m{ 1 [Cc [al ENGINEERING Mass transfer - CH2051 Dr. Tran Tan Viét Vice Dean - Faculty of Chemical Engineering HoChiMinh City University of Technology - VNU HCM Email: [email protected] Room: 114 B2Chapter 2 Molecular Diffusion Dr. Tran Tan Viét ap 8) (o\Modes of mass transfer Mass transfer could occur by the following three ways: Q Diffusion is the net transport of substances ina stationary solid or fluid under a concentration gradient. Q Advection is the net transport of substances by the moving fluid. It cannot therefore happen in solids. It does not include transport of substances by simple diffusion. Q Convection is the net transport of substances caused by both advective transport and diffusive transport in fluids.Modes of mass transfer Diffusion (slower) Water Convection (faster) Stirring the water with a spoon creates forced convection. That helps the sugar molecules to transfer to the bulk water much faster.CONTENTS Molecular Convective Pyrat (oa) Mass Transfer Chapter 21. Diffusion phenomena OFick’s law: linear relation between the rate of diffusion of chemical species and the concentration gradient of that species. O Thermal diffusion: Diffusion due to a temperature gradient. Usually negligible unless the temperature gradient is very large. OPressure diffusion: Diffusion due to a pressure gradient. Usually negligible unless the pressure gradient is very large.1.Diffusion phenomena Whenever there is concentration difference in a medium, nature tends to equalize things by forcing a flow from the high to the low concentration region. Water Sait Before After The molecular transport process of mass is characterized by the general equation: Rate of transfer process = driving force resistance1. Diffusion phenomena Example of Mass Transfer Processes Consider a tank that is divided into two equal parts by a partition. Initially, the left half of the tank contains nitrogen NV, gas while the right half contains H, at the same temperature and pressure. When the partition is removed the Vv, molecules will start diffusing into the air while the H, molecules diffuse into the NV>. If we wait long enough, we will have a homogeneous mixture of V2 and H, in the tank.The temperature and pressure are such that . the diffusion coefficient is 0.1 cm?/sec. Find nitrogen | hydrogen) the molar average velocity v’, the mass average velocity v, the volume average velocity v’ at the average concentration in an the system. ‘ The volume in this system does not move, so v’ = 0. If the gases are ideal, the molar concentration is constant, so v’ = 0. io —¢y)) Diffusion in the thin-film D (SoG = ot tet, = 0.02. cm/sec For nitrogen at an average i” ¢ 10° 0.5 concentration of 0.5 ¢ D (Cy) ~ Cr) _9-1/0-1 For hydrogen at an average ea =— (=) =-0.02 cm/s , Be] cy ) 10 ( 05> em’) concentration of 0.5 ¢ eM, —__—_(0.5(28) — = = ——————__. = 0.933 Mass fraction of nitrogen ¢,M,+e,M, 0.5(28) +0.5(2) 952) 0.067 Mass fraction of hydrogen 0.5(28) + 0.5(2) -933(0.02) + 0.067(20:02) = 0.017 cm/sec1. Diffusion phenomena * Liquid in open pail of water evaporates into air because of the difference in concentration of water vapor at the water surface and the surrounding air. « Adrop of blue liquid dye is added to a cup of water. The dye molecules will diffuse slowly by molecular diffusion to all parts of the water.Understanding diffusion in one dimensionSolvent B concentration of A concentration of A is high is low Mass transfer by diffusion occurs when a component in a stationary solid or fluid goes from one point to another driven by a concentration gradient of the component.7 a * SolventBe ¢ SoluteA . ° ° : concentration of A is the same everywhere Diffusion is the macroscopic result of random thermal motion on a micoscopic scale (Brownian motion). jt occurs even when there is no concentration gradient (but there will be no net flux).Liquids A and B are separated from each other. Separation removed. A goes from high concentration of A to low concentration of A. B goes from high concentration of B to low concentration of B. Molecules of A and B are uniformly distributed everywhere in the vessel purely due to DIFFUSION. Equilibrium is reachedA diffusion problem that occurs in the field of microelectronics is the oxidation of silicon according to the reaction Si + O,— SiO,. When a slab of the material is exposed to gaseous oxygen, the oxygen undergoes a first-order reaction to produce a layer of the oxide. The task is to predict the thickness d of the very slowly- growing oxide layer as a function of time ¢. Oxygen Mies Cag atZ=0 Silicon oxide layer Ca=CagatZ=6Mathematical modelling of steady-state one dimensional diffusive mass transfer+ Molecular diffusion or molecular transport can be defined as the transfer or movement of individual molecules through a fluid by mean of the random, individual movements of the molecules. @ Fre schematic diagram of molecular diffusion process G ° if a greater number of A molecules near point (1) than at (2), then since molecules diffuse randomly in both direction, more A molecules will diffuse from (1) to (2) than from (2) to (1). + The net diffusion of A is from high to low concentration regions.For mass transfer occurring only in z-direction i, =-BDy, (1) dz Mixture Ca of A&B What is J,? Ja Ca + dC, dzJ,=-D —e concentration gradient of A AB ~~~ Ain z-direction | ‘ Unit: mass (or moles) per x \ \ + diffusion coefficient (or diffusivity) of Ain B | \ volume per distance diffusion flux of A in relation to the bulk motion in z-direction Unit: mass (or moles) per area per time What is the unit of diffusivity?For dissolved matter in water: D = 105 cm2/s For gases in air at 1 atm and at room temperature: D =0.1 to 0.01 cm2/s Diffusivity depends on the type of solute, type of solvent, temperature, pressure, solution phase (gas, liquid or solid) and other characteristics.Molecular diffusion of Helium in Nitrogen: A mixture of He and Nz gas is contained in a pipe (0.2 m long) at 298 K and 1 atm total pressure which is constant throughout. The partial pressure of He is 0.60 atm at one end of the pipe, and it is 0.20 atm at the other end. Calculate the flux of He at steady state if Daz of He-N, mixture is 0.687 x 10-4 m?/s. Solution: Use Fick’s law of diffusion given by equation (1) as Ja =- Dap aC dz Rearranging the Fick’s law and integrating gives the following:Zz Cao 2 [4s dz = -[Pas dC, ——— 4 Cay At steady state, diffusion flux is constant. Diffusivity is taken as constant. Therefore, equation (2) gives Ja(Z2—- 21) = - Dag (Caz— Car) 8) Dag is given as 0.687 x 10+ m/s (Z)—2Z,) is given as 0.2m (Cag—Cay) =?Even though C, is not given at points 1 and 2, partial pressures are given. We could relate partial pressure to concentration as follows: _h——-> Number of moles of A V~—+ Total volume __-+ Absolute temperature t "Gas constant Partial pressure of A Combining the above we getEquation (3) can therefore be written as (Pao — Pat) Jn(22— 21) = - Dag which gives the flux as _. (Paz— Pas) “OS RT (2: 2,) (0.6 — 0.2) x 1.01325 x 105 Pa Ja = - (0.687x10- m2/s) — (8314 J/kmol.K) x (298 K) x (0.20—0) m Ja = 5.63 x 10° kmol/m?.sSummary: Modelling diffusion in z-directio! Jn =-Dap (Caz— Cai) =-Das (Pa2— Pat) (Z.-Z,) RT(Z,—Z,) Ca, and pa, at Z, and Crp and Ppp at Zz remain unchanged with time (steady state). Dagis constant Longitudinal flow: Flow area perpendicular to the flow direction is a constant.Derivation of Dag = Dg, under certain conditions Dap: diffusivity of Ain B Dga: diffusivity of B inAConsider steady-state diffusion in an ideal mixture of 2 ideal gases A & B at constant total pressure and temperature. Molar diffusive flux of Ain B: J, = - Dap SA (4) Zz Molar diffusive flux of Bin A: Jp = - Dea _1C3 _ dz Jaand Js : molar diffusive flux of A and B, respectively (moles/area.time) (5) C, and Cg: concentration of A and B, respectively (moles/volume) Dag and Dga: diffusivities of A in B and of B inA, respectively z: distance in the direction of transferSince the total pressure remains constant, there is no net mass transfer. That is, Jat Jg=0 6) For an ideal gas mixture at constant pressure, Cy, + Cg = pa/RT + pp/RT = P/RT = constant (7) Substituting (6) and (7) in (4) and (5), we get Therefore, dC, + dC, =0 Dag = Daa = D (Say) Therefore, (4) & (5) give Ja=-D = — ee iz -p ey dzThis is known as equimolar counter diffusion. This describes the mass transfer arising solely from the random motion of the molecules (i.e., only diffusion) It is applicable to stationary medium or a fluid in streamline flow. J,=-D 108 (8a) =-D—,— — (8a A dz dc J3=-D——2— — @) dz Chapter 2Mathematical modelling of steady-state one dimensional convective mass transferDiffusion is the net transport of substances in a stationary solid or fluid under a concentration gradient. Advection is the net transport of substances by the moving fluid, and so cannot happen in solids. It does not include transport of substances by simple diffusion. Convection is the net transport of substances caused by both advective transport and diffusive transport in fluids. J,is the diffusive flux described by Fick’s law, and we have already studied about it. Let us use N, to denote the total flux by convection (which is diffusion plus advection.Molar diffusive flux of A in B: Ja = = Dap * (4) Z The velocity of the above diffusive flux of Ain B can be given by Ja (mol/m?.s) (9) Vaditfusion (M/S) = C, (mol/m) A The velocity of the net flux of Ain B can be given by Na (mol/m?.s) (10) Va.convection (M/S) = Cq (mol/m?) The velocity of the bulk motion can be given by (Na+ Ng)(mol/m?.s) = ___(41) (C;) (mol/m?) he pte 2™ Total concentration Vbuik (M/s) =Va.convection = Vadittusion + Vbutk Multiplying the above by Ca, we get CaVaconvection = CaVaitusion * Ca Vouk Using equations (9) to (11) in the above, we get —— (12) dC, +Cq (Na+ Np) dz Cy (13) “MABLet us introduce partial pressure p, into (13) as follows: = Reo, JPA (14a) Cay RT ése Me Pe (14b) “~~ Total pressure Total number of moles a Using (14a) and (14b), equation (13) can be written as (15) D d Na=- PA | (8314 J/kmol.K) (293 K) (0.1524 m) (0.988 atm) = 1.595 x 10°77 kmol/m2.sDiffusion in a tube with change in " path length: AP) Diffusion of water vapour in a narrow tube is occurring as in Example 6.2.2 TI * |2 under the same conditions. However, * Zo as shown in the figure, at a given “|| time t, the level is z from the top. As diffusion proceeds the level drops slowly. Drive the equation for the time t; for the level to drop from a starting point z) m at t= 0 to zp att = te s as shown. Water (A) Chapter 2 45Solution: Since the level drops very slowly, we assume pseudo-steady-state condition. Therefore, equation (19) applies in which (z, — z,) must be replaced by z. Dag P Na= — (Pas - Paz) RTZ Pei — (22)Suppose that the level reduce by dz in dt time. Mass balance yields the following: (Na X Ax dt) x Ma = pa x (Ax dz) —— (23) where A is the cross-sectional TI area of the tube, M, is the 2 Z molecular mass of water and F P, is the density of water.Combining equations (22) and (23), we get 0, dz = Rene (Pa ~Pag)dt (24) 'B,LM Rearranging (24), we get Ze te Ma Dag P Is it okay to take > See - dt Pr fe = pen (Pat - Paz | ae Zo 0 constants? (27-20?) _ My DagP = t Pay =RT Pauw At Paz) te t= Pa RT Pei (2? -Z0°) ___ (25) 2 Mg Dag P (Pai - Pao) This above equation is used to experimentally determine diffusivity DagEvaporation of a Naphthalene sphere: A sphere of naphthalene having a radius of 2.0 mm is suspended in a large volume of still air at 318 K and 1.01325 x 10° Pa (1 atm). The surface temperature of the naphthalene can be assumed to be at 318 K and its vapour pressure at 318 K is 0.555 mm Hg. The Dag of naphthalene in air at 318 K is 6.92 x 10°° m2/s. Calculate the rate of evaporation of naphthalene from the surface. If the radius of the sphere decreases slowly with time, drive the equation for the time taken (te) for the sphere to evaporate completely. Chapter 2 Par 49Solution: All this time the mass transfer of Ais taken as Ny (in moles/m?.s) and it is assumed to remain constant for systems at steady-state or pseudo steady-state while carrying out the integrations concerned. For the system considered here, we cannot do that since the area across which the mass transfer occurs vary as the radius changes. Paz Therefore, we use the following definition: Mass transfer Mass transfer per time per area per 5 “ : Ee is time FN iat 5 (26) “\. area of mass transfer For the given system, equation (26) gives Na N,= — (27) Amr?Since the given system is a case of A diffusing through stagnant, non-diffusing B, we could start from equation (15). Na= ~—eae SPA 5 PA (Na+ No) (18) RT dz P Setting Ng = 0 and rearranging the above gives: D dp, Par Neo ‘AB A RT (1-pa/P) dz Using equation (26) and replacing dz by dr, the, above could be written as follows: Na Das dpa Na= “am? =” RT (1-p,/P) dr which can be rearranged to giveNa dre _. Dag dpa 40 e RT (1- p,/P) Integrating the above between r, and some point r, (which is a large distance away) gives the following: Par r Zz mf 8 __ Das [dpa 4m 2 RT J} (1-p,/P) y Pat Ma | 1 _ 1] _ DagP_ |P- Pas 47 [my tp RT [P-PatSince ry >> 1, 1/1, is much smaller than 1/r, in equation (28), and therefore neglected. So, equation (28) gives the following: Na _ —_ DagP [Est - P—Pay Amr, RT Using equation (27) and the definition of pgim, Paz the above could be written as = AL L _DysP_— Pat - Par ___ (29) A” 4m? RTr; Pe. - Equation (29) gives the rate of evaporation of naphthalene from the surface at r, radius.Data provided are the following: Dag = 6.92 x 10 m/s; P=1 atm; T=318 kK; r,=2mm = 2/1000 m; Pai = 0.555/760 atm (saturated vapour pressure); Paz = 0 atm (no naphthalene vapour far away from the sphere) Substituting the data provided, we get the following: (0.555/760 — 0 ) a nya RRRANYW 7 0:9996 atm In{((1 - 0 / (1 — 0.555/760 )] Pam = (6.92x10-° m2/s)(1.01325x105 Pa)(0.555/760 - 0) atm (8314 J/kmol.K) (318 K) (2/1000 m) (0.9996 atm) = 9.686 x 10% kmol/m?.sNow, we need to drive the equation for the time taken (t-) for the sphere to evaporate completely. That is, the radius of the sphere is r, att=0 and itis zero att = tp. Since the radius of the sphere is said to decrease slowly, we assume pseudo-steady- state condition. Therefore, equation (29) applies in which r, is replaced by « as follows: Paz — OM, _ DasP Pat > Pao 30 a” 4t1a? ~ RTa Pam — 6h N. Sphere is 5 enlarged:Suppose that the surface reduce by da in dt time. Mass balance yields the following: (Na XA x dt) x My = Pa X (AX (-da)) —— (31) where A is the surface area of the sphere at radius a, M, is the molecular mass of the Paz sphere and py is the density of the sphere. Scher! is 5 enlarged:Combining equations (30) and (31), we get Ma Dag P (Par - Paz) dt ~pada= RT Pp.im Rearranging and integrating, we get 0 t Paz = MaDas P - dt - Pa [oes = RT Pen (Pat - Paz) tr 0 a Pat . Is it okay to assume that p,, and pap remain constants as the radius of the sphere (a) reduces from r, to 0? Sphere enlarged.Integration gives the following results: (72-0) = Ma DasP - t a =RT Paw PA" Pag) te _ Pat? RT Pam (32) t= ‘ 2 My Dag P (Pat - Paz) The above gives the time taken for the sphere to vanish completely at a slow rate of evaporation. Par Pat Sphere is enlarged.Estimating diffusivitiesDiffusivities for different systems could be estimated using the empirical equations provided in the following slides as well as those provided in other reference texts available in the library and other sources.An example at 1 atm and 298 K: Diffusivity (cm2/s) Air-C,H;0H Air-benzene(4) Diffusivities of gases in the case of absence of experimental data Holman gives a semiempirical equation by T —Absolute temp, K; P— Absolute pressure, atm Mg, Mg - molar weights of A, B, g/mol Vy, Vg - molecular volumes of gases A, B , define by sum of molecular volumes of all elements . In case of benzene, naphtalene or anthracen exist, the molecular volumes must be minus structure constance (benzene = 15; Naphtalen 30; anthracen 47,5) Dag is proportional to 1/P and T1-76Table 24.5 Atomic volumes for complex molecular volumes for simple substances! Atomic volume, Atomic volume, Element in cm*/g mol Element in cm/gmol Bromine 27.0 Oxygen, except as noted below 74 Carbon 148 Oxygen, in methyl esters o1 Chlorine 21.6 Oxygen, in methyl ethers 99 Hydrogen 37 Oxygen, in higher ethers lodine 37.0 and other esters 110 Nitrogen, double bond 15.6 Oxygen, in acids 120 Nitrogen, in primary amines 10.5 Sulfur 25.6 Nitrogen, in secondary amines 12.0 'G. Le Bas, The Molecular Volumes of Liquid Chemical Compounds, Longmans, Green & Company, Lid., London, 1915. .. cat corrections are recommended: for three-membered ring, as ethylene oxide deduct 6 for four-membered ring, as cyclobutane deduct 8.5 for five-membered ring, as furan deduct 11.5 for pyridine deduet 15 for benzene ring deduct 15 for naphthalene ring deduct 30 for anthracene ring deduct 47.5 Chapter 2Table 24.4 Molecular volumes at normal boiling point for some commonly encountered compounds Compound Molecular volume, in cm*/g mol Compound Molecular volume, in cm/g mol Hydrogen, H> Oxygen, O2 Nitrogen, N> Air Carbon monoxide, CO Carbon dioxide, CO, Carbonyl sulfide, COS Sulfur dioxide, SO, 14.3 25.6 31.2 29.9 30.7 34.0 SLS 44.8 Nitric oxide, NO Nitrous oxide, NxO Ammonia, NH; Water, H2O. Hydrogen sulfide, HoS Bromine, Bry Chlorine, Cl, lodine, Chapter 2 23.6 36.4 25.8 18.9 32.9 53.2 48.4 51 103717 (i+4) D.= My, Mz ap = ere) ] Das - diffusivity in cm?/s P - absolute pressure in atm M, - molecular weight T - temperature in K Vi - sum of the diffusion volume for component i Dag is proportional to 1/P and T1.’5Table 24.3 Fuller, Schettler, and Giddings Atomic and structure diffusion-volume increments, v Atomic diffusion volumes for use in estimating Dag by method of Cc 16.5 | 19.5 H 1.98 Ss 17.0 oO 5.48 Aromatic ring —202 N 5.69 Heterocyclic ring —202 Diffusion volumes for simple molecules, v H, 7.07 Ar 16.1 HO 12.7 D2 6.70 Kr 22.8 CCIF, 114.8 He 2.88 co 18.9 SF, 69.7 N> 17.9 oO; 26.9 Ch 37.7 O> 16.6 NO 359 Bry 67.2 Air 20.1 NH; 14.9 SO 41.1Normal butanol (A) is diffusing through air (B) at 1 atm abs. Using the Fuller et al. method, estimate the diffusivity Dy, for the following temperatures with the experimental data: (a) For 0 °C. (b) For 25.9 °C. (c) For 0 °C and 2.0 atm abs.Solution For part (a), P= 1.00 atm, T= 273 + 0 = 273 K, M, (butanol) = 74.1, Msg (air) = 29. From Table 6.2-2, D v4 = 4(16.5) + 10(1.98) + 1(5.48) = 91.28 (butanol) D vp = 20.1 (air) Substituting into Eq. (6.2-45), _ 1.0 x 10°7(273)"5(1/74.1 + 1/29)" 1.0[(91.28)" + (20.1)4P = 7.73 x 10° m/s AB This values deviates +10% from the experimental values of 7.03 x 10° m?/s from Table 6.2-1Solution For part (b), T = 273 + 25.9 = 298.9. Substituting into Eq. (6.2-45), Dag= 9.05 x 10° m2/s. This values deviates by +4% from the experimental value of 8.70 x 10° m2/s For part (c), the total pressure P = 2.0 atm. Using the value predicted in part (a) and correcting for pressure, Dag = 7.73 x 10°9(1.0/2.0) = 3.865x10 m2/sFor very large spherical molecules (A) of 1000 molecular weight or greater diffusing in a liquid solvent (B) of small molecules: 9.96 x 10°12 T Dag = —————————___ applicable yu VA"? for biological Dag - diffusivity in cm2/s solutes such i - temperature in K as proteins y - viscosity of solution in kg/m s Va - solute molar volume at its normal boiling point in m°/kmol Dag is proportional to 1/y and TFor smaller molecules (A) diffusing in a dilute liquid solution of solvent (B): 1.173 x 10-2 (® Mg)? T Dag = 7 co Up Va" applicable Das - diffusivity in cm?/s for biological Mg - molecular weight of solvent B solutes T - temperature in K H - viscosity of solvent B in kg/m s V, - solute molar volume at its normal boiling point in m°/kmol ® - association parameter of the solvent, which 2.6 for water, 1.9 for methanol, 1.5 for ethanol, and so on Dag is proportional to 1/Ug and TFor smaller molecules (A) diffusing in a dilute liquid solution of solvent (B): 8.928 x 101° T (1/n, + 1/n.) De = “e (1A, + 10) D°ag is diffusivity in cm?/s n, is the valence of cation n_ is the valence of anion A, and A_are the limiting ionic conductances in very dilute solutions T is 298.2 when using the above at 25°C Dag is proportional to TEXAMPLE 5 Estimate the liquid diffusion coefficient of ethanol, C>HsOH, in a dilute solution of water at 10°C. ‘The molecular volume of ethanol may be evaluated by using values from Table 24.5 as follows: Veunson = 2Ve + 6Vy + Vo Vesn,on = 214.8) + 6(3.7) + 7.4 = 59.2em*/mol At 10°C, the viscosity of a solution containing 0.05 mol of alcohoV/liter of water is 1.45 centipoises; the remaining parameters to be used are T= 283K Pp forwater = 2.26 Mp for water = 18 Substituting these values into equation (24-52), we obtain _ (74% 10-8(2.26 x 18)? (®) DosHsOH-H20 = way =7.96 x 10° em*/s_—_(7.96 x 107". 145, Chapter 2Diffusion in solids Diffusion in solids are occurring at a very slow rate. However, mass transfer in solids are very important. Examples: Leaching of metal ores Drying of timber, and foods Diffusion and catalytic reaction in solid catalysts Separation of fluids by membranes Treatment of metal at high temperature by gases.Diffusion in solids Diffusion in solids occur in two different ways: - Diffusion following Fick’s law (does not depend on the structure of the solid) - Diffusion in porous solids where the actual structure and void channels are importantDiffusion in solids following Fick’s Law Start with equation (13) from Set 1: Na = -Dag Aa, Sa (Ng +5) iz 1 (13) Bulk term is set to zero in solids Therefore, the following equation will be used to describe the process: Ny = -Dyg 2% (33) dzDiffusion in solids following Fick’s Law Pure molecular Pure knudsen diffusion diffusion AQ A BY Ss dove Random porous Knudsen + molecular Taira diffusion 8 tot yt Deze? Dx Figure 24.3 Types of porous diffusion, Shaded areas represent nonporous solids Chapter 2 7Diffusion in solids following Fick’s Law The Knudsen number, Kn, given by A __ mean free path length of the diffusing species Kn= = pore pore diameter ate, em [8KNT aM, Via = 4850dpore VimDiffusion through a slab Applying equation (33) for steady-state diffusion through a solid slab, we get —— (34) Zo- Zs where N, and Dag are taken as constants. Drive (34) and compare it with heat conduction equivalent. Cas Crz Zy2yRelating the concentration and solubility The solubility of a solute gas in a solid is usually expressed by the notation S. Unit used in general is the following: m solute at STP m solid . atm partial pressure of solute Relationship between concentration and solubility: c= S Pa kmol solute /m? solid 22.414 where Py is in atm STP of 0°C and 1 atmRelating the concentration and permeability The permeability of a solute gas (A) in a solid is usually expressed by the notation P,y. in m? solute at STP (0°C and 1 atm) diffusing per second per m? cross-sectional area through a solid 1 m thick under a pressure difference of 1 atm. Unit used in general is the following: m solute at STP . 1 m thick solid s . m* cross-sectional area . atm pressure difference Relationship between concentration and permeability: Py = Dag S where Dag is in m?/s and S is in m3/m3.atmDiffusion of Hj through Neoprene membrane: The gas hydrogen at 17°C and 0.010 atm partial pressure is diffusing through a membrane on vulcanized neoprene rubber 0.5 mm thick. The pressure of Hz on the other side of neoprene is zero. Calculate the steady-state flux, assuming that the only resistance to diffusion is in the membrane. The solubility S of Hz gas in neoprene at 17°C is 0.051 m® (at STP of 0°C and 1 atm)/mé solid. atm and the diffusivity Dag is 1.03 x 10°? m2/s at 17°C. Answer: 4.69 x 10°12 kmol H2/m2.s Chapter 2Diffusion through a packging film using permeability: Apolythene film 0.00015 m (0.15 mm) thick is being considered for use in packaging a pharmaceutical product at 30°C. If the partial pressure of O2 outside the package is 0.21 atm and inside it is 0.01 atm, calculate the diffusion flux of O2 at steady state. Assume that the resistances to diffusion outside the film and inside are negligible compared to the resistance of the film. Permeability of O2 in polythene at 303 K is 4.17 x 10-1? m? solute (STP)/(s.m2.atm.m). Answer: 2.480 x 10°12 kmol O2/m2.s Would you prefer nylon to polythene? Permeability of O, in nylon at 303 K is 0.029 x 10-1? m3 solute (STP)/(s.m2.atm.m). Support your answer. Chapter 2Diffusion through a cylinder wall Applying equation (33) for steady-state diffusion through a cylinder wall of inner radius r, and outer radius r, and length L in the radial direction outward, we get Mass transfer _Mass transfer per time per area per “ ; n dc time \y-_* -. A 35 Na= dar Das dr (39) ‘\ Area of mass transfer Re= 2mL _2T1L Dag(Car- Cao) Caz) In(ry/ 14) Drive (36) and compare it with heat conduction equivalent.Diffusion through a spherical shell Applying equation (33) for steady-state diffusion through a spherical shell of inner radius r, and outer radius r, in the radial direction outward, we get Mass transfer _Mass transfer per time per area per “ ; n dc time “yy -—* -. A 37 Na 4 Das dr (37) ‘ \ Area of mass transfer n= Atty Dap(Cai- Caz) (2-14) Drive (38) and compare it with heat conduction equivalent. —— (38)Fick’s First Law of Diffusion is written as follows when C, is only a function of z: dC, Jn = - Dag . —— (1) Fick’s First Law is written as follows when C, is a function of z as well as of some other variables such as time: oC, oz Jn =- Dap (39) Observe the use of ordinary and partial derivatives as appropriate.Ja, in Ja, out AC, py we we z z+Az A: cross-sectional area Ma: molecular weight of species A Mass flow of species A into the control volume = Ja, in XAX My where = aC, Ja, in = - Dap i oz at z Mass flow of species A out of the control volume = Ja, out XAX My where aCy Ja, out = - Dag oz at z+AzJa, in AC, Ja, out py YO we z+Az A: cross-sectional area Ma: molecular weight of species A Accumulation of species A in the control volume = FER x (Ax Az) x Ma Mass balance for species A in the control volume gives, Ja, in A Ma = Ja, out A Ma + a (AAz) MaMass balance can be simplified to aC; ‘a = - Dap ey oz - Dag 2CA_ oz atz at z+Az The above can be rearranged to give at Ae (40)In the limit as Az goes to 0, equation (40) is reduced to rn _p., Cn (41) at éz2 which is known as the Fick’s Second Law. Fick’s second law in the above form is applicable strictly for constant Dag and for diffusion in solids, and also in stagnant liquids and gases when the medium is dilute in A.Fick’s second law, applies to one-dimensional unsteady-state diffusion, is given below: 2 aC, = Bs &C, at az? Fick’s second law for one-dimensional diffusion in radial direction only for cylindrical coordinates: a, _ Das 4 |, ay at ~ r or or (42) (43) Fick’s second law for one-dimensional diffusion in radial direction only for spherical coordinates: ay _ Das 2 [nS - ar | — 4) at Pr arFick’s second law, applies to three-dimensional unsteady-state diffusion, is given below: 12 2, 12 aCa == (Dig PC, = PCa 7 PCy —(45a) at ax? oy? az? Fick’s second law for three-dimensional diffusion in cylindrical coordinates: OCy _ Das at r £ (45b) é - aCy + or or a | acy a0| rae é : aC, oz oz Fick’s second law for three-dimensional diffusion in spherical coordinates: ay _ Das | 2 Rr aCy 2 1 \2 sind aCy + 1 Oy at Pr lor or sin8 06 08 sin?@ ah — (45c)12 Cy De PCy (42) at oz? z=0 z20 Z=00 Initial condition: Cy = Cy att0andz=0 Ca = Cao att20andz— ooIntroducing dimensionless concentration change: Ca- Cro Use y= Cas 7 Cro and transform equation (42) to the following: oY PY = Dap at ez? oY aC, / at where = —_ at Cas - Cro ey #C,/ az? a2? Cas— CaoIntroducing dimensionless concentration change: Ca- Cro Use y= ——A ~A0 Cas ~ Cro and transform the initial and boundary conditions to the following: Initial condition: Cy =Cay becomes Y =0 attsOandz20 Boundary condition: Cya=Cas becomes Y =1 att20andz=0 Cya=Cyg becomes Y = 0 att=0andz— ooSolving for Y as a function of z and t: oY 2 & _p,, & at ez? Initial condition: Y=0 atts Oandz20 Y=1 att2>0andz=0 Boundary condition: Y=0 att20andz—0 Since the PDE, its initial condition and boundary conditions are all linear in the dependent variable Y, an exact solution exists.Non-dimensional concentration change (Y) is given by: (46) where the complimentary error function, erfc, is related to the error function, erf, by x erfc(x) = 1-erfx) =1- =| exp(-o) de 0 (47)A little bit about error function: - Error function table is provided (take a look). - Table shows the error function values for x values up to 3.29. - For x > 3.23, error function is unity up to five decimal places. - For x > 4, the following approximation could be used: exp(-x) erf(x) = 1- ) VXExample 3.11 : Determine how long it will take for the dimensionless concentration change (Y) to reach 0.01 ata depth 1 m in a semi-infinite medium. Assume Dag = 0.1 cm/s. Solution: Starting from (46) and (47), we get Y=1- erf —_—__ 2aDigt Using Y = 0.01, z = 1 m (= 100 cm) and Dag = 0.1 cm/s, we get erp | 8°] =1-001=0.99 2V0.1xt 100 24.8214 2V01xt t=2.09hGet back to (46), and determine the equation for the mass flux from it. Jy == Dy, Oa az atz Ja= VDag/ tt exp(-z2/4Dagt) (Cag-Cao) (48) Flux across the interface at z = 0 is Ja , = VDag/tt — (Cas-Cao) (49) tz=0Exercise: Determine how the dimensionless concentration change (Y) profile changes with time in a semi-infinite medium. Assume Dag = 0.1 cm?/s. Work up to 1 m depth of the medium. Solution: Starting from (46) and (47), we get Zz Y=1- ef | ——— 2 DigitLet us get the complete profile using MATLAB which has a built- in error function. clear all DAB=0.1; %em2/s t=0; for i= 1:1:180 %in min t(i) = i*60; z= [0:1:100]; %em x=2/(2 * sqrt(DAB*t(i))); Y(;,i) = 1 - erf(x); end plot(z,Y) xlabel('z (cm)') ylabel('Non-dimensional concentration, Y') grid pause plot(t/3600,Y(100,:)) xlabel('t (h)') ylabel('Y at_ z= 100 cm') grid Chapter 2 102]os 869 6 Oo 6 SO OS A ‘eBueys uonenuesuos jeuolsuewip-uon, z(cm) 103, Chapter 20.035 0.03 0.025 Yat z=100cm ° 2 0,005 L i i 15 t(h) Chapter 2 104Diffusion in semi-infinite medium: In gas: Dag = 0.1 cm/s Time taken for the dimensionless concentration change (Y) to reach 0.01 at a depth 1 mis 2.09 h. In liquid: Dag = 10° cm/s Time taken for the dimensionless concentration change (Y) to reach 0.01 at a depth 1 mis 2.39 year. In solid: Dag = 10° cm/s Time taken for the dimensionless concentration change (Y) to reach 0.01 at a depth 1 mis 239 centuries.Chapter 2 Molecular Diffusion Dr. Tran TAn Viét aH p 8) (o\