2
1.2 PFC Heat Transfer Fluid
Emissions
Perfluorocarbon (PFC) liquids were first used as heat
transfer media in the early 50s, primarily to cool
sensitive military electronics. PFC liquids were used
because they are chemically inert and non-flammable;
they have high dielectric strength and electrical
resistivity; and they evaporate cleanly. In the ’70s and
’80s these liquids debuted in the commercial sector
where they were used to perform thermal test and
reflow soldering operations and to cool
supercomputers, lasers, x-ray targets, etc. In the early
1990s, the semiconductor manufacturing industry
began using PFC liquids as heat transfer media to
control component temperatures in ion implanters, dry
etchers, deposition tools, steppers, automatic test
equipment (ATE) and other tools. Deionized aqueous
liquids (DI water and glycol) were being used in the
industry at that time but applications were emerging
that required the wider liquid range of PFCs. Today, as
operating temperature ranges continue to widen, the
functional limits of deionized liquids are being
challenged and PFCs are being used in an increasing
percentage of applications.
PFC liquids will evaporate when not contained. Given
the long atmospheric lifetime of the PFC vapors, the
vapors ultimately reach the upper atmosphere. Though
published atmospheric data for PFC liquids are
somewhat sparse, they indicate that PFC liquids have
atmospheric lifetimes and IR cross sections similar to
the PFC gases listed in the MOU. If the global warming
properties of a PFC material (or the vapor of that
liquid) have not been measured, it is standard practice
to substitute C
6
F
14
[6] data as a means of estimating the
environmental properties of that material. This practice
is generally sound, however, there are
perfluoropolyether liquids with GWPs as high as
11,500
[7] (the GWP of C
6
F
14
is 9,000 [8]).
On average, a typical high volume fab consumes 500
gallons of virgin PFC liquid during a year
[9] when no
manufacturing capacity is being added. Assuming
C
6
F
14
properties, this equates to 0.0079 million metric
tons of carbon equivalent (MMTCE) per fab per year.
As these fluids are rarely incinerated or recycled, a fab’s
fluid consumption equates roughly with what is emitted
to the atmosphere via equipment leaks, evaporation and
spills. Table 1 compares this PFC liquid emission to
total (gas + liquid) MMTCE emissions using MOU gas
emissions published by Beu.
These data show that even today emissions of PFC heat
transfer liquids are a significant percentage of a
facility’s MMTCE emissions.
2. Reducing PFC Liquid
Emissions
2.1 PFC Liquid Loss Prevention
A critical on-site review of a heat transfer system
installation can often reveal causes of fluid loss. Causes
of liquid leaks include inappropriate or damaged
connections, inappropriate valves or pumps and others.
An often overlooked mechanism is evaporative loss
resulting from temperature cycles.
[10] An
understanding of this mechanism can, in some
instances, reduce losses dramatically.
When effective, the loss reduction strategy will
immediately reduce liquid costs with a nominal
investment of time and little or no cost. It can also
reduce service time and prevent false Fluorine alarm
triggers, etc. However, the effectiveness varies widely.
On average, losses are reduced no more than 20%.
2.2 Alternative Liquids
The most effective strategy for reducing global
warming emissions resulting from PFC liquids is
replacement of these liquids with low-GWP
alternatives. While a variety of alternatives can satisfy
the temperature requirements of semiconductor
applications, many suffer limitations that make their
use impractical.
[11] For example, petroleum-based
products, silicone oils and citrus-based oils tend to be
flammable or unstable. Chlorofluorocarbons are ozone
depleting and other chlorinated materials like
dichloroethylene, trichloroethylene and methylene
chloride are either flammable, unstable or highly
regulated.
6
In lieu of accurate environmental data for PFCs of higher boiling point, the US EPA recommends use of perfluorohexane data (Inventory
of US Greenhouse Gas Emissions and Sinks: 1990-1996, US EPA Doc EPA236-R-98-006, 1998).
7
3M Technical Report, J. G. Owens and L. Tousignant, “Measured IR Cross-Sections and GWPs for FC-3283 and HT-135,” June 2001.
8
Based on an integration of the radiative forcing effect over a 100 year period or integrated time horizon (ITH). Scientific Assessment of
Ozone Depletion: 1998, WMO Global Ozone Monitoring Project, Report No. 44.
9
Based on 3M market studies.
10
Tuma, P.E. and Tousignant, L., Modeling Vapor Leak Rate and System Pressures for Single Phase Heat Transfer Systems that Utilize
Fluorinated Heat Transfer Liquids, To be published.
11
Tuma, P.E., “Segregated Hydrofluoroethers: Long Term Alternative Heat Transfer Liquids,” Proceedings of the 2000 Earth Technologies
Forum, Oct.30-Nov.1, Washington D.C., pp. 266-275.