Specific Process Knowledge/Thermal Process/Dope with Phosphorus

Feedback to this page: click here

The Phosphorus Predep furnace (A4) can be used for phosphorus predeposition of silicon wafers, resulting in N-type doping. In the furnace, the silicon wafers are positioned vertically in a quarts boat.

Test of the Phosphorus Predep furnace

Purpose

To study the correlation between the temperature for the predeposition process and drive-in of the phosphorus doping in the Phosphorus Predep furnace (A4) at DTU Danchip.

Experimental setup

20 boron doped device wafers (p-type) were used - Four wafers for each of the five different predeposition temperatures (see table below). In the furnace five dummy wafers were placed on each side of the device wafers. The dummy wafers nearest to the device wafers were changed in-between the runs to miniminze doping from these dummy wafers.

Run #	Temperature	Process time with POCl3	Annealing time in N2	Wafer #
1	850 oC	15 minutes	20 miuntes	1, 2, 3, 4
2	900 oC	15 minutes	20 miuntes	5, 6, 7, 8
3	950 oC	15 minutes	20 miuntes	9, 10, 11, 12
4	1000 oC	15 minutes	20 miuntes	13, 14, 15, 16
5	1050 oC	15 minutes	20 miuntes	17, 18, 19, 20

After the predeposition two wafers from each run were taken out to be further processed. These wafers were: 1, 2, 5, 6, 9, 10, 13, 14, 17, 18. These wafers were dipped in BHF to remove the phosphorus glass layer before the drive-in process.

The drive-in process was made in the Phosphorus Drive-in furnace (A3) for all the mentioned wafers at same time. At the drive-in process a dummy wafer was placed in-between the wafers that have been predeposited at different temperatures, so doping from wafer to wafer was minimized. The phosphorus drive-in was done with the process "DRY1050" which is a dry oxidation at 1050 ^oC for 100 minutes and 20 minutes annealing. At the oxidation the O₂ flow was 5 SLM, and the N₂ flow for annealing was 3 SLM.

Results

Several measurements were done for the different device wafers. After the predeposition, the thickness of the grown phosphorus glass layer was measured, and the sheet resistance and slice resistivity were measured on the same wafer after a BHF etch.

Measurements after phosphorus predeposition

		Ellipsometer (center point only)		Four Point Probe
Wafer #	Temperature [C]	Oxide thickness [nm]	Refrative index	Sheet resistance [Ωsq]	Slice Resistivity [Ωcm]
3	850	27,4	1,4623	311	6,22*10-3
7	900	45,27	1,4622	138,5	4,85*10-3
11	950	61,36	1,4625	16,1	1,29*10-3
15	1000	80,45	1,4624	7,4	0,89*10-3
19	1050	119,37	1,4623	6,6	1,49*10-3

The oxide layer was etch in BHF before the drive-in of the doping with a dry oxidation at 1050 ^oC for 100 minutes and 20 minutes annealing.

Measurements after phosphorus drive-in

		Avg. five point on Filmtek		Four Point Probe
Wafer #	Temperature [oC]	Oxide thickness [nm]	Refrative index	Sheet resistance [Ωsq]	Slice Resistivity [Ωcm]
2	850	110,44	1,4654	189,7	3,79*10-3
6	900	116,28	1,4629	101,6	3,56*10-3
9	950	137,06	1,4604	10,0	0,8*10-3
14	1000	141,46	1,4651	4,72	0,57*10-3
18	1050	139,87	1,4659	3,23	0,73*10-3
Blank Si wafer		110,71	1,46286	-	-

Resistivity	Sheet resistance
SIMS Measurement After Pre-dep	SIMS Measurement After Drive-in Process at 1050 oC

Looking at the "SIMS Measurement After Drive-in Process at 1050 ^oC" graph, it can be seen that the two curves "Pre-dep at 950 ^oC" and "Pre-dep at 1000 ^oC" are crossing each other, but according to the theory they should not do that. Only one wafer has been meassured, so there is not that much statistical data to verify it with.

Doping profile after Phosphorus Predep process

" Result from GreenBelt, @Danchip,Jan 2015"

Process parameters

Recipe: POCL900

Temperature: 900 °C

Time: 15 min

Anneal: 20 min in N₂

After removing the oxide in BHF the wafers were oxidized in Phosphor Drive in Furnace (A3) using DRY1050 at 1050°C for 70 min without further annealing in N₂. Results are listed in the table below:

Measurements after phosphorus predeposition

	Afer predeposition	After Oxidation
Test date	Sheet resistance [Ω/sq]	Oxide thickness [nm]	Sheet resistance [Ω/sq
Aug2012	47.0	93.1	41.8
Jan2013	50.5	91.3	55.6
Aug2013	58.0	91.8	41.8
Jan2014	54.1	93.3	33.8
Jan2015	50.8	94.2	39.4

One wafer from Jan2015 was analyzed with SIMS and the profiled is shown on the graph below:

SIMS Measurement After Predeposition, before and after oxidation respectively

Result from Trine Holm Christensen, @Space, Feb. 2015

Process parameters

POCL1000 for 15 min, no anneal

WET1050 for 30 min in Furnace(A3)