weight between the pans:
16 (2.1.5) In Eq. 2.1 .5 is the difference in mass between the ref erence pan and the pan used to hold the sapphire, c pan is the heat capacity of the alum inum pan, and E is defined as ( 2 1 6 ) In Eq. 2 .1. 6 is the difference in mass between the reference pan an d the pan used to hold the sample. 2.2 Calculating the Latent Heat The latent heat of fusion is defined as the heat required to convert a liquid into a solid with no temperature change. The heat flow curves are analyzed to calculate the lat ent heat o f fusion at temperatures where the phase change occurs According to the Universal Analysis manual, the latent heat of fusion is the integral of the heat flow curve. Before integrating, t he heat flow curves are fit with a polynomial. The polynomial is t hen integrated to give the change in heat from the solid phase to the liquid p hase. 2.3 Calculating the Melting Point The melting point is calculated with Universal Analysis. I t involves finding the slope at a point on the endothermic heat curve near the phase transition and determining where that crosses the heat capacity baseline. The heat capacity baseline is a portion where it is relatively constant near the phase transition. Further details of this method are prop rietar y
17 3 Experimental Procedure 3 .1 The Differential Scanning Calorimeter In this study, a Thermal Analysis (TA) Q20 differential scan ning calorimeter (DSC) was used to measure the melting point, heat capacity and the latent heat of Al 2 O 3 /H 2 O nanoflu id with varying particle concentrations. A picture of the DSC is shown in Figure 3 .1.1. A picture of two of the nanofluid samples are shown in Figure 3 .1.2. Figure 3 .1.1: The DSC used in the experiment.
18 Figure 3 .1.2: The nanofluids measured in the experiment are shown. The nanoparticles give the water its white translucent appearance. The thermocouple connecting the reference pan and sample pan inside of the cell is shown in Figure 3 .1.3 Figure 3 .1.3 : A cell with a n aluminum reference pa n
19 The cell is enclosed by three layers of metal covers during experiments. 3 .2 Calibration The DSC requires regular calibration to obtain accurate data Three types of calibration are performed before the sample data are acquired Baseline slope and offset calibration. This calibration involves heating the cell through the entire temperature range to flatten the baseline and zero the heat flow signal. This calibration subtracts any heat flow present between the thermocouples without pans in th e cell. Enthalpy (cell) constant calibratio n. In this calibration, indium, with a known mass, is heated through its melting point. The associated latent heat ( the heat of fusion ) is measured and calculated by integrating over the peak in the heat flow versus temperature graph In principle, the thermocouple and indium should be in thermal equilibrium. As the sample melts, it draws more heat and it becomes a different temperature than the thermocouple. The purpose of this calibration is to account for the temperature difference between a sample and the thermocouple. The heat of fusion is then compared to its theoretical value and a parameter called the cell constant is calculated. Temperature calibration. It is performed during the enthalpy constant calibration. As the indium is heated, the calculated melting point temperature is set to its theoretical melting point 156.6C 3 .3 Data Acquisition Each measurement consists of three temperature runs from 50C to150C The first run is called the empty pan run. First, the mass of two empty Tzero aluminum pans
20 with Hermetic lids are measured A picture of these pans is shown in Figure 3 .3 .1. The first pan is the reference pan to be used for subsequent experiments. The second pan is placed in pl ace of the sample holding pan The measurement is accurate up to the microgram. Figure 3 .3 .1: The pans used to hold the samples in the experiment. The reference pan is placed on the left side of the cell, and the empty pan is placed on the right sid e of the cell. Before each experiment, a Liquid Nitrogen Cooling System (LNCS) tank is filled with liquid nitrogen. The cell is purged with dry nitrogen at 50mL/minute and helium at 25 mL/minute. The helium replaces nitrogen below 0C. Purge gases serv e the following purposes: 2 9 Control sample environment Purge volatiles from the cell
21 Prevent contamination Provide convection currents inside the cell to prevent ice formation The measurements were performed with identical procedures The cell is equilib rated at 150 C and held for 5 minu tes. The cell is heated at 10 C /minute to 50 C and it is held for 5 minutes. The cell is cooled at 10 C /minute to 150 C and it is again held for 5 minutes. This process is repeated such that there are two endothermi c cycles where the cell goes from 150 C to 50 C. Next, the sapphire s tandard run is performed. All of the steps associated with the empty pan run are followed, except the empty sample pan is replaced with a pan with sapphire inside of it. The sapph ire pan was not sealed with a hermetic lid. It is not necessary to do so, because the sapphire is not a liquid. This is a second type of calibration. T he nanofluid is put inside of an empty pan with a micropipette. Special care is taken to prevent the nanofluid from leaking out of the pan during data acquisition. The empty pan procedure is followed, with the empty pan replaced by the pan with the nanofluid. Th is proce dure is repeated for nine samples of three different concentrations and a sample of deionized water.
22 4 Experimental Results and Analysis 4.1 Heat Flow Curves After each experiment, the mass was measured to ensure that no appreciable mass loss occurred during data acquisition. When a sample lost more than 3% of its mass during data acquisition, the measurement was deemed invalid and the data was not used. Table 4.1.1 shows the masses measured in the experiment. Sample Concentration (%) Pan mass Sample mass before measurement (mg) Sample mass after measurement (mg) Percent loss (%) Pan mass/S ample Mass Ratio Empty reference pan n/a 52.59 1 n/a n/a 0 n/a Sapphire pan n/a 50.73 6 26.131 26.131 0 n/a 11 2 52.80 9 8.102 8.102 0 .15 13 2 52.92 7 7.489 7.487 0.03 .14 14 2 51.85 7 5.301 5.301 0 .10 16 5 52.77 1 6.876 6.852 0.3 .13 18 5 52.57 6.921 6.88 0.6 .13 21 5 52.52 5 7.551 7.548 0.04 .14 24 10 52.31 5 7.502 7.502 0 .14 27 10 52.64 1 3.55 3.55 0 .07 28 10 51.59 2 5.555 5.555 0 .11 Water 0 52.25 5 13.363 13.363 0 .26 Table 4.1.1: The measured masses before and after data acquisitio n
23 Raw data from the DSC is a set of data points of heat flow versus temperature. For the purposes of comparing samples with different masses each of these graphs are normalize d by the mass of the sample. Figs. 4.1.1 4.1.3 show the relative heat flow for samples with concentrations of 2%, 5%, and 10%, respectively. Three samples of each concentration were measured over two temperature cycles each. Sample 11, Sample 13, and Sample 14 contained nanofluids with a 2% concentration. Sample 16, Sample 18, and Sample 21 contained nanofluids with a 5% concentration. Sample 24, Sample 27, and Sample 28 contained nanofluids with a 10% concentration. C1 and C2 denote the two endothermic cycles over which each pan was measured. Figure 4.1.1: Low temperature normalized heat flow for 2% concentration nanofluids. Normalized heat flow over the entire temperature range (inset).
24 Figure 4.1.2: Low temperature normalized heat flow for 5% concentration nanoflu ids. Normalized heat flow over the ent ire temperature range (inset). Figure 4.1.3: Low temperature normalized heat flow for 10% concentration nanofluids. Normalized heat flow over the entire temperature range (inset).
25 The lege nds in Figs 4.1.1 4.1.3 apply to the rest of the graphs in this work. The difference between two cycles of one sample is small (see Figs. 4.1.1 4.1.3) compared to the difference between two heat flow curves of different samples of the same concentration. For example, at 26 C the difference in heat flow per unit mass between C 1 and C 2 for S ample 13 (2%) is .00734 W/g. At 26 C, the d ifference in heat flow per unit mass bet ween C 1 for S ample 11 (2%) and C 1 for S ample 13 (2%) is .05297 W/g. The average hea t flow curve of each sa mple is shown in Figure 4.1 .4 and Figure 4.1.5 Figure 4.1.4: The heat flow per unit mass plots for the 9 samples.
26 Figure 4.1.5 : P art of the plot in Figure 4.1.4 where the heat fl ow is approximately constant. This grap h summarizes Figure 4.1.1 4.1.
27 4.2 Specific Heat Capacity Calculations The specific heat capacity of the samples is calculated using Eq. 2.1.5 and the results are plotted in Figure 4 2 .1 and Figure 4.2 .2 The tem perature range for Figure 4.2 .1 is chosen to represent the sample while it is in its liquid phase. The temperature range for Fig ure 5.1.2 is chosen to represent the sample wh ile it is in its solid phase. All theoreti cal values come from The Chemical Rubb er Publishing Company Handbook of Chemistry and Physics 30 and the ASTM E 1269 Standard Test Method for Determining the Specific Heat Capacity by Differential Scanning Calorimetry. 18 Figure 4.2 .1: The specific heat capacity for ni ne different samples at three different concentrations and one water sample is plotted when the sample is a liquid
28 Figure 4.2 .2: The specific heat capacity for nine different samples at three different concentrations and one wat er sample is plotted when the sample is a solid The heat capacity versus percent concentration for two tempera tures is plotted in Figs 4.2 .3 4.2 .4 The theoretical values for specific heat capacity are calculated using Eq. 1.3 .5, and shown on the same plot.
29 Figure 4.2 3 : S pecific heat capacity versus percent concentration at 27 C (solid phase) Figure 4.2 4 : S pecific heat capacity versus percent concentration at 23 C (liquid phase) A plot of the heat capacity of all of the samples at 27C is shown in Figure 4.2 .5
30 Figure 4.2.5: The specific heat capacity of all of the samples at 27C. E vinced by Figur e 4.2 .1 4.2 5 this study measured no correl ation between concent ration and specific heat capacity. The aver age specific heat capacity of all of the nanofluids at 27C is 3. 8 9 and at 23C is 2.3 7 The measured specific heat capacity of water at 27C is 4.3 and at 23C is 2.2 These values are within one standard deviation of this average of the specific heat capacities of the nanofluids Thus, the effects of the nanoparticles in the nanofluids are undetected by the experiments. 4 3 Additional Thermal Analys is Four other p arameters of the nanofluids are calculated: melting point, latent heat, the temperature at which the sample is entirely melted and the full width at half
31 maximum (FWHM) of the heat flow curve during the phase transition from liquid to solid The m ethod for calculating the melting point is described in Section 2.3. The temperature at which the sample is entirely melted is denoted T F. It is calculated by finding the temperature at which the heat flow is at a minimum during the phase transition. At this temper ature, melting is complete 3 1 The latent heat is denoted L. The FWHM comes from the heat flow curves. The peak of the heat curve during the phase transition is measured. The magnitude of the peak is halved, and the width of this curve at the halfway point is calculated The values for these quantities are shown in Table 4.3 .1, Table 4.3 .2, Table 4.3 .3, and Table 4.3 .4. Sample 11 Percent Change (%) Sample 13 Percent Chang e (%) Sample 14 Percent Change (%) Average T M (C) .11, .09 18 3.31, 3.80 15 2.64, 2.64 0 .32.8 T F (C) 1.27, 1.19 6 2.28, 1.95 14 2.56, 2.44 5 2.0.6 L (J/g) 249 n/a 310 n/a 278 n/a 24916 FWHM (C) 5.25, 4.82 8 4.64,7.02 51 5.65,5.60 .88 5.5.8 Table 4.3 .1: The calculated thermal properties of samples of a 2% concentration.
32 Sample 16 Percent Change (%) Sample 18 Percent Change (%) Sample 21 Percent Change (%) Average T M (C) .23, .39 70 1.20, .85 29 1.31 1.19 14 .9.5 T F (C) .15, .33 120 .38, .35 8 1.97, 1.70 14 .8.8 L (J/g) 216 n/a 228 n/a 309 n/a 255 FWHM(C) 4.26,4.59 8 5.2,4.19 19 5.71,5.86 3 4.97.7 Table 4.3 .2: The calculated thermal properties of sa mples of a 5% concentration Sample 24 Percent Change (%) Sample 27 Percent Change (%) Sample 28 Percent Change (%) Average T M (C) 2.15, 2.21 3 .31, .28 10 .75, .74 1 .91.1 T F (C / ) .03, .04 33 .07, .03 57 .75, .72 4 .3.4 L (J/g) 270 n/a 274 n/a 302 n/a 28217 FWHM(C) 3.94, 3.93 .25 3.18, 3.18 0 3.72, 3.74 .54 3.6.3 Table 4.3 .3: The calculated thermal properties of sa mples of a 10% concentration Table 4.3 .4 : The calculated therm al properties of water The theoretical values are presented when applicable. The values from Table 4.3 .1 4.3 .4 are plotted in Figure 4.3 .1 Figure 4.3 .4 W ater Experiment Percent Change Theoretical Values T M (C) .11, .11 n/a 0 T F (C) 2.69, 2.73 1 n/a L (J/gC) 334 n/a 334 FWHM(C) 7.41, 7.32 1 n/a
33 Figure 4.3 .1: T M as a function of concentration. No trend appears in T M The average melting poin t of all of the nanof luids is .5 1. 7 C The melting point for water is within one standard deviation of this average. Contr ary to the results reported in the literature 17 this study finds that adding nanoparticles to water will not change its melting point.
34 Figure 4.3 .2 : T F as a function of concentration. The temperature at which the sample is fully melted seems to depend on temperature. The samples with a higher concentration of nanoparticles are fully melted at lower temperatures. According to Table 4 3 .1 4 3 .4, none of the averages are within one standard deviation of each other. The average value for T F is 1.0.93 C This is more than one standard deviation for the values obtained for water, 2.69 C and 2.73 C Thus, this study finds that adding nanoparticles to water will lower the temperature at which the sample is fully melted.
35 Figure 4.3 .3 : The latent heat of fusion as a function of concentration There is no correlation found between the latent heat of f usion and concentration. The average latent heat of fusion for all of the nanofluid sa mples is 27030 The measured latent heat of fusion for water is 334 which is not within one standard deviation of the average heat of fusion for a ll of the nanofluids. Thus, this study finds that adding nanoparticles to water will decrease its latent heat of fusion.
36 Figure 4.3 .4 : The FWHM as a function of concentratio n. The FWHM is generally lower for higher concentr ations. The FWHM at 10% is more than one standard deviation from the average of the FWHMs measured for 5%. However, the same relationship does not hold between the samples at 2% and the samples at 5%. The average FWHM for all of the nanofluids is 4.46 1 .10 C The FWHM for water is 6.32C, which is more than one standard deviation away from the average FWHM of all of the nanofluids. Thus, it is concluded that the FWHM was different for nanofluids. The mass of the water is approximately double the mass of the samples. Therefore, the mass of the samples could play a role in the width of the FWHM.
37 The rate of change of the heat flow plots shown in Figure 4.1.1 4.1.3 was calculated using a linear fit with Universal Anal ysis The results are shown in Tabl e 4 3 .5. The rate of change is measured over two tempe rature ranges: from 73C to 30 C and from 15C to 27C. From 73C to 30 C, the samples are in their solid phase. From 15C to 27C, the samples are in their liquid phase. Sample Concentration Concentration Solid C1, x10 4 Solid C2, x10 4 Percent change (%) Liquid C1, x10 4 Liquid C2, x10 4 Percent change (%) 11 2 60 61 2 48 61 27 13 2 47 50 6 21 28 33 14 2 19 92 5 13 49 277 16 5 99 92 7 15 113 653 18 5 84 88 5 63 83 32 21 5 65 67 3 14 14 0 24 10 125 122 2 4.7 20 325 27 10 46 50 9 8.2 10 222 28 10 68 77 13 19 .026 100 Water 104 104 0 31 33 6.45 Table 4.3 .5: The rate of change of the heat flow plots is presented. The slopes are presented for the first cy cle and second cycle for two temperature ranges. C 1 to C 2. Of the s values for T M T F and FWHM had the smallest percent change (0%, 5%, and .88%, respectively) Of the samples with a 5% M and FWHM had the smallest percent change
38 (14%, and .3% respectively) values for T M and T F had the smallest percent chan ge (1% and 4%, respectively) These samples did not have the lowest percent change in rate of change, suggesting that the rate of change of heat flow is not a poor metric for sample evolution. There is also correlation with mass loss, or pan mass to samp le mass ratio. The fact that these samples were more stable suggests that these are the most reliable sources of data for each concentration The thermal parameters for these values are summarized in Table 4.3 .6. Sample Concentration Specific Heat Capacity at 27C (J/gK) Specific Heat Capacity at 23C (J/gK) T M (C) T F (C) Latent Heat of Fusion (J/g) FWHM (C) 14 2 1.2.006 2.6.1 2.6, 2.6 2.6, 2.4 270 5.7,5.6 21 5 2.4.2 4.4.2 1.3, 1 2 2.0, 1.7 310 5.7 ,5.6 28 10 2.5.02 4.1.1 .8, .7 .8, .7 300 3.7, 3.7 Table 4.3 .6: Parameters for Samples 14, 21, and 28. It is possible that the alumina, water, and aluminum pan reacted to change the chemical composition of the nanofluid. The reaction 2Al+6H 2 O 2Al(OH) 3 +3H 2 may have occurred. The reactio n of aluminum and water can occur at room temperature. 3 2 Although the reaction is thermodynamically favorable, it does not usually happen because there is an adherent layer of aluminum oxide on the surface of the aluminum particles. However, if catalyst such as Al 2 O 3 is present, the rate of the reaction increases
39 in the studied temperature range. Thus, less reactive pans may yield more stable data. Raman spectra taken on Sample 16 showed the presence of 2Al(OH) 3 in the pan, suggesting this reaction occu rred. When the nanofluids were studied, it was assumed that the thermal history of the samples would not affect the heat capacity measurement values. Thus, the samples were not treated the same before measurement. Some samples were refrigerated, and some samples were made days before they were measured. Some samples were heated to test for mass loss. Perhaps the thermal history of the nanofluid affected the measurements. 5. Conclusions and Further Experiments The chemical reaction between the Al 2 O 3 water, and aluminum pans rendered the data inconclusive. Some correlations were found between concentration and T F FWHM, and latent heat of fusion. S ince the pans were not chemically inert, however, it is not possible to make a conclusion about nanofluids based on the correlations found in this study. For future experiments, improvements in the experimental method can be made 1) Baseline. In the current study, the same baseline run and sapphire r un is used f or all of the data. The ASTM E 1269 Standard Test Method for Determining the Specific Heat Capacity by Differential Scanning Calorimetry suggests performing a new calibration for each sample. 18 2) Maximizing the sample mass. 3) Minimizing mass l oss during measurement. 4) Sample preparation should be immediately followed by data acquisition. 5) Chemically inert pans.
40 6) Different heating rates.
41 Bibliography 1. J. Buongiorno, Journal of Heat Transfer 128, 240 (2006). 2. A. Diana, Confirmation of Canditure/Art iculation, Queensland University of Technology, 2011. 3. M. J. V. Lourenco, F. J. V. Santos, M. L. V. Ramires, and C. A. Nieto de Castro, Journal of Chemical Thermodynamics 38, 970 (2006). 4. D. W. Oxtoby, H. P. Gillis, and A. Campion, Principles of Modern Chem istry (Thomson Brooks/Cole, Belmont, CA, 2008), p. 1.6. 5. H. Akoh, Y. Tsukasaki, S. Yatsuya, and A. Tasaki, Journal of Crystal Growth 45, 495 (1978). 6. J. A. Eastman, U.S. Choi, S. Li, L.J. Thompson, and S. Lee in Material Research Society Symposium Proceedin g. 7. C. H. Lo, T. T. Tsung, L. C. Chen, C. H. Su, and H. M. Lin, Journal of Nanoparticles Research 7, 313 (2005). 8. C. H. Lo, T. T. Tsung, L. C. Chen, C.H. Su, and H.M. Lin, Journal of Nanoparticles Research 277, 636 (2005). 9. S. Lee, U. S. Choi, S. Li, and J. A. Eastman, Journal of Heat Transfer 121, 280 (1999). 10. T. K. Hong, H. S. Yang, and C. J. Choi, Journal of Applied Physics 97, 64311 1 (2005). 11. S. Zhou and R. Ni, Applied Physical Letters 92, 093123 (2008). 12. C. Kittel, Thermal Physics (W. H. Freeman, New Yor k, NY, 1980), 2 nd edition.
42 13. C. Kittel, Introduction to Solid State Physics (Wiley, Johns & Sons, Inc., 1995), 7 th edition. 14. A. Adamson, Physical Chemistry of Surfaces (Wiley, Johns & Sons, Inc., 1990). 15. S. Wu, D. Zhu, X. Li, H. Li, and J. Lei, Thermochimica Acta 483, 73 (2008). 16. J. W. Moore, C. L. Stanitski, and P. C. Jurs, Chemistry The Molecular Science (Brooks/Cole, Cenage Learning, 2011), 4 th edition. 17. H. Xie, J. Wan, L. and L. Chen, Journal of Material Science Technology 5, 742 (2008). 18. G. W. H. Hohne, W. F. Hemminger, and H. J. Flammersheim, Differential Scanning Calorimetry (Springer Verlag Berlin Heidelberg New York, Germany, 2003), p. 10. 19. ASTM, E 1269, Standard Method for Determining Specific Heat Capacity by Differential Scanning Calorimetry ( West Co nshohocken, PA). 20. S. Kasap and P. Capper, Springer handbook of Electronic and Photonic materials (Springer Science+Business Media, Inc., New York, NY, 2006), p. 400. 21. D. D. Pollock, Thermocouples: Theory and Properties (CRC Press, Inc., Boca Raton, FL, 1991) p. 1. 22. TA Instruments Waters LLC, Thermal Specialty Library ( 2003). 23. W. Wagner, A. Prub, Journal of Physical Chemistry Reference Data 31, No. 2, (2002). 24. J. J. Manya, M. J. Antal Jr., C. K. Kinoshita, and S. M. Masutani, Industria Engineering Chemical Res ource 50, 6470 (2011). 25. S. Zhou and R. Ni, Applied Physical Letters 92, 093123 (2008).
43 26. R. Saidur, K.Y. Leong, and H.A. Mohammad, Renewable and Sustainable Energy Reviews 15, 1646 (2011). 27. C. Choi, H.S. Yoo, and J.M. Oh, Current Applied Physics 8, 710 (200 8). 28. H. Tyagi, P. Phelan, R. Prasher, Journal of Solar Energy Engineering 131(4), 041004 (2009). 29. P. Gabbott, A Practical Introduction to Differential Scanning Calorimetry (Blackwell Publishing Ltd, 2008). 30. D. Lide, CRC Handbook of Chemistry and Physics (C RC Press, Boca Raton, FL, 1996), 2 nd edition. 31. G. P. Johari, Journal of Chemical Physics 107, 10154 (1997). 32. J. Petrovic and G. Thomas, DOE Hydrogen Program (2008).