Last Update -- BobVitale? - 27 Sep 2007

MOCVD - Metal-Organic Chemical Vapor Deposition

Table of Contents

Personnel & Contacts

Below is a list of persons involved with the BSOE MOCVD/ALD lab. For convenience and record keeping, an email list has been established bsoe_mocvd -at-

Name email Notes
UCSC Personnel    
Nobuhiko "Nobby" Koboyashi Professor
Bob Vitale Lab Director & SOE Project Coordinator
    SOE Backup to Bob for Project
Wenbo Yuan UCSC Project Manager/Architects Office
Steve Paul PP&C Supervisor Mechanical Engineer
Rhonda Tramble UCSC Electrical Engineer
Brent Cooley EH&S Industrial Hygienist
Buddy Morris   EH&S Director
Rick Rodewald Fire Marshal & Asst Fire Chief
Engineering Team    
Rizik Michael (Integrated Engineering Services) Engineering Code Consulting and Design
Jack Emerson (GLP Architects) GLP Architects - Alt2/3 Architect
Martin Gicklhorn (GLP Architects) GLP Architects - Alt2/3 Architect
MOCVD Manufacturing Structured Materials  
Gary Provost MOCVD Designer
Dr. Gary S. Tompa President of Stuctured Materials  

Project Description - SOE MOCVD Lab System

Lab Objective

The proposed modifications to laboratory space in the Baskin Engineering building will seek to accommodate a Group III-IV material Metal Organic Chemical Vapor Disposition (MOCVD) reactor vessel and an Atomic Layer Disposition (ALD) reactor for metal oxides. These systems will be used to grow thin films and study how they form in order to promote the advancement of nanotechnology. Detailed information on the lab objective is listed below

* MOCVD Lab Objective

An explanation of typical MOCVD processes can be found on wikipedia at:

Location Issues

The proposed lab should be located in Baskin Engineering Building to meet current building fire codes while keeping construction costs low. We would also like to maximize flexiblity of chemicals in known control zones as we do not yet know what in coming researchers will require. According to the UCSC Fire Marshal, Baskin Engineering Building (BEB) is a B-rate occupancy building with 2 known control zones and one large remainder zone. Per current codes, a B-rated building is allowed a maximum of 4 control zones and thus BEB already has 3 zones. The three known control zones are (a) 2nd floor cleanroom + unfinished area; (b) Alts 2/3 space; (c) remainder of building. The FM would be very happy if a 4th zone was prepared and finished in what is considered the remainder of the building. That is not a required condition for lab location but would be a bonus.

As this lab will have an appreciable chemical load and likely the limit on pyrophoric material for a zone, it is preferable to site the lab outside the Alts 2/3 space. The FM indicated there may be too many researchers in one zone and this may be difficult to keep some classifications of materials under the limit.

After the 6/14/07 meeting, the consultants were asked to look primarly at the unfinished clean room space and a location on the 3rd floor. The intent is to develop a location that would be less costly for construction. Floor plans may be reviewed at:

IES and GLP asked for specific locations on 3rd floor to look at. The below draft document was drafted by Vitale and sent to IES, GLP and Ricker (cc'd to Nobby, BSOE Dean and BSOE Asst Dean).

Proposed Site Plans

MOCVDLab-BE258-Option1.pdf: Option 1: BE-258 Only (Gas Cabinet in Lab)
Concern about exhausted enclosure partially blocking second exit (out back door)

MOCVDLab-BE258-260-Option2.pdf: Option 2: BE258/260 (Gas Cabinet in Room Next Door)
Distance between MOCVD and gas cabinets + agents may require longer lines than desired. This also permits longer lines of hazardous materials.

MOCVDLab-BE340-Option3.pdf: 3rd Level Location (BE-340)
Building codes may require reduced levels of materials. The reduction is likely lower than what we need to achieve.

Project Status and To Dos


8/21/07 - Rizik told Wenbo that preliminary design should be ready next week and cost estimates in 1.5 weeks

* MOCVDlabPastStatus - past status updates.

To Do List

Who What
Rizik preliminary design & Costing
Rhonda electrical load on existing emergency generator
?? can existing emergency generator be upgraded to supply more power
Nobby/Bob spec out MOCVD with appropriate changes recommended by Rizik
Dean/Asst Dean Determine funding approach
Bob Write up ACF proposal

Design Elements

Lab Chemicals

Chemical Listing

SOE facilities has a list of chemicals for the MOCVD lab on our server under /admin/facilities/SOEprojects/MOCVD-Lab/Chemicals

  • We are limited to 1lb (440 grams) of pyrophoics in the control area. We need to total the container size not the amount of material in the container. We need to find smaller containers (100g vs 200g).

Note: Sigma-Aldrich sells this material in two different packages. Examples below for trimethylaluminum (TMAl): 100g - Surpack 20g - steel container

Need to verify with MOCVD manufacturer that these supplies will work.

Hydrogen Gas Limitations

  • Volume of H2 gas must not exceed 3,000 cubic feet (c.ft). Each cylinder contains 237-240 ft3, two sixpacs of gas would be 240*6* 2 = 2800 c.ft flammable gas in the control area. This might use all gas for the clean room. This limits flammable gas in the clean room area. Table 3B of building code; 750 cft is allowed; footnotes 6 & 10; sprinklers allow doubling to 1500, gas cabinet allows an additional doubling to 3,000 c.ft.

  • We might consider installing the H2 gas supply in another control area or outside the building and run a pipe to the MOCVD lab.

H2 Gas Use (from Nobby email 8/13/07)

A. Please define approx H2 consumption per week and per month.
Once I reach regular operation mode, I would think I run a two-hour process two times a day so, I would say, 10-slm x 120 minutes/run x 2 runs/day x 2 days/week x 4 weeks/month ~ 19200 liters/month. If this exceeds any number imposed by the building code, I will cut down the usage.

B. Please define peak flow rate of H2.
It would be around 10 (base flow) + 3 (process flow) slm depending on CVD processes. Gary (SMI) would be the best person who would be able to answer this question. The base flow rate 10 slm may not be necessary for the SMI reactor to establish a stable process in terms of its hydrodynamics.

Metal Organic Vendor Recommendations from Gary Provost of SMI

1) Epichem Group (now SAFC Hightech with sigmaaldrich)

  • Minimum container size from Sigma-Aldrich Technical Services: 100g

2)Rohm Haas

Original Message -------- Subject: 257222-100G Resent-Date: Mon, 18 Jun 2007 06:21:41 -0700 Resent-From: <> Date: Mon, 18 Jun 2007 08:26:28 -0500 From: Sigma-Aldrich Technical Service <> Reply-To: To:

Dear Randy,

Thank you for contacting Sigma-Aldrich Technical Service. The smallest unit size that we offer of product number 257222 is the 100g. It is packaged in a 280 mL capacity carbon steel cylinder known as the Sure-Pac. Please let us know if you have any additional questions.

Best regards, Nicole Hanson Scientist Sigma-Aldrich Technical Service

TechServ? _Number: 257222-100G TechServ? _Name: trimethylaluminum

Here are the conversions of grams to liquid measures of the chemicals that Prof.Kobayashi will be using in his lab:
From Randy Porter, Mon, 21 Aug 2006 16:26:06 -0700 ,BR>

  • trimethylaluminum: 150g = 189ml (density = 0.79grams/ml)

  • Trimethylgallium: 150g = 135 ml. (density = 1.1g/ml)

  • tertiarybutylphosphine: 150g = 190ml (density = 0.79g/ml)

  • tertiarybutyl--sine: 150g = 95ml (density = 1.35g/ml)

  • diethylzinc: 150g = 125 ml (density = 1.182 g/ml)

MOCVD Elements

MOCVD Facilities Requirements and Design Drawings

  • Size: 44'wide x 88"long x 76.5in high for moving
  • Footprint: 44" wide x 88" long x 81" high (frame size)
  • Electrical: 208VAC, 3 Phase, 70 amps (100 amp is better)
  • Compressed Air: 80-120 psi dry
  • Cooling Water: 2-4 gpm (recommended)
  • Process Exhaust: 1000 cfm
  • Weight: less than 2000lbs
  • Hazmat Abatement Plan - not posted (SMI Proprietary, ask SOE facilities if needed)
  • Sketch and Flow Diagram - not posted (SMI Proprietary, ask fSOE facilities if needed)

Frame Review for California Seismic Code

Frame design of MOCVD needs to be reviewed by a California Registered Structural Engineer to ensure fame meets seismic code.

Vacuum Pump

This pump evacuates the reactor chamber. Rizik and SM recommend that a dry pump be used however the standard base model of MOCVD comes with a wet pump.

Here are the various pumps available:

Toxic Gas Scrubber

There is a recommendation to have a dual TGA scrubber. Should the single scrubber become full during operation, this could force a shutdown and then a re-oxidation. Two options have been presented:
a. 2nd parallel scrubber
Nobby email 8/13/07 -- This is definitely the option I prefer. Please be aware that I cannot financially cover this part by my research fund as I mentioned before.
b. a small downstream (in-line scrubber)

Meeting Notes

* June 14, 2007, meeting with Fire Marshal Chuck Hernandez. Attending: Rizik Michael (IES), Jack Emerson (GLP Architects), Hernandez, Ricker, Vitale, Porter (UCSC) FMMeeting-070614-RizikNotes-BizHub35020070615092137.pdf: Fire Marshal Meeting Notes (from Rizik)

* April 20, 2007; initial meeting with Nobby and PP&C (Randy Porter brief notes)
Nobby gave brief presentation. General process is atomic layer deposition (ALD). Related processes are all characteriized by sources being deposited on a substrate. Nobby uses metallo-organic sources. Gases containing these as vapors sent into a reaction chamber where they are heated. They then decompose into --senic and phosphorous (or --sine and phosgine) plus various C-H compounds. Once decomposed they are deposited on a substrate as a film. |


  • MOCVD Bubbler Explaination from RPI