MAT 227
“Vapor Pressures, Solid Composition of
MOCVD”
Lecture 2
April 2, 2025
S. DenBaars
Materials and ECE Departments
University of California Santa Barbara
Santa Barbara, CA 93106
Vapor Pressures-Metalorganic compound
The vapor pressure of the MO source is an important consideration in
MOCVD, since it determines the concentration of source material in the
reactor and the deposition rate. Too low a vapor pressure makes it difficult
to transport the source into the deposition zone and to achieve reasonable
growth rates. Too high a vapor pressure may raise safety concerns if the
compound is toxic. Further more, it is easier to control the delivery from a
liquid than from a solid.
Vapor pressures of Metalorganic compounds are calculated in terms of the
expression
Log[p(torr)]=B-A/T
Vapor pressure of most common MO compounds
Log[p(torr)]=B-A/T T is in Kelvin
Compound P at 298 K A B Melt point
(torr) (oC)
(Al(CH3)3)2 TMAl 14.2 2780 10.48 15
Al(C2H5)3 TEAl 0.041 3625 10.78 -52.5
Ga(CH3)3 TMGa 238 1825 8.50 -15.8
Ga(C2H5)3 TEGa 4.79 2530 9.19 -82.5
In(CH3)3 TMIn 1.75 2830 9.74 88
In(C2H5)3 TEIn 0.31 2815 8.94 -32
Zn(C2H5)2 DEZn 8.53 2190 8.28 -28
Mg(C5H5)2 Cp2Mg 0.05 3556 10.56 175
Calculate the molar flow rate of MO sources
When we read some reference, we often see people use mol/min to indicate the flow
rate. Normally, we use the formula below to calculate it.
F (mol/min)=p MO/p Bubbler*[flow rate (ml/min)]/22400 (mol/ml)
Note 1sccm= std. cm3 per min = 1ml/min(milliliter/min). We need to calculate the
molar flow rate before we determine the growth condition. If we want to grow
alloys, we can use the mole flow rate to estimate the alloys’ composition.
For example, if we grow AlGaN, we can estimate the Al concentration use the
following formula if we assume the efficiency of Al and Ga sources is the same.
x Al=F Al/(F Al+ F Ga)
Calculating Solid Composition of InGaAs from Bubbler Flow Rates
To calculate the solid composition (x) of InₓGa₁₋ₓAs grown by MOCVD using TMIn and TMGa:
1. **Determine molar flow of precursors:**
- Use the equation:
F = (P_vapor / P_total) × (f_carrier / 22,400)
where:
- P_vapor = vapor pressure of precursor
- P_total = total pressure in bubbler (e.g., 760 torr)
- f_carrier = carrier gas flow rate (sccm)
- 22,400 = molar volume at STP (mL/mol)
2. **Calculate In and Ga molar flows:**
- F_In = flow from TMIn bubbler
- F_Ga = flow from TMGa bubbler
3. **Determine gas-phase mole fraction of In:**
- x_gas = F_In / (F_In + F_Ga)
4. **Assume solid composition reflects gas ratio:**
- x_solid ≈ x_gas (valid at low V/III ratio and mass transport-limited regime)
Group III
Linear Incorporation
-AlGaAs
-InGaP
-InGaAs all linear
(InGaN is the exception
On-linear)
EXAMPLE Group III Grown Lattice Match InxGa1-xP on GaAs
x In=F TMIn/(F TMIn+ F TMGa)
F (mol/min)=p MO/p Bubbler*[flow rate (ml/min)]/22400 (mol/ml)
Figure 1
SOLUTION Group III composition
Group V
Non-linear Incorporation
How do we make a lattice matched compound to InP a=5.87ang??
Lattice matching
line InP
1eV->1.24m
x=0.2
y=0.45
If compound is GaxIn1-xAsyP1-y emtting at 1.0eV than the composition
Is the intersection of the energy countoru with lattice match point,
i.e. Ga0.2In0.8As0.45P0.55
Mixing on Group V Site (As,P,N)
XsolidP = k1PPH3/(k1PPH3 + k2PAsH3)
At T=400-800C Due to incomplete pyrolysis of
PH3 rate k1 is much less than pyrolysis of AsH3
rate k2 and is highly temperature dependent (see
Fig 3.36)
ksegregation= k2/k1
Rewriting we get
XsolidP = PPH3/(PPH3 + ksegPAsH3)
Group V - Example
• What is ksegregation for XsolidP= 0.45 at 650C
using phosphine and arsine?
Example solution-Group V Mixing
• XsolidP = PPH3/(PPH3 + ksegPAsH3)
• From chart we can see for 40% P
• 0.4= PPH3/(PPH3 + ksegPAsH3)
• PPH3/(PPH3 + PAsH3) =0.9
• PPH3=0.9*(PPH3 + PAsH3). -> (0.1/0.9 )PPH3 = PAsH3-> PPH3/(PAsH3)=9
Solving for kseg
• kseg=1/(PAsH3)*(PPH3 /0.4 - PPH3)=1.5 PPH3/ PAsH3
kseg = 1.5*9 = 13.5
AsH3 pyrolizes at 13.5 times the rate of phosphine at 650C
DOPANT INCORPORATION
Case 1 Vapor pressure of dopant is low compared to group III species
Xdopantsolid= Jdopant/JGroupIII= Pdopant/PGroupIII
Case 2 Vapor perssure of dopant is high compared to group III
(Most typical Zn,Mg doped GaAs, InP, GaN)
Xdopantsolid== kdistribution*Pdopant/PGroupIII
For example if kdist =1e-2 you must add 100 times more DEZn to get 1 to incorporate
Memory effect of Cp2Mg (adsorption
on reactor surfaces)
OFF ON
17
P-type doping of GaN
Maximum in hole conc [p] peaks and is
not equal to maximum of {Mg} conc.
[p]
{Mg}
Mg precipitates
18
N-type doping of GaN
19
How does MOCVD Process Works??
SIMPLY VIEW
MOCVD Crystal Growth of GaAs at the Surface
Actual MOCVD reaction pathways
Gas phase and surface reaction
The basic reaction describe GaN growth can simply write as
Ga(CH3)3+NH3 → GaN+3CH4
The growth procedure as follows:
1. MO sources and hydrides inject to the reactor.
2. The sources are mixed inside the reactor and transfer to the deposition area
3. At the deposition area, high temperature result in the decomposition of
sources and other gas-phase reaction, forming the film precursors which
are useful for film growth and by-products.
4. The film precursors transport to the growth surface
5. The film precursors absorb on the growth surface
6. The film precursors diffuse to the growth site
7. At the surface, film atoms incorporate into the growing film through
surface reaction
8. The by-products of the surface reactions absorb from surface
9. The by-products transport to the main gas flow region away from the
deposition area towards the reactor exit
MOCVD Chemical reaction Types
Special Chemical Considerations
-Carbon, Oxygen Contamination
Typical MOCVD process for GaN and related materials
Two Step MOCVD Growth procedure
TMGa
NH3
1150oC 1050oC
Temperature 550oC
High temperature Buffer Epilayer
treatment layer Growth
Ga(CH3)3(vapor)+NH3(vapor) GaN(solid)+ 3CH4 (vapor)
2-step Buffer Layer process for AlN and GaN
Low temp 1st /high temp 2nd
Some basic problem related to GaN growth
MOCVD and other epitaxial techniques have developed more than 50 years, but
high quality GaN and related compound only available in since 1993. There are
some special problems for GaN and related materials.
•No suitable substrate (bulk GaN very expensive)
•Difficult to obtain p-type epilayer
•High Reactor Temperatures (>1050C for GaN and >1100C for AlN)
•InGaN difficult to incorporate Indium (T< 900C)
•Lattice Mismatch of AlN, GaN and InN is large (>5-11%)
Lattice Mismatch of GaN and InGaN is large (~2.3%)
460nm blue LED
%Strain =100*(aepi-asub)/(asubstrate)
For blue LED %strain = (3.275-3.2/3.2)*100 = 2.34%
Vegard’s Law for Alloys
Vegard's law for alloyed materials is a linear
function: Ealloy = xEA + (1 – x)EB, where EA, EB, and
Ealloy are the respective properties of pure A, pure B, and
the alloy AxB1–x, and x is the fraction of one ingredient in
a material point
Critical Layer Thickness in Epitaxial Growth
During epitaxial growth first few layers are coherent with a substrate
crystalline structure. As film thickness increases, growing stress causes
nucleation of dislocations. This partially relaxes the strain due to lattice
mismatch. A thickness at which this occurs is defined as a critical
thickness.
Recent developments and future directions in the growth of nanostructures by
van der Waals epitaxy
Utama, MIB; et al 2013 | NANOSCALE 5 (9) , pp.3570-3588
For blue LED xIn=18% so Hcritical ~ 8nm (80A)
Matthews-Blakeslee Limit
8nm
