1. Theory

Definition: Spin-on and gas-source doping systems supply dopant from a coated glass film or from a dopant gas at the wafer surface, then diffuse it in with a furnace or rapid thermal anneal, with no implant damage (Lee and Lee; Kiyota et al.).
Constant source: A spin-on dopant layer holds the surface near solid solubility, giving an erfc-type profile (Lee and Lee; Diffusion Furnace page).
Gas-phase doping: Diborane, phosphine or halide gases adsorb on a clean surface and diffuse in during a short anneal (Kiyota et al.; TI patent).
Damage-free: Without implant damage there is no transient enhanced diffusion, so very shallow junctions are possible (Kiyota et al.; my reading).

2. Working Principle

Coat: Spin the dopant solution at about 3000 $rpm$ and bake at 200 $^{\circ} C$ to remove solvent (Lee and Lee).
Diffuse: RTP at 1000 to 1100 $^{\circ} C$ for 1 to 3 minutes, or a furnace for hours (Lee and Lee; Cal Poly).
Gas dose: Flow diluted B2H6 or a halide over a hydrogen-cleaned surface at about 900 $^{\circ} C$ (Kiyota et al.; TI patent).
Strip and measure: Remove the dopant glass in dilute HF and measure sheet resistance with a four-point probe (PV-Manufacturing; Cal Poly).

3. Applications

Solar cells: Boron back-surface fields and phosphorus emitters (Lee and Lee).
Ultra-shallow junctions: Source and drain extensions without implant damage (Kiyota et al.; TI patent).
Research and teaching labs: Doping without an implanter (Cal Poly).
Monolayer doping: Molecular monolayers of phosphorus or boron give sub-5 $nm$ junctions after a spike anneal (Ho).

Spin-On Doping

Name	Formula	Related Formulae	Variables/Constants
Spin-coat thickness	$h \propto ω^{- 1/2}$	$ω = 3000 rpm, 750 μ l per 100 mm wafer$	$h : d o p an t f i l m t hi c k n ess$ $ω : s p in s p ee d; M i cr o C h e mi c a l s : t hi c k n ess f a l l s w i t h t h e r ec i p r oc a l s q u a r e r oo t o f s p in s p ee d; L ee an d L ee : 3000 rpm f or 5 s t h e n 15 s, bak e o n a h o t pl a t e a t 200^{\circ} C f or 15 min t o d r i v e o u t so l v e n t; S O D s B or o n A (ab o v e 4.0 p er ce n t B), B 155 (4.0), B 153 (2.0), B 0.008, A l / B 120, A l / G a 130; C a l P o l y : F i l m t r o ni cs B 155 an d P 507$ $(co m p u t e d : d o u b l in g s p in s p ee d t hin s t h e f i l m b y a f a c t or o f 1.41)$
RTP diffusion from SOD	$T = 1000 to 110 0^{\circ} C, t = 60 to 180 s$	$\dot{T} = 100 K/s, N_{2} ambient$	$L ee an d L ee : b or o n S O D d i f f u se d b y R T P; p h os p h or u s S O D a t 850 t o 1025^{\circ} C f or 5 t o 80 s; so l a r - ce l l se q u e n ce B ba c k - s u r f a ce f i e l d a t 1100^{\circ} C f or 2 min, P e mi tt er a t 950^{\circ} C f or 50 s, R T O a t 1050^{\circ} C f or 2 min; C a l P o l y : f u r na ce 900 t o 1200^{\circ} C f or 0.5 t o 2 h, l o a d e d a t 900^{\circ} C, ni t r o g e n r am p s an d o x y g e n d u r in g p r ocess in g$ $(m y r e a d in g : t h e S O D l a y er a c t s a s a co n s t an t s u r f a ce so u r ce, so t h e er f c so l u t i o n a ppl i es)$
SOD layer results	$C_{S} = 2 to 3 \times 1 0^{20} c m^{- 3}, x_{j} = 0.5 to 0.65 μ m$	$R_{S} < 40 Ω/ □ (105 0^{\circ} C, 60 s), R_{S} < 20 Ω/ □ (110 0^{\circ} C, 120 s)$	$C_{S} : s u r f a ce co n ce n t r a t i o n b y E C V$ $R_{S} : s h ee t r es i s t an ce; L ee an d L ee : C_{S} f e l l t o 1 \times 1 0^{20} c m^{- 3} a f t er t h r ee l a t er R T P s t e p s; x_{j} q u o t e d a s 0.5 t o 0.6 μ m (ab s t r a c t), 0.5 t o 0.65 μ m (sec t i o n I I I) an d 0.5 t o 0.7 μ m (co n c l u s i o n); f u r na ce C T D e mi tt er ab o u t 120 Ω/ □; B 0.008 an d A l / G a 130 r e a d in g s u n r e l iab l e$ $(a ss u m e d : h o l e m o bi l i t y 50 c m^{2} / V s; co m p u t e d : 20 Ω/ □ cor r es p o n d s t o ab o u t 6.2 \times 1 0^{15} c m^{- 2} o f a c t i v e b or o n)$

Profile and Measurement Checks

Name	Formula	Related Formulae	Variables/Constants
erfc check of SOD depth	$x_{j} = 2 D t erfc^{- 1} (\frac{C _{B}}{C _{S}})$	$D_{n ee d e d} = \frac{1}{t} [\frac{x _{j}}{2 erfc ^{- 1} ( C _{B} / C _{S} )}]^{2}$	$C_{B} : s u b s t r a t e d o p in g; co n s t an t - so u r ce so l u t i o n (D i f f u s i o n F u r na ce p a g e)$ $(a ss u m e d : C_{B} = 1 \times 1 0^{16} c m^{- 3}, C_{S} = 2.5 \times 1 0^{20} c m^{- 3}, 1100^{\circ} C f or 120 s; co m p u t e d : t h e in t r in s i c b or o n D o f 1.43 \times 1 0^{- 13} c m^{2} /s p r e d i c t s o n l y 0.24 μ m, an d 0.6 μ m n ee d s 8.9 \times 1 0^{- 13} c m^{2} /s, 6.2 t im es l a r g er)$
Four-point sheet resistance	$R_{S} = \frac{π}{l n 2} \frac{V}{I} = 4.532 \frac{V}{I}$	$ρ = R_{S} t_{l a y er}$	$V / I : v o l t a g e o v er c u r r e n t b e tw ee n inn er an d o u t er p r o b es; C a l P o l y : f a c t or 4.532, r es i s t i v i t y f r o m s h ee t r es i s t an ce t im es an 8 μ m d i f f u se d l a y er, j u n c t i o n s in co n c l u s i v e b ec a u se o f p oor s t ainin g (F o u r - P o in t P r o b e p a g e o f t hi s v a u l t f or g eo m e t r y cor r ec t i o n s)$ $(co m p u t e d : π / ln 2 = 4.5324)$
Phosphorus against boron diffusivity	$\frac{D _{P}}{D _{B}} = \frac{D _{0, P}}{D _{0, B}} e^{- (E_{P} - E_{B}) / k_{B} T}$	$E_{P} = 3.66 eV, E_{B} = 3.46 eV$	$C a l P o l y : a c t i v a t i o n e n er g i es a s l i s t e d, D_{0} u p t o 3.9 c m^{2} /s, p h os p h or u s t o t a l d i f f u s i v i t y hi g h er t han b or o n b y u p t o 19 p er ce n t in o n e sec t i o n an d b y u p t o a f a c t or o f 3 in an o t h er; b or o n so l u bi l i t y ab o u t t e n t im es t ha t o f p h os p h or u s a s s t a t e d$ $(co m p u t e d : w i t h t h e n e u t r a l - co m p o n e n t se t (P 3.9, B 0.037) t h e r a t i o i s 17 a t 1000^{\circ} C an d 19 a t 1100^{\circ} C; w i t h t h e S t an f or d t o t a l s (P 4.70/3.68, B 1.0/3.5) i t i s 0.91 an d 1.03)$

Gas-Source Doping

Name	Formula	Related Formulae	Variables/Constants
Rapid vapor-phase doping	$T \approx 90 0^{\circ} C, t = 60 s, B_{2} H_{6} / H_{2}$	$x_{j} < 60 nm, C_{p e ak} = 1 0^{19} to 1 0^{20} c m^{- 3}$	$K i y o t a e t a l . : b or o n s h ee t co n ce n t r a t i o n se t b y t h e d ib or an e f l o w, P M O S t h r es h o l d r o l l - o f f s u pp r esse d d o w n t o 0.22 μ m g a t e l e n g t h a g ain s t 0.5 μ m w i t h B F_{2} im pl an t s, d r ain c u r r e n t 1.5 t im es hi g h er (t e m p er a t u r e u ni t l os t in e x t r a c t i o n, r e a d a s d e g r ees C)$ $(m y r e a d in g : n o im pl an t d ama g e m e an s n o t r an s i e n t e nhan ce d d i f f u s i o n, w hi c h k ee p s t h e j u n c t i o n s ha l l o w)$
Halide gas sources	$B C l_{3} : T_{b} = 12. 5^{\circ} C; P F_{3} : T_{b} = - 101. 5^{\circ} C; P C l_{3} : T_{b} = 75. 5^{\circ} C$	$x_{j} ≲ 1000 \overset{˚}{A}$	$T_{b} : b o i l in g p o in t; T I p a t e n t : p r i or - a r t B_{2} H_{6} / H_{2}, P H_{3} / H_{2}, A s H_{3} an d T B A; p r o p ose d B C l_{3}, A s F_{5}, P F_{3}, P F_{5}, P C l_{2} F_{3} g a ses an d h e a t e d P C l_{3} v a p or w i t h H_{2} c a r r i er an d o pt i o na l r e m o t e h y d r o g e n pl a s ma, f l u or i d e so u r ces r e m o v in g t h e se p a r a t e na t i v e - o x i d e c l e an, im pl an t a t i o n l imi t e d b e l o w 10 k e V f or j u n c t i o n s s ha l l o w er t han 1000 \overset{˚}{A}$ $(m y r e a d in g : c h e mi sor pt i o n a t l o w t e m p er a t u r e f o l l o w e d b y d i f f u s i o n a t hi g h er t e m p er a t u r e se p a r a t es d ose co n t r o l f r o m d e pt h co n t r o l)$
Adsorption-limited dose	$Q \leq θ N_{s}, N_{s, S i (100)} = 6.78 \times 1 0^{14} c m^{- 2}$	$Q_{t o t a l} = n_{cy c l es} θ N_{s}$	$θ : f r a c t i o na l d o p an t co v er a g e p er a d sor pt i o n cy c l e$ $N_{s} : s u r f a ce s i t e d e n s i t y; H a d l ey : S i (100) s u r f a ce a t o m d e n s i t y 6.78 \times 1 0^{14} c m^{- 2}; H o : m o n o l a y er d o p in g o f P an d B g i v in g s u b - 5 nm j u n c t i o n s a f t er s p ik e ann e a l; T I p a t e n t : r e p e a t ab l e c h e mi sor pt i o n an d d i f f u s i o n cy c l es$ $(co m p u t e d : o n e f u l l m o n o l a y er i s 6.8 \times 1 0^{14} c m^{- 2}, ab o u t o n e t e n t h o f t h e S O D l a y e r^{'} s 6.2 \times 1 0^{15} c m^{- 2})$

Important Questions

Ques: How does a spin-on dopant layer form a junction?

The dopant glass is spun on, baked and then annealed, acting as a constant surface source. Lee and Lee spun 750 $μ l$ of boron SOD per 100 mm wafer at 3000 $rpm$ , baked at 200 $^{\circ} C$ for 15 minutes and diffused by RTP at 1000 to 1100 $^{\circ} C$ for 60 to 180 $s$ . They obtained surface concentrations of 2 to $3 \times 1 0^{20}$ $c m^{- 3}$ , sheet resistances below 40 ohm per square after 60 $s$ at 1050 $^{\circ} C$ and below 20 after 120 $s$ at 1100 $^{\circ} C$ . Twenty ohm per square corresponds to about $6.2 \times 1 0^{15}$ $c m^{- 2}$ of active boron for an assumed hole mobility of 50 $c m^{2} / (V s)$ (computed). The furnace-diffused BBr3 reference emitter was about 120 ohm per square. The dopant glass is then stripped in dilute HF (PV-Manufacturing). Film thickness falls as the inverse square root of spin speed, so doubling the speed thins it by about 1.4 (MicroChemicals). Reference: J. Y. Lee and S. H. Lee, ‘Boron Back Surface Field Using Spin-On Dopants by Rapid Thermal Processing’, Journal of the Korean Physical Society 44, 1581 (2004) (fetch-tool summary); MicroChemicals GmbH, ‘Fundamentals of Microstructuring’, chapter 1 (spin coating), hosted by Arizona State University cores: thickness falls with the reciprocal square root of spin speed, doubling speed thins by about 1.4 (fetch-tool summary); PV-Manufacturing.org, ‘Tube diffusion’: dopant glass removed in dilute HF for 1 to 2 min (fetch-tool summary).

Ques: Why are SOD junctions deeper than the intrinsic erfc model predicts?

Lee and Lee measured junctions of 0.5 to 0.65 $μ m$ after 120 $s$ at 1100 $^{\circ} C$ , but an erfc profile with the intrinsic boron diffusivity of $1.43 \times 1 0^{- 13}$ $c m^{2} /s$ gives only 0.24 $μ m$ for an assumed $1 0^{16}$ $c m^{- 3}$ substrate (computed). Reaching 0.6 $μ m$ needs a diffusivity 6.2 times larger (computed). At $2 \times 1 0^{20}$ $c m^{- 3}$ the boron concentration far exceeds the intrinsic carrier concentration, so charged-defect diffusion and the built-in field speed diffusion up (Diffusion Furnace page; my reading). Ramp heating at 100 $K/s$ adds further time. The paper itself quotes the depth as 0.5 to 0.6, 0.5 to 0.65 and 0.5 to 0.7 $μ m$ in different sections, so 0.5 to 0.65 $μ m$ , the value in its results section, is more likely the measured range (my reading). Reference: J. Y. Lee and S. H. Lee, ‘Boron Back Surface Field Using Spin-On Dopants by Rapid Thermal Processing’, Journal of the Korean Physical Society 44, 1581 (2004) (fetch-tool summary); Diffusion Furnace page of this vault.

Ques: Does phosphorus diffuse 19 per cent faster than boron, or 19 times faster?

The Cal Poly report says phosphorus total diffusivity exceeds boron’s by up to 19 per cent in one section and by up to a factor of 3 in another. Its activation energies, 3.66 $eV$ for phosphorus and 3.46 $eV$ for boron, match the neutral-component set with $D_{0}$ of 3.9 and 0.037 $c m^{2} /s$ , which gives a phosphorus-to-boron ratio of 17 at 1000 $^{\circ} C$ and 19 at 1100 $^{\circ} C$ (computed). The ‘19 per cent’ is therefore more likely a misprint for a factor of 19 (my reading). With the Stanford total intrinsic values instead, the ratio is 0.91 to 1.03 (computed), meaning the two dopants diffuse at almost the same rate, which is the more realistic picture for lightly doped silicon (Diffusion Furnace page; my reading). Reference: D. Gonzalez and H. Le (advisor J. Lee), ‘Characterization of Thermal Diffusion Profiles of Boron and Phosphorus Dopants in Silicon and Silicon Dioxide’, Cal Poly Project Expo (2021) (fetch-tool summary); Diffusion Furnace page of this vault.

Ques: How does rapid vapor-phase doping make shallow junctions?

By adsorbing boron from diborane on a clean surface and diffusing it briefly without implant damage. Kiyota et al. doped at about 900 $^{\circ} C$ for 60 $s$ with B2H6 in hydrogen, controlling the sheet boron concentration through the diborane flow, and obtained junctions below 60 $nm$ with peak concentrations of $1 0^{19}$ to $1 0^{20}$ $c m^{- 3}$ . PMOS transistors with these source and drain regions suppressed threshold roll-off down to 0.22 $μ m$ gate length, against 0.5 $μ m$ for $B F_{2}$ -implanted devices, and gave 1.5 times higher drain current. Since no implant damage is present, there is no transient enhanced diffusion during the anneal, which is likely the reason for the shallower profile (my reading). The temperature unit was lost in the abstract text and is read here as degrees Celsius. Reference: Y. Kiyota, T. Nakamura, S. Suzuki and T. Inada, ‘Shallow p-Type Layers in Si by Rapid Vapor-Phase Doping for High-Speed Bipolar and MOS Applications’, IEICE Trans. Electron. E79-C, 554 (1996), abstract (fetch-tool summary).

Ques: Why use halide gas sources and adsorption-limited cycles?

Texas Instruments’ patent replaces diborane, phosphine and arsine with halides such as BCl3 (boiling at 12.5 $^{\circ} C$ ), PF3, PF5, AsF5 and PCl2F3, or heated PCl3 vapor, with hydrogen carrier and an optional remote hydrogen plasma. Fluoride sources remove native oxide, eliminating a separate clean, and the process alternates low-temperature chemisorption with higher-temperature diffusion to reach junctions of about 1000 angstrom or less, where implants below 10 $keV$ become difficult. Each adsorption cycle is limited by surface sites, about $6.78 \times 1 0^{14}$ $c m^{- 2}$ on Si(100) (Hadley), so the dose per cycle is self-limiting and repeatable (my reading). Ho’s molecular monolayer doping of phosphorus and boron reaches sub-5 $nm$ junctions. One monolayer is about a tenth of the active dose of Lee and Lee’s 20 ohm per square SOD layer (computed). Reference: M. M. Moslehi, ‘Method for doped shallow junction formation using direct gas-phase doping’, US patent 5,599,735, Texas Instruments (filed August 1994) (fetch-tool summary); P. Hadley, TU Graz course 513.160 materials page ‘Silicon’: a = 0.543 nm, 4.995e22 atoms/cm3, surface atom density 6.78e14 cm-2 on (100), 9.59e14 on (110), 7.83e14 on (111) (fetch-tool summary); J. C. Ho, ‘Controlled Nanoscale Doping of Semiconductors via Molecular Monolayers’, Berkeley Sensor and Actuator Center abstract BPN463 (2009): monolayer doping of P and B, sub-5 nm junctions on 4 inch wafers, spike anneal (fetch-tool summary); J. Y. Lee and S. H. Lee, ‘Boron Back Surface Field Using Spin-On Dopants by Rapid Thermal Processing’, Journal of the Korean Physical Society 44, 1581 (2004) (fetch-tool summary).

References

J. Y. Lee and S. H. Lee, “Boron Back Surface Field Using Spin-On Dopants by Rapid Thermal Processing,” Journal of the Korean Physical Society, vol. 44, no. 6, pp. 1581-1586, 2004. Available: https://www.eet.bme.hu/~mizsei/Nanoelektronika,%20nanotechnol%c3%b3gia/Nanoelektronika/RTP/J012004044R061581.pdf
MicroChemicals GmbH, “Fundamentals of Microstructuring,” chapter 1, spin coating. Available: https://cores.research.asu.edu/sites/default/files/2025-01/spin_coating_photoresist.pdf
P. Hadley, “Silicon,” Microelectronics and Micromechanics course materials, TU Graz. Available: https://lampz.tugraz.at/~hadley/memm/materials/silicon/silicon.php
J. C. Ho, “Controlled Nanoscale Doping of Semiconductors via Molecular Monolayers,” Berkeley Sensor and Actuator Center, BPN463, 2009. Available: https://bsac.berkeley.edu/node/1158
PV-Manufacturing.org, “Tube diffusion.” Available: https://pv-manufacturing.org/solar-cell-manufacturing/tube-diffusion/
Diffusion Furnace page of this vault.
D. Gonzalez and H. Le, “Characterization of Thermal Diffusion Profiles of Boron and Phosphorus Dopants in Silicon and Silicon Dioxide,” Cal Poly Project Expo, 2021. Available: https://projectexpo.calpoly.edu/2021/characterization-of-boron-and-phosphorus-dopant-diffusion-profiles-in-si-and-sio2/
Y. Kiyota, T. Nakamura, S. Suzuki and T. Inada, “Shallow p-Type Layers in Si by Rapid Vapor-Phase Doping for High-Speed Bipolar and MOS Applications,” IEICE Trans. Electron., vol. E79-C, no. 4, pp. 554-559, 1996. Available: https://globals.ieice.org/en_transactions/electronics/10.1587/e79-c_4_554/_p
M. M. Moslehi, “Method for doped shallow junction formation using direct gas-phase doping,” US Patent 5,599,735, Texas Instruments, 1997. Available: https://patents.google.com/patent/US5599735

Quartz 5

Explorer

Spin-On and Gas-Source Doping System