Diffusion and its applications.

Σχετικά έγγραφα
Διάχυση και εφαρμογές. Αυτο-διάχυση (self-diffusion), π.χ. διάχυση ραδιενεργών ισοτόπων.

HOMEWORK 4 = G. In order to plot the stress versus the stretch we define a normalized stretch:

Homework 8 Model Solution Section

Απόκριση σε Μοναδιαία Ωστική Δύναμη (Unit Impulse) Απόκριση σε Δυνάμεις Αυθαίρετα Μεταβαλλόμενες με το Χρόνο. Απόστολος Σ.

[1] P Q. Fig. 3.1

Jesse Maassen and Mark Lundstrom Purdue University November 25, 2013

6.1. Dirac Equation. Hamiltonian. Dirac Eq.

Phys460.nb Solution for the t-dependent Schrodinger s equation How did we find the solution? (not required)

Mean bond enthalpy Standard enthalpy of formation Bond N H N N N N H O O O

Concrete Mathematics Exercises from 30 September 2016

Approximation of distance between locations on earth given by latitude and longitude

Partial Differential Equations in Biology The boundary element method. March 26, 2013

Matrices and Determinants

CHAPTER 25 SOLVING EQUATIONS BY ITERATIVE METHODS

Section 8.3 Trigonometric Equations

Capacitors - Capacitance, Charge and Potential Difference

Inverse trigonometric functions & General Solution of Trigonometric Equations

ST5224: Advanced Statistical Theory II

Srednicki Chapter 55

the total number of electrons passing through the lamp.

(1) Describe the process by which mercury atoms become excited in a fluorescent tube (3)

Second Order Partial Differential Equations

1 String with massive end-points

ΠΟΛΥΤΕΧΝΕΙΟ ΚΡΗΤΗΣ ΣΧΟΛΗ ΜΗΧΑΝΙΚΩΝ ΠΕΡΙΒΑΛΛΟΝΤΟΣ

Thi=Τ1. Thο=Τ2. Tci=Τ3. Tco=Τ4. Thm=Τ5. Tcm=Τ6

Second Order RLC Filters

4.6 Autoregressive Moving Average Model ARMA(1,1)

EE512: Error Control Coding

ADVANCED STRUCTURAL MECHANICS

DESIGN OF MACHINERY SOLUTION MANUAL h in h 4 0.

Μελέτη Πρότυπων Καταλυτικών Συστηµάτων. µε Επιφανειακά Ευαίσθητες Τεχνικές

Surface Mount Multilayer Chip Capacitors for Commodity Solutions

5.4 The Poisson Distribution.

Strain gauge and rosettes

Instruction Execution Times

Technical Information T-9100 SI. Suva. refrigerants. Thermodynamic Properties of. Suva Refrigerant [R-410A (50/50)]

Example Sheet 3 Solutions

3.4 SUM AND DIFFERENCE FORMULAS. NOTE: cos(α+β) cos α + cos β cos(α-β) cos α -cos β

b. Use the parametrization from (a) to compute the area of S a as S a ds. Be sure to substitute for ds!

MECHANICAL PROPERTIES OF MATERIALS

Εργαστήριο Ανάπτυξης Εφαρμογών Βάσεων Δεδομένων. Εξάμηνο 7 ο

Econ 2110: Fall 2008 Suggested Solutions to Problem Set 8 questions or comments to Dan Fetter 1

ΕΙΣΑΓΩΓΗ ΣΤΗ ΣΤΑΤΙΣΤΙΚΗ ΑΝΑΛΥΣΗ

Solutions to the Schrodinger equation atomic orbitals. Ψ 1 s Ψ 2 s Ψ 2 px Ψ 2 py Ψ 2 pz

Main source: "Discrete-time systems and computer control" by Α. ΣΚΟΔΡΑΣ ΨΗΦΙΑΚΟΣ ΕΛΕΓΧΟΣ ΔΙΑΛΕΞΗ 4 ΔΙΑΦΑΝΕΙΑ 1

DuPont Suva. DuPont. Thermodynamic Properties of. Refrigerant (R-410A) Technical Information. refrigerants T-410A ENG

The Simply Typed Lambda Calculus

H επεξεργασία πληροφορίας απαιτεί ανίχνευση πληροφορίας

Homework 3 Solutions

DuPont Suva 95 Refrigerant

Reminders: linear functions

Nuclear Physics 5. Name: Date: 8 (1)

Calculating the propagation delay of coaxial cable

Ανάπτυξη οξειδίου του πυριτίου σε αντιδραστήρα πλάσματος και ηλεκτρικός χαρακτηρισμός του

SMD Transient Voltage Suppressors

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Potential Dividers. 46 minutes. 46 marks. Page 1 of 11

SOLUTIONS TO MATH38181 EXTREME VALUES AND FINANCIAL RISK EXAM

ΠΡΟΣΤΑΣΙΑ ΑΤΜΟΣΦΑΙΡΙΚΗΣ ΔΙΑΒΡΩΣΗΣ ΑΛΟΥΜΙΝΙΟΥ/ΑΝΟΔΙΩΣΗ Al

Mock Exam 7. 1 Hong Kong Educational Publishing Company. Section A 1. Reference: HKDSE Math M Q2 (a) (1 + kx) n 1M + 1A = (1) =

DuPont Suva 95 Refrigerant

Math221: HW# 1 solutions

2 Composition. Invertible Mappings

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Figure 1 T / K Explain, in terms of molecules, why the first part of the graph in Figure 1 is a line that slopes up from the origin.

6.3 Forecasting ARMA processes

Problem Set 9 Solutions. θ + 1. θ 2 + cotθ ( ) sinθ e iφ is an eigenfunction of the ˆ L 2 operator. / θ 2. φ 2. sin 2 θ φ 2. ( ) = e iφ. = e iφ cosθ.

Macromechanics of a Laminate. Textbook: Mechanics of Composite Materials Author: Autar Kaw

Higher Derivative Gravity Theories

EE101: Resonance in RLC circuits

ΘΕΩΡΗΤΙΚΗ ΚΑΙ ΠΕΙΡΑΜΑΤΙΚΗ ΙΕΡΕΥΝΗΣΗ ΤΗΣ ΙΕΡΓΑΣΙΑΣ ΣΚΛΗΡΥΝΣΗΣ ΙΑ ΛΕΙΑΝΣΕΩΣ

Discontinuous Hermite Collocation and Diagonally Implicit RK3 for a Brain Tumour Invasion Model

Forced Pendulum Numerical approach

The kinetic and potential energies as T = 1 2. (m i η2 i k(η i+1 η i ) 2 ). (3) The Hooke s law F = Y ξ, (6) with a discrete analog

forms This gives Remark 1. How to remember the above formulas: Substituting these into the equation we obtain with

Mechanical Behaviour of Materials Chapter 5 Plasticity Theory

Pg The perimeter is P = 3x The area of a triangle is. where b is the base, h is the height. In our case b = x, then the area is

A Bonus-Malus System as a Markov Set-Chain. Małgorzata Niemiec Warsaw School of Economics Institute of Econometrics

Aluminum Electrolytic Capacitors (Large Can Type)

1. Ηλεκτρικό μαύρο κουτί: Αισθητήρας μετατόπισης με βάση τη χωρητικότητα

Enthalpy data for the reacting species are given in the table below. The activation energy decreases when the temperature is increased.

C.S. 430 Assignment 6, Sample Solutions

( y) Partial Differential Equations

Section 9.2 Polar Equations and Graphs

상대론적고에너지중이온충돌에서 제트입자와관련된제동복사 박가영 인하대학교 윤진희교수님, 권민정교수님

Space-Time Symmetries

Multilayer Ceramic Chip Capacitors

; +302 ; +313; +320,.

Areas and Lengths in Polar Coordinates


ANSWERSHEET (TOPIC = DIFFERENTIAL CALCULUS) COLLECTION #2. h 0 h h 0 h h 0 ( ) g k = g 0 + g 1 + g g 2009 =?

ΕΦΑΡΜΟΓΗ ΕΥΤΕΡΟΒΑΘΜΙΑ ΕΠΕΞΕΡΓΑΣΜΕΝΩΝ ΥΓΡΩΝ ΑΠΟΒΛΗΤΩΝ ΣΕ ΦΥΣΙΚΑ ΣΥΣΤΗΜΑΤΑ ΚΛΙΝΗΣ ΚΑΛΑΜΙΩΝ

UDZ Swirl diffuser. Product facts. Quick-selection. Swirl diffuser UDZ. Product code example:

Introduction to Theory of. Elasticity. Kengo Nakajima Summer

D Alembert s Solution to the Wave Equation

ΕΛΛΗΝΙΚΗ ΔΗΜΟΚΡΑΤΙΑ ΠΑΝΕΠΙΣΤΗΜΙΟ ΚΡΗΤΗΣ. Ψηφιακή Οικονομία. Διάλεξη 7η: Consumer Behavior Mαρίνα Μπιτσάκη Τμήμα Επιστήμης Υπολογιστών

Multilayer Ceramic Chip Capacitors

Στο εστιατόριο «ToDokimasesPrinToBgaleisStonKosmo?» έξω από τους δακτυλίους του Κρόνου, οι παραγγελίες γίνονται ηλεκτρονικά.

Statistical Inference I Locally most powerful tests

ΕΛΛΗΝΙΚΗ ΔΗΜΟΚΡΑΤΙΑ ΠΑΝΕΠΙΣΤΗΜΙΟ ΚΡΗΤΗΣ

Transcript:

Diffusion and its applications. Concentration gradient diffusion In materials technology concentration gradients are used in order to change the surface composition. Self-diffusion: diffusion of isotopes. Important diffusion applications: Steel hardening via diffusion of N and C. Impurity diffusion in semiconductors (e.g. fabrication of the base and emitter region, resistors etc). Si oxidation Powder metallurgy which relies on mass transport for powder bonding or sintering. W. G. Pfann in 195 patented the application of diffusion for the introduction of impurities in semiconductors with the purpose of varying their conductivity. 1

Diffusion systems: 1. Diffusion at high T from a gas source (infinite source).. Diffusion from an oxide which is rich in the impurity (limited source). 3. Diffusion and annealing from an implanted layer (limited source). Diffusion theories Fick s theory describes the phenomenon using appropriate diffusion coefficients. Applies when the impurity concentrations are low D constant. Atomistic theory: Interaction of impurities with defects. Boltzmann-Matano analysis: is used when the diffusion coefficient depends on the impurity concentration.

Mathematical description of diffusion J = D dc 1 st Fick s law dx where J is the flux (atoms/cm s), C is the concentration (atoms/cm 3 ) & D is the diffusion coefficient (cm /s). The concentration gradient and the direction of diffusion. The quantity dc/dt 0 when J x J x+dx Mass conservation: the time dependent variation of the impurity concentration must be equal to the local reduction of the diffusion flux, i.e. C(x, t) t C(x, t) = D x x When the impurity concentration is low D= constant: C(x, t) t C(x, t) = D x nd Fick s law. It describes diffusion under non-steady-state conditions and under the assumption that D is constant. 3

Steady-state diffusion C (x, t) t = 0, i.e. linear variation of the concentration and J (mass flow) is constant at every cross section of the system. The most common solutions for D=constant are: Constant surface concentration (limitless source). Constant total impurity concentration (S) (limited source or predeposition condition). Constant surface concentration Initial condition: C(x,0)=0 (for t=0) Boundary conditions: C(0,t)=C s & C(,t)=0 C(x, t) = Cs erfc x Dt where C s is the constant surface concentration (atoms/cm 3 ) and erfc is the complementary error function which is defined by equation: erfc x Dt ( x 4Dt ) = 1 exp( z )dz π 0 For small values of the integral (όρισμα): erfc( x 4Dt ) 1 x 4Dt 4

Time evolution of the normalized impurity distribution described by an erfc. Predepostion diffusion- Constant total amount of dopant Initial condition: C(x,0)=0 Boundary conditions : C (x, t)dx = S και C(x, )=0 0 The distribution is Gaussian: C(x, t) = S πdt x exp 4Dt The surface concentration C s (x=0) is: C s = C(0, t) = S πdt Time evolution of the normalized impurity distribution described by a Gaussian. 5

Differences between erfc and Gaussian in logarithmic and linear scales x = Dt Correlates the 3 basic parameters x, t & Τ. The temperature dependence is introduced via the relation: D = Do exp( ΔED RT). 6

The microscopic theory of diffusion The diffusion coefficient. The atom movement between adjustment cross sections is possible when: 1. A suitable empty site exists, e.g. a vacancy.. The diffusing atom has sufficient energy to overcome the energy barrier ΔΕ D imposed by its neighboring atoms. When the atom occupies a neighboring vacancy, the vacancy moves in the opposite direction and occupies the initial atom s site. C 1 C C(x) C 1 S 1 α Atomic jumps of equal probability across the atomic planes 1 & C α x Diffusion occurs against the concentration gradient. Assumptions: 1. The atoms on the cross sections 1 & perform atomic jumps with frequency ν.. The atom jumps or or have equal probability the atoms tend to cross surface S only during 1/ of the number of the jumps. 7

The diffusing atom must have sufficient energy to overcome the energy barrier ΔΕ D. α is the distance between adjacent atom planes, n i (cm - ) & c i (cm -3 ) are the surface & bulk concentrations, respectively. n i =αc i. Atomic kinetic model -Ατομικό κινητικό πρότυπο. The diffusion current J is: 1 Replace n i =αc i J = αν( ) The concentration gradient J 1 c να x = (1) 1 1 J = n1ν n ν c 1 c c x is: c c c 1 = α x the minus sign indicates that diffusion occurs against the concentration gradient. Via a comparison of (1) with Fick s law c J = D x D 1 να = (-dimensions) & D 1 να 6 = (3-dimensions) 8

The temperature dependence of the diffusion coefficient: E It is experimentally demonstrated that: D = D exp D o where D o is kt a constant, E D is the activation energy for diffusion (diffusion introduces strain in the lattice =>energy barrier). Diffusion mechanisms Diffusion with mutual exchange of lattice-atoms Diffusion via hopping of interstitial atoms. Vacancy assisted diffusion. If the atom vibrates with frequency ν ο around its equilibrium position (i.e. makes ν ο attempts/sec to overcome the energy barrier), the probability that it makes a successful jump is given by the Boltzmann factor : exp(-e m /kt). Therefore, it changes position with a frequency: E ν = ν ο exp m kt The atom can move only if there is an adjustment empty lattice position in its 1 st nearest neighbor shell (coordination number Z). The probability that there exists and empty lattice position is exp(-e v /kt), where Ε v is the formation energy of a vacancy. the probability that an atom migrates from one lattice position to a neighboring one is: E ν = Z ν ο exp m kt E exp v kt E = Z ν ο exp D kt, where E D =E m +E v. 9

Therefore: 1 = ν α E D Z ο exp D 6 kt E = D exp D o kt, where D Zν α o 1 = ο. 6 Arrhenius plot (logd-1/t) D o & E D D(T) για Fe:Ni. D varies by 16 orders of magnitude for ΔΤ = 0-1000 o C. 10

Results based on experimental observations: 1. In solids: ΔΕ D = 0. - ev per atom.. In several materials ΔΕ D Τ m 3. In liquids D has a weak Τ-dependence (10-4 -10-6 cm s -1 ) 4. The atoms diffuse fast on surfaces, grain boundaries & dislocations Extrinsic diffusion When the material is doped and the carrier concentration > the intrinsic carrier concentration, the value of D depends on n. Representative values of n i at 1000 o C: n i (Si)= 5x10 18 cm -3 & n i (GaAs)= 5x10 17 cm -3. D depends on n because when the carrier concentration increases both the position of E F and the vacancy concentration change. The vacancy concentration C V as a function of T is E E C = C exp F i V i kt where C i & E i are the vacancy concentration and Fermi level for the intrinsic material. When the diffusion is vacancy assisted, the value of D will depend on their concentration. For small values of n, E F & E i coincide. However, when the semiconductor becomes extrinsic E F moves in the gap and thus 11

exp ( E E kt) F of D increase. i >1. both the vacancy concentration and the value Variation of D versus the carrier concentration. The extrinsic diffusion is described by the equation C t F = x = C D instead of x x C t = D x C where D=const. The variation of D versus the concentration C is given by γ C D = D S where D S & C S are the values on the surface. In this C S case the diffusion equations are solved arithmetically. Normalized impurity distributions in extrinsic diffusion under the assumption that the concentration on the surface is constant. 1

When γ>0 the value of D decreases with decreasing C the distributions look like step functions (abrupt junction) abrupt junctions can be fabricated via diffusion in a substrate with the opposite kind of dopant. 13

Diffusion systems. Open vertical or horizontal quartz reactors operating at 900-100 ο C. Sources : solid, liquid or gaseous (better control). The T-gradients cause thermal stress introduction of defects (e.g. dislocations, wafer deformation etc) precise T control is necessary. Typical T-gradient: 3-10 ο C/min while at the center of the diffusion system (length 10-100 cm) the T is controlled within ±0.5 o C. Schematic of a diffusion system with gas sources. 14

Diffusion applications. Metal hardening 1. Carburization of steel is a surface hardening process which relies on diffusion. Takes place at 900 ο C in C-containing atmosphere, e.g. CH 4 -CO-H. The thickness of the C-rich surface layer increases with a rate t.. Hardening via diffusion of N and/or N & C. Oxidation. Proceeds with oxygen or metal diffusion via the existing native oxide Native oxidation is undesirable since it is not precisely controlled. SiO Gate dielectric & mask against impurity diffusion in μ-electronics Oxidation saturates dangling bonds at the Si surface reduces the density of surface states in the gap and the density of immobile charge. The native SiO is 15-40Å thick & ragged useless 15

The Si oxidation takes place at 900-100 ο C in dry or wet Ο : Dry oxidation: Si+O SiO (density.5gr/`cm 3 ) Wet oxidation: Si+H O SiO +H (density.15gr/cm 3 ) Oxidation mechanism 1. The oxygen is transferred to the wafer surface via a stagnant layer (remember CVD). Oxygen (Ο or Ο) diffuses via the native SiO to the Si-SiO interface consumption of Si from the substrate and growth of SiO. When the thickness of the consumed Si is x the thickness of the grown SiO is.7x. The thickness of the grown oxide is x 1/ D N k t = 1 + o 1 k Dn Where k is the reaction constant, n is the number of oxygen molecules in the unit volume of SiO (n=.x10 cm -3 for dry O & 4.4x10 cm -3 for H O). Limiting cases 1. short t (x 40Å in dry Ο & x 1000Å in Η Ο) limiting step is the oxidation rate (diffusion is fast while the reaction constant k is small) : N k x = o t x = n B A t. For longer t growth is limited by diffusion 1/ 1/ NoD x = t x = n B 1/ 1/ t 16

Β/Α : is the linear growth rate constant & Β : is the parabolic rate growth constant. Si oxidation rate in the T-range 700-1300 ο C. For short t the oxidation rate depends linearly on t (reaction rate limited). For longer t the growth is parabolic (diffusion limited) The activation energy for oxidation (E ox ) is calculated from Arrhenius plots lnx-1000/t. Ε ox in the linear region is -.05 ev/molecule ( 1.83eV/molecule is needed to break Si-Si bonds). In the parabolic region Ε ox for dry oxidation is 1.3eV/molecule ( 1.18eV/molecule is needed for oxygen diffusion in SiO ). In the parabolic region Ε ox for wet oxidation: E ox =0.78eV/molecule and E d =0.79eV/molecule for H O diffusion in SiO ). 17

Wet oxidation is 3 times faster than dry due to the larger solubility of Η Ο in SiO, which counter balances the smaller diffusion coefficient. Modifications of thermal oxidation: rapid thermal oxidation, introduction of halogens (F, Cl etc) 18

Redistribution of impurities during oxidation. Κατά την οξείδωση του Si οι προσμείξεις που βρίσκονται κοντά στην επιφάνεια του υποστρώματος ανακατανέμονται λόγω : (1) της βαθμίδας συγκέντρωσης μεταξύ υποστρώματος και αναπτυσσόμενου οξειδίου και () της διαφορετικής συγκέντρωσης ισορροπίας της πρόσμειξης ανάμεσα στο Si και το SiO. Το φαινόμενο είναι ανάλογο της αποβολής πρόσμειξης από κρύσταλλο που αναπτύσσεται από το τήγμα. Ορίζεται ο συντελεστής k που περιγράφει την διαφορά της συγκέντρωσης ισορροπίας στο Si και το SiO : συγκέντρωση ισορροπίας της πρόσμειξης στο Si k =. συγκέντρωση ισορροπίας της πρόσμειξης στο SiO Ένας άλλος μηχανισμός που συμβάλλει στην ανακατανομή της πρόσμειξης απαντάται όταν ο συντελεστής διάχυσης της πρόσμειξης στο οξείδιο είναι μεγάλος οπότε είναι δυνατόν τα άτομα της πρόσμειξης να φτάσουν στην επιφάνεια του οξειδίου και να διαφύγουν στην αέριο φάση. Τέλος ανακατανομή της πρόσμειξης συμβαίνει λόγω του μεγαλύτερου όγκου του αναπτυσσόμενου SiO σε σύγκριση με τον όγκο του Si που καταναλώνεται (V SiO V Si ή d Si =0.44d SiO ). 19

Σχήμα ΙΙΙ.1: Ανακατανομή των προσμείξεων κατά την οξείδωση του Si. Στις περιπτώσεις (α) & (β) : k<1 το Si αποβάλλει πρόσμειξη στο SiO. Ειδικότερα στο (β) η πρόσμειξη διαχέεται γρήγορα μέσω του SiO ο βαθμός μείωσης της συγκέντρωσης της πρόσμειξης στο Si είναι μεγάλος. Στις περιπτώσεις ( c) & (d): k>1 το οξείδιο αποβάλλει πρόσμειξη στο Si. Ειδικότερα, όταν η πρόσμειξη διαχέεται βραδέως μέσα στο SiO, όπως στο ( c), εμφανίζεται συσσώρευση της πρόσμειξης στην επιφάνεια του Si. Αντίθετα όταν η πρόσμειξη διαχέεται γρήγορα μέσα στο SiO, συμβαίνει απώλεια στην αέριο φάση και εκκένωση της επιφάνειας του Si από πρόσμειξη. 0

Η επίδραση του κρυσταλλογραφικού προσανατολισμού στην ταχύτητα οξείδωσης. Παραβολική περιοχή: η ταχύτητα οξείδωσης είναι ανεξάρτητη του κρυσταλλογραφικού προσανατολισμού του Si (η ανάπτυξη περιορίζεται από την διάχυση των οξειδωτικών ριζών μέσα από το αναπτυσσόμενο οξείδιο). Στην γραμμική περιοχή η ταχύτητα οξείδωσης εξαρτάται από την ταχύτητα ενσωμάτωσης των ατόμων του Si στο πλέγμα του SiO εξαρτάται από την επιφανειακή πυκνότητα του Si και τον προσανατολισμό του υποστρώματος. 1 η προσέγγιση: Η ταχύτητα οξείδωσης του Si (111) >> του Si (100) Si Si κατά ( B A) (111) : ( B A) (100) 1.16:1. Ακριβέστερη προσέγγιση: 3D πλέγμα του Si και το σχετικό μέγεθος των ατόμων του Si και του οξυγόνου η πειραματική τιμή ( B A) Si (111) : ( B A) Si (100) 1.68:1. 1