Topic 3: Solved Problems
Full step-by-step analytical resolutions for the 6 official problem set exercises on Diffusion in Solids and Mass Transport, formatted with rigorous 4-phase pedagogical derivations.
| $z$ | $\text{erf}(z)$ | $z$ | $\text{erf}(z)$ | $z$ | $\text{erf}(z)$ | $z$ | $\text{erf}(z)$ |
|---|---|---|---|---|---|---|---|
| 0.00 | 0.0000 | 0.25 | 0.2763 | 0.70 | 0.6778 | 1.30 | 0.9340 |
| 0.01 | 0.0113 | 0.30 | 0.3286 | 0.75 | 0.7112 | 1.40 | 0.9523 |
| 0.02 | 0.0226 | 0.35 | 0.3794 | 0.80 | 0.7421 | 1.50 | 0.9661 |
| 0.03 | 0.0338 | 0.40 | 0.4284 | 0.85 | 0.7707 | 1.60 | 0.9763 |
| 0.04 | 0.0451 | 0.45 | 0.4755 | 0.90 | 0.7969 | 1.70 | 0.9838 |
| 0.05 | 0.0564 | 0.50 | 0.5205 | 0.95 | 0.8209 | 1.80 | 0.9891 |
| 0.10 | 0.1125 | 0.55 | 0.5633 | 1.00 | 0.8427 | 1.90 | 0.9928 |
| 0.15 | 0.1680 | 0.60 | 0.6039 | 1.10 | 0.8802 | 2.00 | 0.9953 |
| 0.20 | 0.2227 | 0.65 | 0.6420 | 1.20 | 0.9103 |
Problem 1
Problems T3_Diffusion.pdf #1
Fick's 2nd Law · erf(z)
Click to Collapse ▲
Cementation of Grade 1018 Steel Gear
Original statement · Problems T3_Diffusion.pdf
- Semi-Infinite Medium: The penetration depth $x = 0.6\text{ mm}$ is small compared with the size of the part, allowing one-dimensional semi-infinite modeling [Session 5 Slide 20].
- Constant Surface Concentration: The carburizing atmosphere maintains $C(0, t) = C_s = 1.0\text{ wt}\%$ for $t > 0$.
- Homogeneous Initial State: The AISI 1018 steel core has initial carbon concentration $C(x, 0) = C_0 = 0.18\text{ wt}\%$.
- Parameters: $C_0 = 0.18\text{ wt}\%$, $C_s = 1.00\text{ wt}\%$, $C_x = 0.30\text{ wt}\%$, $x = 0.6\text{ mm} = 6.0 \times 10^{-4}\text{ m}$, $D = 1.28 \times 10^{-11}\text{ m}^2/\text{s}$ at $T = 927^\circ\text{C}$.
In non-steady-state diffusion into a semi-infinite solid with constant surface composition, Fick's Second Law yields [Session 5 Slide 20]:
$$\frac{C_s - C_x}{C_s - C_0} = \text{erf}\left(\frac{x}{2\sqrt{Dt}}\right) = \text{erf}(z)$$We evaluate the concentration ratio, locate the corresponding argument $z$ in the official error function table via exact linear interpolation, and solve for time $t$:
$$z = \frac{x}{2\sqrt{Dt}} \implies t = \frac{x^2}{4 \cdot z^2 \cdot D}$$1. Concentration Ratio:
$$\frac{C_s - C_x}{C_s - C_0} = \frac{1.00 - 0.30}{1.00 - 0.18} = \frac{0.70}{0.82} \approx 0.8536585 \implies \text{erf}(z) = 0.8536585$$2. Linear Interpolation in Error Function Table [Slide 21]:
- For $z_1 = 1.00 \implies \text{erf}(z_1) = 0.8427$
- For $z_2 = 1.10 \implies \text{erf}(z_2) = 0.8802$
3. Time of Carburization:
$$x^2 = (6.0 \times 10^{-4}\text{ m})^2 = 3.60 \times 10^{-7}\text{ m}^2$$ $$4 \cdot z^2 \cdot D = 4 \times 1.05930 \times (1.28 \times 10^{-11}\text{ m}^2/\text{s}) = 5.4236 \times 10^{-11}\text{ m}^2/\text{s}$$ $$t = \frac{3.60 \times 10^{-7}\text{ m}^2}{5.4236 \times 10^{-11}\text{ m}^2/\text{s}} \approx \mathbf{6638\text{ s}} = \mathbf{1.844\text{ h}}$$Discrepancy with the official solution. The official key gives $t = 6636\text{ s}$. Linear interpolation in the table gives $z = 1.0292$ and $t = 6637.6\text{ s} \approx 6638\text{ s}$ (the exact inverse error function, $z = 1.0271$, would give $6665\text{ s}$ because the table is linear between $1.00$ and $1.10$). The key is $0.02\%$ below the interpolated value; this is a rounding difference, and the interpolated value is reported.
Dimensional Check: $[t] = \frac{\text{m}^2}{\text{m}^2/\text{s}} = \text{s}$. Physical Insight: About $1.84\text{ hours}$ at $927^\circ\text{C}$ bring the carbon content at $0.6\text{ mm}$ to $0.30\text{ wt}\%$. Carbon is small, so it diffuses interstitially [Slide 13]; in the surface layer it blocks the movement of dislocations and hardens the steel for wear resistance, while the core stays tough, and the excess carbon leaves residual compressive stresses [Slide 30].
Problem 2
Problems T3_Diffusion.pdf #2
Arrhenius · Semiconductor Doping
Click to Collapse ▲
Diffusion of Aluminum in Monocrystalline Silicon
Original statement · Problems T3_Diffusion.pdf
- Volume Diffusion in a Single Crystal: A single crystal has no grain boundaries, so diffusion takes place through the volume, the slowest path ($D_{surface} > D_{grain\ boundary} > D_{volume}$) [Slide 25].
- Arrhenius Behavior: The temperature dependence of $D$ obeys Arrhenius kinetics with constant $D_0$ and $E_D$ [Slide 26].
- Parameters: $D = 10^{-14}\text{ m}^2/\text{s}$, $D_0 = 1.55 \times 10^{-4}\text{ m}^2/\text{s}$, $E_D = 73\,000\text{ cal/mol}$, $R = 1.987\text{ cal/mol}\cdot\text{K}$.
Starting from the Arrhenius equation [Session 5 Slide 26]:
$$D = D_0 \exp\left(-\frac{E_D}{RT}\right) \implies \ln\left(\frac{D_0}{D}\right) = \frac{E_D}{RT}$$Isolating absolute temperature $T$:
$$T = \frac{E_D}{R \cdot \ln(D_0 / D)}$$1. Diffusivity Ratio:
$$\frac{D_0}{D} = \frac{1.55 \times 10^{-4}\text{ m}^2/\text{s}}{1.0 \times 10^{-14}\text{ m}^2/\text{s}} = 1.55 \times 10^{10}$$2. Natural Logarithm:
$$\ln(1.55 \times 10^{10}) = \ln(1.55) + 10 \cdot \ln(10) = 0.438255 + 23.025851 = 23.464106$$3. Absolute Temperature Calculation:
$$T = \frac{73\,000\text{ cal/mol}}{(1.987\text{ cal/mol}\cdot\text{K}) \times (23.464106)} = \frac{73\,000}{46.62318} \approx \mathbf{1565.74\text{ K}} \approx \mathbf{1566\text{ K}}$$Conversion to Celsius:
$$T = 1565.74 - 273.15 = \mathbf{1292.59^\circ\text{C}} \approx \mathbf{1293^\circ\text{C}}$$Dimensional Check: $[T] = \frac{\text{cal/mol}}{(\text{cal/mol}\cdot\text{K})} = \text{K}$. Physical Insight: A diffusivity of $10^{-14}\text{ m}^2/\text{s}$ needs $1293^\circ\text{C}$: $D$ depends exponentially on $T$, and the high activation energy ($73\text{ kcal/mol} \approx 305\text{ kJ/mol}$) means the atoms need a lot of thermal energy to jump [Slides 4, 26]. Diffusion of impurities in Si wafers is the process used to fabricate integrated circuits [Slide 29].
Problem 3
Problems T3_Diffusion.pdf #3
Two-Point Arrhenius · Activation Energy
Click to Collapse ▲
Activation Energy and Diffusivity of Carbon in Mild Steel
Original statement · Problems T3_Diffusion.pdf
Data: $R = 8.314\text{ J/mol}\cdot\text{K}$
• At $T_1 = 850^\circ\text{C}$: $D_1 = 4.826 \times 10^{-12}\text{ m}^2/\text{s}$
• At $T_2 = 950^\circ\text{C}$: $D_2 = 1.805 \times 10^{-11}\text{ m}^2/\text{s}$
Using these data, calculate:
a) The activation energy for the $\text{C}$ diffusion in mild steel, for the temperature interval of $850^\circ\text{C}$ to $950^\circ\text{C}$ (Solution: $E_D = 36\text{ kcal/mol}$)
b) The diffusivity at $1000^\circ\text{C}$, assuming that the activation energy does not change at this temperature range. (Solution: $D = 3.23 \times 10^{-11}\text{ m}^2/\text{s}$)
- Single Arrhenius Law (assumption): The problem uses one Arrhenius law with constant $D_0$ and $E_D$ for $\text{C}$ in mild steel from $850^\circ\text{C}$ to $1000^\circ\text{C}$, as part b) states [Session 5 Slide 26].
- Parameters: $T_1 = 850^\circ\text{C} = 1123.15\text{ K}$, $D_1 = 4.826 \times 10^{-12}\text{ m}^2/\text{s}$ $T_2 = 950^\circ\text{C} = 1223.15\text{ K}$, $D_2 = 1.805 \times 10^{-11}\text{ m}^2/\text{s}$ $T_3 = 1000^\circ\text{C} = 1273.15\text{ K}$, $R = 8.314\text{ J/mol}\cdot\text{K}$.
Dividing Arrhenius equations at two temperatures eliminates $D_0$ [Session 5 Slide 26]:
$$\ln\left(\frac{D_2}{D_1}\right) = \frac{E_D}{R}\left(\frac{1}{T_1} - \frac{1}{T_2}\right) \implies E_D = \frac{R \cdot \ln(D_2 / D_1)}{\frac{1}{T_1} - \frac{1}{T_2}}$$For part (b), extrapolating to $T_3 = 1000^\circ\text{C}$:
$$D_3 = D_2 \cdot \exp\left[\frac{E_D}{R}\left(\frac{1}{T_2} - \frac{1}{T_3}\right)\right]$$Part (a): Activation Energy $E_D$
$$\frac{D_2}{D_1} = \frac{1.805 \times 10^{-11}}{4.826 \times 10^{-12}} = 3.7401575 \implies \ln(3.7401575) \approx 1.319128$$ $$\frac{1}{T_1} - \frac{1}{T_2} = \frac{1}{1123.15} - \frac{1}{1223.15} = 8.903530 \times 10^{-4} - 8.175612 \times 10^{-4} = 7.27918 \times 10^{-5}\text{ K}^{-1}$$ $$E_D = \frac{8.314 \times 1.319128}{7.27918 \times 10^{-5}} = 150\,666\text{ J/mol} = 150.67\text{ kJ/mol}$$ $$E_D = \frac{150\,666\text{ J/mol}}{4184\text{ J/kcal}} \approx \mathbf{36.01\text{ kcal/mol}} \approx \mathbf{36\text{ kcal/mol}}$$Part (b): Diffusivity at $1000^\circ\text{C}$ ($T_3 = 1273.15\text{ K}$)
$$\frac{1}{T_2} - \frac{1}{T_3} = \frac{1}{1223.15} - \frac{1}{1273.15} = 8.175612 \times 10^{-4} - 7.854534 \times 10^{-4} = 3.21078 \times 10^{-5}\text{ K}^{-1}$$ $$\text{Exponent} = \frac{150\,666}{8.314} \times (3.21078 \times 10^{-5}) = 18121.96 \times 3.21078 \times 10^{-5} \approx 0.581857$$ $$D_3 = (1.805 \times 10^{-11}\text{ m}^2/\text{s}) \times e^{0.581857} = (1.805 \times 10^{-11}) \times 1.78936 \approx \mathbf{3.23 \times 10^{-11}\text{ m}^2/\text{s}}$$Consistency with Official Data: The official reference table [Session 5 Slide 23] lists $D(\text{C in Fe-FCC}) \approx 3 \times 10^{-11}\text{ m}^2/\text{s}$ at $1000^\circ\text{C}$, matching our derived $3.23 \times 10^{-11}\text{ m}^2/\text{s}$; both are of the same order, the slide value being a one-significant-figure figure ($\approx 8\%$ lower). Mechanism Check: $E_D \approx 36\text{ kcal/mol}$ is lower than the activation energy for self-diffusion of iron ($57.5\text{ kcal/mol} = 240\text{ kJ/mol}$ for $\alpha\text{-Fe}$) [Slide 9]. This agrees with carbon diffusing interstitially, which needs less energy than vacancy diffusion [Slide 13].
Problem 4
Problems T3_Diffusion.pdf #4
Fick's 1st Law · Ceramic Barrier Transport
Click to Collapse ▲
Ionic Transport of Nickel Through an MgO Ceramic Plate
Original statement · Problems T3_Diffusion.pdf
- Quasi-Steady State: Because the plate thickness ($\Delta x = 0.1\text{ cm} = 1000\ \mu\text{m}$) is much larger than the layer consumed ($\Delta h = 1\ \mu\text{m}$), flux through $\text{MgO}$ is in steady state [Slide 16].
- Interfacial Boundary Conditions: At the $\text{Ni}/\text{MgO}$ interface, the concentration equals the atomic density of nickel metal: $C_1 = \rho_{\text{at, Ni}} = 4/a^3$ (FCC, as the statement gives). At the $\text{MgO}/\text{Ta}$ interface, the nickel that arrives is taken as removed at once: $C_2 \approx 0$ (assumption of the model).
- Parameters: $\Delta x = 0.1\text{ cm}$, $D = 9 \times 10^{-12}\text{ cm}^2/\text{s}$, $a = 3.6 \times 10^{-8}\text{ cm}$, $\Delta h = 1\ \mu\text{m} = 10^{-4}\text{ cm}$.
1. Atoms of $\text{Ni}$ removed per unit area: $N/A = \rho_{\text{at, Ni}} \cdot \Delta h = \frac{4 \cdot \Delta h}{a^3}$.
2. Steady-state flux via Fick's First Law [Session 5 Slide 16]:
3. Conservation of mass equates flux to removal rate:
$$J = \frac{N/A}{t} \implies \frac{\rho_{\text{at, Ni}} \cdot \Delta h}{t} = D \frac{\rho_{\text{at, Ni}}}{\Delta x} \implies t = \frac{\Delta x \cdot \Delta h}{D}$$The atomic density $\rho_{\text{at, Ni}}$ elegantly cancels out.
1. Direct Evaluation with General Formula:
$$\Delta x = 0.1\text{ cm}, \quad \Delta h = 10^{-4}\text{ cm}, \quad D = 9 \times 10^{-12}\text{ cm}^2/\text{s}$$ $$t = \frac{(0.1\text{ cm}) \times (10^{-4}\text{ cm})}{9 \times 10^{-12}\text{ cm}^2/\text{s}} = \frac{10^{-5}\text{ cm}^2}{9 \times 10^{-12}\text{ cm}^2/\text{s}} = \frac{1}{9} \times 10^7\text{ s} \approx 1.1111 \times 10^6\text{ s}$$2. Conversion to Hours:
$$t = \frac{1.1111 \times 10^6\text{ s}}{3600\text{ s/h}} \approx \mathbf{308.64\text{ h}} \approx \mathbf{309\text{ h}}$$3. Step-by-Step Atomic Confirmation:
$$\rho_{\text{at, Ni}} = \frac{4}{(3.6 \times 10^{-8}\text{ cm})^3} = 8.5734 \times 10^{22}\text{ atoms/cm}^3$$ $$N/A = (8.5734 \times 10^{22}) \times (10^{-4}) = 8.5734 \times 10^{18}\text{ atoms/cm}^2$$ $$J = (9 \times 10^{-12}) \times \frac{8.5734 \times 10^{22}}{0.1} = 7.71605 \times 10^{12}\text{ atoms/cm}^2\cdot\text{s}$$ $$t = \frac{8.5734 \times 10^{18}}{7.71605 \times 10^{12}} = 1.1111 \times 10^6\text{ s} \approx \mathbf{309\text{ h}}$$Dimensional Check: $[t] = \frac{\text{cm} \cdot \text{cm}}{\text{cm}^2/\text{s}} = \text{s}$. Physical Insight: Nearly 13 days ($309\text{ hours}$) at $1400^\circ\text{C}$ are needed to remove $1\ \mu\text{m}$ of nickel: diffusion is slower for charged species, in compact structures and in materials with high $T_m$ [Slide 27]. At fixed concentrations the flux falls as $1/\Delta x$, so a thicker plate would take proportionally longer [Slide 16].
Problem 5
Problems T3_Diffusion.pdf #5
Fick's 2nd Law + Arrhenius Coupled
Click to Collapse ▲
Cementation Temperature for Steel 1010 in 8 Hours
Original statement · Problems T3_Diffusion.pdf
Data: $D_0 (\text{C in Fe}) = 16.2\text{ mm}^2/\text{s}$; $E_D = 137\,800\text{ J/mol}$; $R = 8.314\text{ J/mol}\cdot\text{K}$.
- Coupled Transient Diffusion: Combines semi-infinite error function diffusion with Arrhenius thermal kinetics [Session 5 Slides 20, 26].
- Parameters: $C_0 = 0.10\text{ wt}\%$, $C_s = 1.31\text{ wt}\%$, $C_x = 0.30\text{ wt}\%$ $x = 1.16\text{ mm} = 1.16 \times 10^{-3}\text{ m}$ $t = 8\text{ h} = 8 \times 3600\text{ s} = 28\,800\text{ s}$ $D_0 = 16.2\text{ mm}^2/\text{s} = 16.2 \times 10^{-6}\text{ m}^2/\text{s}$ $E_D = 137\,800\text{ J/mol}$, $R = 8.314\text{ J/mol}\cdot\text{K}$.
1. Evaluate the concentration ratio $\text{erf}(z) = \frac{C_s - C_x}{C_s - C_0}$.
2. Determine $z$ via linear interpolation in the official table [Slide 21].
3. Calculate the required diffusivity: $D = \frac{x^2}{4 \cdot z^2 \cdot t}$.
4. Invert the Arrhenius equation to calculate the processing temperature $T$:
1. Error Function Value:
$$\text{erf}(z) = \frac{1.31 - 0.30}{1.31 - 0.10} = \frac{1.01}{1.21} \approx 0.83471074$$2. Linear Interpolation in Error Function Table [Slide 21]:
- For $z_1 = 0.95 \implies \text{erf}(z_1) = 0.8209$
- For $z_2 = 1.00 \implies \text{erf}(z_2) = 0.8427$
3. Diffusivity Calculation:
$$x^2 = (1.16 \times 10^{-3}\text{ m})^2 = 1.3456 \times 10^{-6}\text{ m}^2$$ $$4 \cdot z^2 \cdot t = 4 \times 0.963688 \times 28\,800\text{ s} = 111\,016.86\text{ s}$$ $$D = \frac{1.3456 \times 10^{-6}\text{ m}^2}{111\,016.86\text{ s}} \approx \mathbf{1.21207 \times 10^{-11}\text{ m}^2/\text{s}}$$4. Inverting Arrhenius for Temperature $T$:
$$\frac{D}{D_0} = \frac{1.21207 \times 10^{-11}\text{ m}^2/\text{s}}{16.2 \times 10^{-6}\text{ m}^2/\text{s}} = 7.48191 \times 10^{-7} \implies \ln\left(\frac{D}{D_0}\right) = \ln(7.48191 \times 10^{-7}) = 2.01249 - 16.11810 = -14.10561$$ $$T = \frac{-137\,800\text{ J/mol}}{(8.314\text{ J/mol}\cdot\text{K}) \times (-14.10561)} = \frac{137\,800}{(8.314)(14.10561)} = \frac{137\,800}{117.275} \approx \mathbf{1175.02\text{ K}} \approx \mathbf{1175\text{ K}}$$ $$T = 1175.02 - 273.15 = \mathbf{901.87^\circ\text{C}} \approx \mathbf{902^\circ\text{C}}$$Sensitivity to Temperature: To reach $0.30\text{ wt}\%$ at $1.16\text{ mm}$ in $8\text{ h}$ the steel must be carburized at $902^\circ\text{C}$. Because $D$ grows exponentially with $T$, a lower temperature would need a much longer time and a higher one a shorter time [Slide 26].
Problem 6
Problems T3_Diffusion.pdf #6
Fick's 1st Law · Membrane Purification
Click to Collapse ▲
Hydrogen Purification Through a Palladium Membrane
Original statement · Problems T3_Diffusion.pdf
- Pure Steady State: Constant external pressures create time-invariant boundary concentrations $C_{\text{high}}$ and $C_{\text{low}}$ [Session 5 Slide 35].
- Parameters: $A = 0.2\text{ m}^2$, $\dot{m} = 1.73 \times 10^{-3}\text{ kg/h}$ $C_1 = 1.5\text{ kg/m}^3$, $C_2 = 0.3\text{ kg/m}^3$ $D = 1.0 \times 10^{-8}\text{ m}^2/\text{s}$.
1. Compute mass flux density [Session 5 Slide 15]:
$$J = \frac{\dot{m}}{A} \quad \left[\frac{\text{kg}}{\text{m}^2\cdot\text{s}}\right]$$2. Apply Fick's First Law across the flat membrane [Slide 16, 35]:
$$J = D \frac{\Delta C}{\Delta x} \implies \Delta x = \frac{D \cdot (C_1 - C_2)}{J}$$1. Flux Density ($J$):
$$\frac{\dot{m}}{A} = \frac{1.73 \times 10^{-3}\text{ kg/h}}{0.2\text{ m}^2} = 8.65 \times 10^{-3}\text{ kg/m}^2\cdot\text{h}$$ $$J = \frac{8.65 \times 10^{-3}\text{ kg/m}^2\cdot\text{h}}{3600\text{ s/h}} \approx \mathbf{2.4028 \times 10^{-6}\text{ kg/m}^2\cdot\text{s}}$$The worked solution stored in the text layer of Official Slide 35 gives $J = 8.65 \times 10^{-3}\text{ kg/m}^2\cdot\text{h} = 2.4 \times 10^{-6}\text{ kg/m}^2\cdot\text{s}$ and $\Delta x = 5\text{ mm}$ (the answer box on the rendered slide appears empty).
2. Concentration Difference:
$$\Delta C = C_1 - C_2 = 1.5\text{ kg/m}^3 - 0.3\text{ kg/m}^3 = 1.2\text{ kg/m}^3$$3. Membrane Thickness ($\Delta x$):
$$\Delta x = \frac{(1.0 \times 10^{-8}\text{ m}^2/\text{s}) \times (1.2\text{ kg/m}^3)}{2.4028 \times 10^{-6}\text{ kg/m}^2\cdot\text{s}} \approx 4.994 \times 10^{-3}\text{ m} \approx \mathbf{5.0 \times 10^{-3}\text{ m}} = \mathbf{5\text{ mm}}$$Dimensional Check: $[\Delta x] = \frac{(\text{m}^2/\text{s}) \cdot (\text{kg/m}^3)}{\text{kg}/(\text{m}^2\cdot\text{s})} = \text{m}$. Physical Insight: At fixed concentrations the flux scales as $J \propto 1/\Delta x$ [Slide 16], so $5\text{ mm}$ is the thickness at which steady-state permeation equals the required purified flow. Hydrogen is small with respect to the matrix, so it diffuses interstitially, which is faster [Slide 13]; its diffusivity ($10^{-8}\text{ m}^2/\text{s}$) is far above the values of the diffusivity table [Slide 23]. This is the example of $\text{H}_2$ purification through a $\text{Pd}$ layer [Slide 35].