Are Bacteria Really a Problem to be Feared in a Well-Managed Ultrapure Water System?

The UPW International Roadmap for Devices and Systems, or UPW IRDS, has removed bacteria as a roadmap parameter. Why? Does that mean that bacteria are no longer an important contaminant in UPW?

The UPW IRDS experts agreed that controlling bacteria is not a technology challenge.  Instead, it’s a task to be addressed by proper system design and operating practices, as partially addressed in the SEMI F61 standard. Nevertheless, the fear seems to still be there…

Let’s take a step back and look at two misguided but common perspectives regarding bacteria in ultrapure water:

Ultrapure Water Myths

Myth 1

UPW is too pure for bacteria to grow, as there are not enough nutrients present. 

Reality

It has been well demonstrated that bacteria are capable of surviving and reproducing in high purity water.  (1) (2) (3) However, they are limited by two key characteristics:

1. Slow Growth Rate – The nutrient content and types of bacteria in UPW result in slow reproduction.  Studies support a doubling rate of about once every 24 hours. (2) (3)  

2. Efficient removal by semiconductor final filters – Bacteria are essentially particles with a diameter of 0.1 – 1.0 microns (4). If a semiconductor system is designed to comply with a particle specification of less than 1000 particles per liter at the critical size (typically 0.05 microns and smaller), then the final filters should remove nearly all bacteria.

Myth 2

Bacteria are a high-risk contaminant in UPW. They require constant sanitization, highly turbulent flows, and 254 nm ultraviolet radiation (UV) to keep them in control. 

Reality

These excessive bacterial control strategies, which were once considered best–known methods, are now thought to be out-of-date.

These practices do not help control bacteria, but do result in wasted CAPEX and OPEX:

1. 254 nm UV

Contrary to what used to be considered best practice, disinfection by 254 nm UV has not been demonstrated to control bacteria in UPW.  While it does deactivate or damage bacteria within the unit itself, the byproducts of bacteria after exposure to UV result in increased nutrient content downstream.  Bacteria counts have been found to increase in the process piping and equipment downstream of 254 nm UV.  (5)

2. Excessively high turbulence

Some outdated rules of thumb may have recommended velocities of at least 3 – 7 ft/sec or Reynolds numbers above 10,000 in hopes of high turbulence helping to remove bacteria from the pipe walls.  This concept has been disproven from a number of angles (6) (7) (8).  Independently of the level of turbulence in the pipe, the thickness of the stagnant layer is much larger than the size of the bacterium, suggesting inability to remove them by shear force. Many state-of-the-art semiconductor facilities have demonstrated success with systems designed for minimum Reynolds numbers in the range of 3,000-4,000, or even lower.

3. Frequent preventative sanitizations

Sanitizations are never prescribed in any modern high volume semiconductor manufacturer, aside from post-construction startup or re-qualification after maintenance activities. If the system is properly designed and operated, bacteria are well-controlled already, making preventative sanitizations unnecessary.

How should bacteria in UPW be dealt with?

In addition to final filtration, implement standard practices that are important for controlling bacteria. None of these concepts are new– they are likely to be already well established in UPW system construction and operation. 

• Materials of construction:  Use semiconductor grade materials with demonstrated performance regarding leachables and internal smoothness specifications.

• Dead legs:  Eliminate any section of pipe that is stagnant for a length of more than four pipe diameters. 

• Standard operating procedures: Establish and follow post-construction startup, ongoing maintenance, and shutdown SOP’s; including sanitization and qualification of newly installed or temporarily stagnant equipment and piping.

• A robust particle monitoring program:  Since bacteria are essentially particles, their numbers are included in compliance with particle specifications.

• Proper sampling and analysis techniques:  If bacteria monitoring is part of compliance with specifications for your UPW system, the correct sample location and type of valve, sample size, and incubation period or analysis technique can affect results significantly. (3)  These considerations are important to take into account when setting SOP’s and interpreting bacteria sampling results.

References

Analysis of Bacteria Contaminating Ultrapure Water in Industrial Systems. Kulakov, Leonid A, et al. 2002 Apr, 68(4), Applied and Environmental Microbiology, pp. 1548-1555.

M B McAlister, L A Kulakov, J F O’Hanlon, M J Larkin, K L Ogden. Survival and nutritional requirements of three bacteria isolated from ultrapure water. J Ind Microbiol Biotechnol. Aug, 2002.

Bacteria in Ultrapure Water: Myth Versus Reality, Part 2. Libman, Slava, Sullivan, Lindsey and Zerfas, Bernie. Portland, Or : Ultrapure Micro, 2016.

Isolation, identification, and microscopic properties of biofilms in high-purity water distribution systems. Patterson, Michael K, et al. s.l. : Tall Oaks Publishing, Inc, May/June 1991, Ultrapure Water, pp. 18-23.

Analysis of bacterial contamination in different sections of a high-purity water system. McAlister, M B, et al. s.l. : Tall Oaks Publishing, Inc., January 2001, Ultrapure Water, pp. 18-24.

An examination of pipe self-cleaning in a high purity water system. Klauer, J. s.l. : Tall Oaks Publishing, Inc., 2001, Ultrapure Water, pp. 56-60.

Part 1: Flow requirements, pressure differential, and pressure control of distribution systems. Buesser, David S. s.l. : Tall Oaks Publishing, Inc., November 2002, Ultrapure Water, pp. 41-48.

Use of Reynolds Number as a criteria for design of high purity water systems. Libman, Vyacheslav (Slava), Ph.D. s.l. : Tall Oaks Publishing, Inc., October 2006, Ultrapure Water, pp. 26-32.