Environmental technology transfer synergism: health risk assessment and community rejection of commercial electronic chip fabrication in upstate New York

Oct 30

424 Int. J. Technology Transfer and Commercialisation, Vol. 1, No. 4, 2002

Robert A. Michaels

RAM TRAC Corporation, 3100 Rosendale Road, Schenectady, NY 12309, USA E-mail: ram@ramtrac.com

Abstract: A prerequisite to technology transfer is community acceptance of a new technology to be transferred. Health risk assessment is the instrument typically used for evaluating technologies relative to public health issues, which often constitute one basis for community concern about a technology. Technology transfer, therefore, also depends upon community acceptance of risk assessment technology, which itself is new to many communities. Interdependence of a hardware technology, such as electronic chip fabrication supporting the computer industry, with a software technology, specifically health risk assessment, may be termed ‘technology transfer synergism’. Here, a case study is presented in which transfer of chip fabrication technology to an upstate New York community failed. The primary basis for this technology transfer failure was community rejection of the chip fabrication technology, not mistrust of risk assessment as an evaluating instrument.

Keywords: Technology transfer; chip fabrication; health risk assessment; technology transfer synergism; facility siting; facility permitting.

Reference to this paper should be made as follows: Michaels, R.A. (2002) ‘Environmental technology transfer synergism: health risk assessment and community rejection of commercial electronic chip fabrication in upstate New York’, Int. J. Technology Transfer and Commercialisation, Vol. 1, No. 4, pp.424-452.

Biographical notes: Dr. Michaels uses health risk assessment to assist communities to evaluate new technologies. By congressional invitation he testified in support of a major piece of technology transfer legislation: the Superfund Technology Demonstration Act. He has consulted for the US Congress, Office of Technology Assessment; and the California Governor’s Office of Appropriate Technology. He serves on the Editorial Board of the journal, Environmental Engineering and Policy and most recently has assisted the US AID to introduce comparative risk assessment technology to Caspian Sea nations to enable them to prioritise energy development options in that region. He is also president of RAM TRAC Corporation.

— os

Environmental technology transfer synergism 425

1 Introduction

‘Environmental technology’ equals ‘technology’. Most if not all technologies affect the environment. The adjective ‘environmental’ describing a technology connotes that the technology is not a source of environmental degradation but instead prevents or mitigates environmental degradation. For example, the modern automobile exhaust system is recognised as an environmental technology because it is designed with catalytic converters and other components to reduce engine emissions, whereas the engine is not recognised as an environmental technology because it is the emission source. This distinction, however, is misleading.

Nearly any technology can be configured in a manner which varies in the degree of its environmental effect. Thus, the degree to which nearly any technology degrades the environment can be increased. To the degree that it is configured to prevent potentially more extreme environmental degradation, therefore, nearly any technology qualifies as an ‘environmental technology’. Thus, like a modern exhaust system, a modern automobile engine is an environmental technology because it averts more extreme degradation, which would result from alternative, less efficient engines. Ultimately, any technology with a non-zero environmental effect parameter is an environmental technology.

Health risk assessment is an environmental technology. A second misleading connotation of ‘environmental technology’ is that, as a technology, it must be mechanical, as in the case of engines and exhaust systems. Health risk assessment, under this definition, would be disqualified as a technology, whether environmental or otherwise. This connotation is misleading. The hallmark of a technology is not that it is mechanical, but that it is technical. This is amply illustrated in the current information age, in which information technologies are typically classified as hardware vs. software. In this case study, health risk assessment is deemed to be a software example of environmental technologies.

The relationship of health risk assessment to mechanical environmental technologies is illustrated in Figure 1. In the figure, mechanical technologies are envisioned as a container of substances of potential concern, which undergo a process which directs some of the substances to a product and some to the environment. The difference is the controlled portion, often measured via an appropriately selected control efficiency parameter. This difference represents the mass of all substances of potential concern which could have been released to the environment, but was not. Health risk assessment is envisioned as an information technology, which should focus upon the process and the product(s) of the subject technology.

Technology transfer. Technology transfer denotes dissemination of a technology beyond the boundaries of its inventor, which may be a person or an institution. It came into existence with the onset of technologies, such as fire, stone tools and the wheel. Historically, technology transfer connotes the problem of an inventor precluding others from making gainful use of an invention at the expense of the inventor. In recent history, it connotes the national security problem of containing weaponry technologies. This problem became most urgent during the Cold War, when technologies for producing atomic bombs and later hydrogen bombs, were transferred by spies to the Soviet Union and beyond. Lately, however, technology transfer has come to connote a goal, as embodied in the Superfund Technology Demonstration Act and in ongoing efforts of the

426 R.A. Michaels

US Agency for International Development to transfer comparative risk assessment (‘comp risk’) technology to nations in the Caspian Sea region for use in prioritising options for developing oil and gas resources.

Figure 1 Environmental technology

product container of substances of potential << | _ environment concern | | : controlled health risk health risk assessment of assessment of process product

Environmental technology transfer synergism. Environmental technology transfer synergism here denotes the relationship of health risk assessment to other environmental technologies and, specifically, the effect of that relationship upon the prospects for technology transfer. This is illustrated by Figure 2, in which technology | is a process that yields product 1 and technology 2 is health risk assessment. Its most immediate ‘product’ is a health risk assessment report but, ultimately, its product is either successful technology transfer to a host community or rejection of the subject technology by the host community.

Health risk assessment has been criticised severely, sometimes justly. The scope of this case study excludes explication of such criticisms and defences. More important, the result of such criticism has been that community acceptance of proposed technologies has depended critically upon its acceptance of health risk assessment as well as its acceptance of the proposed technology which might be sited in the community. This interdependence constitutes synergism and is the focus of the present investigation of technology transfer synergism in the case of the proposed introduction of electronic chip fabrication technology to the town of North Greenbush, Rensselaer County, New York.

Environmental technology transfer synergism 427

Figure 2 Environmental technology transfer synergism

technology 1 |——______» product 1

technology 2 (HRA)

Electronic chip fabrication environmental impact statement (EIS). Corporations in the fast moving electronic chip fabrication (chip fab) industry must expect to receive permits to construct and operate their facilities following a permit application review process of only six to nine months. In New York State, permit review for major facilities (for example, the Inter-Power of New York coal-fired cogenerator proposed for siting on the Hudson River at the town of Halfmoon in Saratoga County) historically has occupied years rather than months. In the Inter-Power example, the permit was ultimately issued too late for the facility to be built (owing to expiration of the applicable power sale agreement with the local utility). Such events are nightmares to permit applicants and, in recent years, have tended to keep proposers of potentially controversial facilities out of New York State.

To overcome anticipated reticence on the part of potential chip fab applicants, an industry-academic consortium named the Center for Advanced Thin Film Technology was established at the State University of New York at Albany. Its purpose, most essentially, was to create the type of education, research and development (R&D) and labour infrastructure, which would attract chip fab facilities to the region. Approximately ten communities vied for primacy in the race to attract chip fab applicants. Empire State Development Authority funded preparation of a Draft Supplemental Generic Environmental Impact Statement (EIS) for chip fabrication facilities at the Rensselaer Technology Park in the town of North Greenbush. The EIS was termed ‘draft’ because it was not finalised. It was termed ‘supplemental’ because it was designed to supplement

428 R.A. Michaels

rather than act in place of an EIS prepared by the eventual chip fab applicant(s). It was termed ‘generic’ because the actual technology to be utilised by a potential applicant could not be divined ahead of an actual proposal being received. Hence the awkward acronym, DSGEIS, which will be rejected in favour of the more familiar EIS henceforth.

Evaluation of the ELS as a health risk assessment instrument. Chip fab supporters sought to shorten New York State’s lengthy permit process by identifying and resolving potentially controversial issues before a permit application was received. In contrast, opponents sought to use the EIS to generate environmental resistance and thereby thwart local plans to attract the chip fabrication industry. Opponents challenged chip fab based upon impacts reported in the EIS; such as traffic and increased water use, as well as undesirable development and health risks. Health risks most notably provided a potential statutory basis for rejecting chip fab facilities. Accordingly, the town of North Greenbush hired the RAM TRAC Corporation to evaluate objectively the adequacy of the chip fab EIS as a health risk assessment instrument.

RAM TRAC Corporation was engaged to perform three tasks:

1 critically evaluate documents filed in support of efforts to site such facilities in North Greenbush, most notably the EIS [1]

2 prepare a written report to the town of North Greenbush regarding the findings made and conclusions drawn

3 discuss the report at a Town Board meeting.

The tasks were directed toward achieving two objectives:

1 identifying public health risks which potentially may be posed to residents in and around North Greenbush

2 _ supporting preparation of the town’s Statement of Findings responding to filed documents.

The Statement of Findings represented an early opportunity for the town to request incorporation of safeguards into plans for siting chip fabrication facilities. To be effective, however, such requests would have to identify risks which the planned facilities might pose unless mitigated and also show how such risks would be mitigated if the requests were implemented.

Scope. The scope of this report is implied by its purpose: to identify risks to public health potentially associated with chip fabrication facilities which might be sited in North Greenbush. Consistent with a report by SEMATECH supporting the EIS [2], electronic chip fabrication:

“encompasses semiconductor process design and manufacturing operations

from the raw silicon wafer start to the packaged integrated circuit... [P]roduct

design, use of the product after manufacture and final disposition are not

included in this scope.” [2, p.3] Accordingly, the present report identifies public health issues of potential concern and evaluates the EIS and the adequacy with which it addresses the concerns identified in any of the stages of chip fabrication within the scope defined above.

The report excludes from its scope any determination of the completeness of the EIS with respect to New York State and Federal permit application requirements. A completeness assessment would identify required EIS components and evaluate the adequacy of each component with respect to proposed facilities. In contrast, a risk

Environmental technology transfer synergism 429

assessment identifies risks of potential concern and evaluates the adequacy with which each risk is addressed in the EIS, considering all components within the EIS where the identified risk(s) might be addressed. Further, in the absence of a specific chip fabrication proposal, potential risks only could be identified not quantified. The risk assessment is, therefore, qualitative not quantitative. Finally, except as necessary, this report does not repeat chip fabrication proposal details provided on a hypothetical basis in the EIS.

2 Methods

The principal method applied in this qualitative risk assessment is evaluation of available documents. Specifically, this risk assessment evaluates the EIS [1] that was published in print as well as electronically, at the town of North Greenbush web site <www.northgreenbush.net>. Further, members of the public have been invited to examine and evaluate the risk assessment and submit comments if interested. This risk assessment identifies and addresses issues of generic concern, which potentially may be associated with risks to public health that should be addressed in the EIS. In addition, it examines public comments submitted in writing, including electronic submittals via the internet and identifies issues of specific concern to members of the interested public as reflected in their submitted comments. Finally, this risk assessment identifies issues of concern expressed orally by members of the interested public attending a public hearing held by the town of North Greenbush on 4 March 1999. These comments have been recorded in a transcript of the meeting [3]. Following are examples of issues to be addressed on behalf of the town.

e Substances: identify substances of potential concern to be used at planned facilities and evaluate the appropriateness of product formulations to be purchased (such as liquid vs. pressurised gas) and the methods of containment and safeguards proposed to prevent releases.

e Transport: evaluate methods of transport of chemical products in and waste products out of facilities.

e Emergency response: evaluate plans to provide adequate information to emergency responders and adequate advance notice of plans to transport substances of potential concern.

e Sensitive subpopulations: evaluate plans to survey the community to identify sensitive subpopulations, such as children attending schools near proposed facilities and/or transport routes. To protect such sensitive subpopulations by maintaining adequate buffer distances and adopting risk management practices documenting protective measures to be taken routinely and in emergency circumstances.

e Training: evaluate plans to train and update training of employees. To educate community emergency response personnel and residents to assure implementation of promised safe practices.

e Maintenance: evaluate plans to maintain the facility to assure continued safety as equipment and containment structures age.

430 R.A. Michaels

¢ Monitoring: evaluate plans to monitor processes involving substances of potential concern and areas for handling and storing such substances. To assure early warning of releases and to prevent catastrophic releases of high volumes of such substances.

3 Findings 3.1 Identification and treatment of issues of potential concern

Generic issues of potential concern that were identified a priori are set forth in the Methods section above. Additional issues were sought in public comments submitted orally [3], in writing (for example, [4]) or electronically. Most issues identified from these sources were deemed to fit within the broader issues already set forth above. The additional issues, therefore, are addressed in appropriate sections and subsections below, within each generic issue. This approach was adopted with the intent to assure that specific issues, or specific aspects of issues, identified as being of potential concern to interested parties are addressed in this qualitative risk assessment.

3.2 Compatibility of chip fabrication proposal review with SEORA procedures

Some commenters expressed concern about possible incompatibility of the chip fabrication review process with procedures specified by the New York State Environmental Quality Review Act (SEQRA). As stated by a comment submitted in writing by allied environmental organisations:

“In the absence of an applicant, the town of North Greenbush is essentially

acting as the lead agency and the applicant, preparing the environmental analysis and reviewing it as well.”

Resolution of the issue awaits the appearance of an applicant and a permit application, at which time the SEQRA review presumably can proceed with all elements in place. Before project applications are received, however, local governments in New York State often request proposals for facilities, such as municipal waste combustors to cogenerate steam and electricity and then review applications submitted in response to such statements of interest. Similarly, with respect to chip fabrication, the only element that commenters appear to miss presently is the applicant. Inasmuch as the applicant may represent the strongest proponent of a proposed project, the temporary absence of the applicant does not lead logically to the adverse consequence, apparently envisioned by commenters, that perhaps project review under SEQRA might be prejudiced against project opponents.

The SEQRA review issue is addressed here and findings above are made, because the project review process potentially is relevant to the project public health risk outcome. The applicant presumably will be able to support the proposed project more strongly than the EIS can, because some data can be provided only by an applicant and only in connection with a specific project application. Such data were found, necessarily, to be missing from the EIS. Findings made regarding issues addressed below may include concerns about the absence or insufficiency of project-specific and/or applicant-specific data. However, such deficiencies cannot be expected to be remediated in the EIS until a project applicant provides the relevant data.

Environmental technology transfer synergism 431

3.3 Substances of potential concern to be used at planned facilities

Substances of potential concern that may be used in chip fabrication facilities should be identified in the EIS. Amounts expected to be used and amounts to be stored on-site should be quantified. The main body of the EIS [1] sets forth emissions to air without first identifying the commercial and/or experimental chemical products to be utilised at the facility, from which the emissions are assumed to arise. However, EIS [1] Table 8 estimates annual use of selected chemicals and fuels. A larger list is provided in EIS Appendix 3 [5] Table 2-1.

The EIS estimates annual usage of selected chemicals. However, electronic chips and their manufacture are rapidly evolving. Chips in the marketplace today are being manufactured in facilities already constructed whose capacity may be assumed to be capable of accommodating anticipated increases in demand. Existing facilities presumably can increase production to meet such demand via additions of equipment, staff and work shifts. Intrinsic to the prospect of attracting a new chip manufacturing facility, or more than one facility, to North Greenbush is the proximity of research facilities and talent, such as at Rensselaer Polytechnic Institute (RPI). These resources improve the prospect of commercialising newly emerging research findings via R&D. In short, a facility that eventually may be located in North Greenbush is unlikely to be identical to any facility already in existence and is unlikely to be described accurately by the EIS.

EIS Appendix 3 [5] includes a brief chip fabrication process description in the main body, supported by a more detailed description in Appendix A of Appendix 3 (an appendix in the appendix). However, these descriptions fail to identify many, if not the majority, of chemical components and commercially available chemical products utilised in chip manufacture, even in existing chip manufacturing facilities. Appendix A also is documented inadequately because it relies upon only four sources of technical information and the four sources are out of date (1984, 1989, 1994 and 1995). A complete description of technical developments in chip manufacturing is beyond the scope of this risk assessment. However, such developments are exemplified below via examination of articles that are recent (published in 1999). The articles were selected from the journal Science, which is widely valued for its selection of highly significant articles in its fields of coverage. Articles in the journal Science in branches of materials science relevant to chip manufacture would seem likely to presage significant advances for microchip technology evolution in the foreseeable future, as described in the paragraphs that follow.

TlBayCuOg1x, Laz,Sr¢CuO, and YBa,CuOg5 superconductors. High transition temperature (T.) superconductors have been discovered, but their high transition temperatures have not been explained theoretically, precluding reliable prediction of ways to increase the temperatures at which superconductors may function. Basov, ef al. 1999 [6] report advances in the theoretical explanation of high T, superconductivity in cuprates (copper based substances). This may enable manufacture of faster computers utilising superconducting chips, as the temperatures at which such computers must function begin to exceed the temperatures at which nitrogen and helium remain liquids at sea level pressures. The EIS, however, did not address the possible significance of high T, superconductor use in chip fabrication in North Greenbush.

432 R.A. Michaels

Superconductive GdBazCu3O;.. ultrathin films. Ultrathin films have been reported to switch between superconducting and insulating behaviour [7]. The authors say this switching represents the largest reversible modulation of superconductivity yet achieved using a field effect and the effect may be persistent because no voltage is maintained and therefore no voltage leakage can occur across the insulating ferroelectric field. The authors report that such transitions have applications in Josephson junction arrays on chips. Superconducting quantum interference devices rely upon small magnetic fields to modulate the critical current of a Josephson junction. The process involves a variety of potentially toxic substances: ferroelectric oxide lead zirconate titanate [Pb(Zr,TI,.,)O3] and super-conducting cuprate gadolinium barium copper oxide (GdBazCu;O;.,). Such substances, if not managed properly, might potentially pose risks to public health. The EIS, however, did not address the possibility, or possible significance, of using superconductive GdBa,Cu;0;., ultrathin films or the commercial chemical products needed to prepare them, should this technology be used in chip fabrication facilities in North Greenbush.

Bismuth oxide (Bi,O3) films epitaxially deposited on gold substrate. Microchips and other small devices can exploit intrinsic properties of their constituent materials or, less effectively, extrinsic properties arising from their grain boundaries. Switzer, ef al. [8] reported epitaxial electrodeposition of the high temperature polymorph of bismuth oxide from an aqueous solution onto single-crystal gold substrates. This electrodeposition process, the authors suggest, is a simple, inexpensive method for producing ceramics exploiting intrinsic material properties. This is essential for devices based upon superconductor, giant magnetoresistance, thermionic, piezoelectric and ferroelectric metal oxides. Products fabricated via electrodeposition epitaxy include non-equilibrium phases of materials that cannot be fabricated via traditional thermal processes because of thermally induced equilibration of the non-equilibrium phases. The EIS does not address the possible significance of electrodeposition of Bi,O; or other metal oxide films in chip fabrication, should this technology be proposed for use in a chip fab facility in North Greenbush.

Gadolinium oxide films on gallium arsenide (Gd)O3;/GaAs MOSFETs). Fabrication of apparently the first electrically and thermodynamically stable metal oxide semiconductor [gadolinium oxide-gallium arsenide (Gd,O;/ GaAs)] diode was reported by Hong, et al. [9]. The authors suggest that a material like this metal oxide semiconductor has has been sought for a long time for use as the substrate (base) for Field-Effect Transistors (FETs) using compounds of groups III and V of the periodic table of elements (MOSFETs). The Gd,0;/GaAs samples used to fabricate a GaAs MOSFET survived severe processing conditions, suggesting the feasibility of commercial MOSFET production via epitaxial growth of similar structures involving other rare earth oxides and semiconductor substrates, including silicon (Si). The EIS, however, did not address the possible significance of using such dielectric films for semiconductor MOSFETs in chip fabrication in North Greenbush.

Pentacene plastic-based Insulated Gate Field-Effect Transistors (IGFETs). FieldEffect Transistors (FETs) of the IGFET type (compare with FETs of the MOSFET type above) have been fabricated from organic semiconductors including conjugated polymers, oligomers (polymers with relatively few repeating monomer units) and other molecules. Dimetrakopoulos, et a/. [10] reported fabricating low voltage, high dielectric constant organic transistors on pentacene plastic-based IGFETs. The fabrication process was reported to have involved only room temperature processes, enabling demonstration

Environmental technology transfer synergism 433

of high-performance organic IGFETs on transparent plastic substrates, at low operating voltages compared with other organic devices. The EIS, however, did not address the possible significance of using organic semiconductor IGFETs in chip fabrication in North Greenbush.

Digital logic gates using Aluminium (Al) quantum-dots. A functioning logic gate was reported, based upon Al quantum-dots and Aluminium Oxide (AIO,) tunnel junctions [11]. This device may be significant in allowing integrated circuit fabrication technologies to continue the trend toward increasing the density of logic gates on chips as smaller gate sizes and increased densities approach quantum limits. The logic gate reported by AmLani, ef al. departs from traditional FETs that have formed the basis for complementary metal oxide semiconductors. Quantum-dot cellular automata (QCA) encode binary information via arrangements of individual electrons rather than via currents and voltages. The EIS, however, did not address the possibility, or potential public health significance, of this technology being commercialised in a chip fabrication facility in North Greenbush, especially with respect to identifying and quantifying potentially toxic chemical products which might be used and stored on site to fabricate such devices.

Platinum-triggering of remote waves in an electrochemical system. Formic acid was reported to be oxidised on a platinum wire, perturbing a passive (OH-poisoned, lowcurrent) state into an active (high-current) state [12]. The active state was observed to spread, as a wave front, switching the entire system from the passive to the active state (low-current to high-current state). The propagation rate included a non-local component, apparently producing ‘instantaneous’ coupling (inducing the low-current or high-current transition in distant points within 10° second) due to spreading of an electric field across the system at light speed. This non-local high-speed coupling in systems otherwise exhibiting slow local dynamics (such as ion migration) might be exploited for switching between on-off states in future chip technologies. The EIS, however, did not address the possibility and potential public health significance, of this technology being applied at a chip fabrication facility in North Greenbush.

‘Dip pen’ nanolithography improves on electron-beam lithography. Lithography as described in the EIS is intrinsic to microfabrication, nanotechnology and molecular electronics; but lithography is evolving. Piner, ef a/. [13] reported development of a direct-write ‘dip-pen’ nanolithography (DPN) capable of delivering collections of molecules from an atomic force microscope to a solid substrate of interest. The authors reported writing alkanethiols (an electronic chip version of ink) with high resolution onto a thin gold film (an electronic chip version of paper). DPN delivers molecules to a substrate via capillary transport. The authors suggest that DPN may represent an improvement over more conventional lithographic methods for detailed ‘functionalisation’ of nanoscale devices. The EIS, however, did not address the possibility, or potential public health significance, of alkanethiol use in lithography for chip fabrication in North Greenbush.

Photo-assisted electrochemical etching in gallium phosphide (GaP). Development of a photo-assisted electrochemical etching technique was reported by Schuurmans, ef al. [14]. Use of the technique was reported for fabricating a completely connected network of macropores in single crystalline GaP with variable porosity. The technique produces a strongly light-scattering (photonic) material with periodic spatial variation of its refractive index — a photonic crystal. The authors observed a scattering layer (slab)

434 R.A. Michaels

exhibiting favourable optical transparency and scattering properties. They also reported success in using the etching technique to control the thickness of the scattering slab. The authors reported that the technique allows preparation of strongly photonic systems with negligible absorption in the red region of the visible spectrum. The EIS, however, did not address the possibility, or potential public health significance, of using GaP for etching microchips, should this technology be proposed for a chip fab facility in North Greenbush.

Optical computing via titanium (Ti) and GaAs photonic memory cells. Fabrication of a semiconductor-based photonic memory cell was reported by Zimmerman, ef al. [15]. It consisted of a GaAs-based quantum well (QW) that was grown via molecular beam epitaxy and sandwiched between AlAs-GaAs superlattice barriers on which a thin film of titanium was deposited. The resulting device stored light for a period sufficiently long for optical signal processing applications. Storing light as an array of pixels on a chip is the optical computing equivalent of storing electrons in a capacitor in electronic computing and memory devices. The authors suggested that their technology can be used to fabricate optical dynamic random access memory (DRAM) analogous to electronic DRAM used in present day computers. The EIS, however, did not address the possibility, or the potential public health significance, of applying optical DRAM and optical chip fabrication technologies in North Greenbush.

Organic light-emitting devices based upon tris (8-hydroxyquinoline). Chip fabrication is linked to ancillary technologies, such as computer flat panel displays. Organic lightemitting devices (OLEDs) have high brightness, a wide range of available colours and low cost, but are unstable [16]. However, approaches have been applied successfully to stabilise OLEDs. The EIS, however, did not address such technologies as potential secondary industries, which might co-locate with chip fabrication facilities in or near the Rensselaer Technology Park.

Phase-separated composite films (PSCOFs) for liquid crystal displays. An emerging technology also ancillary to chip fabrication is PSCOF [17]. The authors report that PSCOF offers advantages including ease of fabrication, mechanical ruggedness, high resistivity and flexibility of fine-tuning the optical path length. PSCOF, the authors also reported, appears to enable commercial fabrication of electro-optical devices with ferroelectric and antiferroelectric liquid crystals. Such devices involve use of polymers, such as polyimides (polymers with -NH(-C=O), monomer units), for configuring the liquid crystal optical axis as needed. The EIS, however, did not address such technologies as potential secondary industries which might colocate with chip fabrication facilities in or near the Rensselaer Technology Park.

Lithium gadolinium fluorine europium (LiGdF,:Eu*') display panels. Alternative luminescent materials have been sought for a long time for use in plasma display panels and other applications. Wegh, et a/. 1999 [18] report synthesis of polycrystalline powder samples of Eus.-doped LiGdF, (LiGdF,:Eu’’) for use in display panel phosphors. They demonstrated down-conversion of ultraviolet photons to twice as many visible photons (200% quantum efficiency) in the orange and red spectral range. The authors opine that this achievement represents an important step toward developing more efficient plasma display panels, triggering the search for complementary down-converting phosphors exhibiting quantum efficiencies exceeding 100% in the blue and green spectral regions. The EIS, however, did not address such technologies as potential secondary industries which might colocate with chip fabrication facilities in or near the Rensselaer Technology Park.

Environmental technology transfer synergism 435

Substances of potential concern in electronic chip fabrication which reasonably should be addressed in the EIS based upon the above review of literature on emerging technologies are set forth in Table 1. As Table 1 suggests, multiple emerging technologies utilise a common list of substances. Although substances in Table | represent only a sample of emerging technologies, consideration of additional emerging technologies is likely to add substances of potential concern for EIS consideration far more slowly than emerging literature will add technologies for EIS consideration. Thus, the primary onus for EIS augmentation is a literature review task, whereas the explication of the public health implications of additional substances associated with emerging technologies will constitute only a secondary onus for EIS preparation.

Table 1 Substances of potential concern in electronic chip fabrication Substance Formula NYS Air Technology innovation High speed chip or Source Guide-1 other device Jeature(s) alkane-thiols = Hyn+1)C,-SH Ree ‘Dip pen’ using etched 13 nanolithography nanoscale features aluminium Al cod quantum-dot logic gates using quantum-dot 11 with aluminium oxide cellular automata tunnel junctions barium Ba moderate high transition superconducting 6 toxicity temperature superconductors barium Ba moderate Josephson junctions superconducting Uf toxicity using superconducting ultrathin films bismuth Bi a epitaxial applying metal 8 electrodeposition of oxide Bi,O; onto gold film semiconductors to ceramics or other materials coppér Cu moderate high transition superconducting 6 toxicity temperature superconductors copper Cu moderate Josephson junctions superconducting 7

toxicity using superconducting ultrathin films

europium Eu ap UV photon conversion computer plasma 18 to twice as many visible _display panels photons (200% quantum efficiency)

fluorine F moderate UV photon conversion computer plasma 18

toxicity to twice as many visible _ display panels photons (200% quantum

efficiency) formic acid HCOOH moderate _ on-off switching via exploiting non- 12 toxicity non-local high-speed locality principle in coupling in slow ion quantum mechanics

migration systems

436 R.A. Michaels

Table 1 Substances of potential concern in electronic chip fabrication (continued) eee Substance Formula NYS Air Technology innovation High speed chip or Source

Guide-1 other device feature(s) gadolinium Gd eee UV photon conversion computer plasma 18

to twice as many visible display panels photons (200% quantum efficiency)

gadolinium Gd aS Josephson junctions superconducting 7 using superconducting ultrathin films

gadolinium Gd op first electrically and using MOSFETs 9 thermodynamically stable metal oxide semiconductor field effect transistors (MOSFETs)

gallium Ga op first electrically and using MOSFETs g) thermodynamically stable metal oxide semiconductor field effect transistors

(MOSFETs) gallium GaAs fs semiconductor-based using optical 15 arsenide photonic memory cell dynamic random using quantum well access memory configuration (DRAM) gallium GaP aa photo-assisted using photonic 14 phosphide electrochemical etching _ etching of microfeatures gold Au a epitaxial applying metal 8 electrodeposition of oxide Bi,O; onto gold film semiconductors to ceramics or other materials gold Au Be ‘Dip pen’ using etched 13 nanolithography nanoscale features lanthanum La Ba high transition superconducting 6 temperature superconductors lead Pb criteria Josephson junctions superconducting 7 pollutant using superconducting ultrathin films lithium Li cc UV photon conversion computer plasma 18

to twice as many visible _ display panels photons (200% quantum efficiency)

pentacene CoHi4 2) high-performance using IGFETs 9 organic insulated-gate field effect transistors (IGFETs)

Environmental technology transfer synergism 437 Table 1 Substances of potential concer in electronic chip fabrication (continued) Substance Formula NYS Air Technology innovation High speed chip or Source

Guide-I other device feature(s) platinum Pt on-off switching via exploiting non- 12 non-local high-speed locality principle in coupling in slow ion quantum mechanics migration systems polyimides [(O=C),NH], phase-separated computer liquid 17 composite films cystal displays (PSCOFs) for liquid using PSCOFs crystal displays strontium Sr high transition superconducting 6 temperature superconductors thallium Tl moderate _ high transition superconducting 6 toxicity temperature superconductors titanium Ti Josephson junctions superconducting 7 using superconducting ultrathin films titanium Ti semiconductor-based using optical 15 photonic memory cell dynamic random using quantum well access memory configuration (DRAM) tris 8-hydroxy CyH;NO stabilisation of organic computer flat panel 16 quinoline light-emitting devices displays using (OLEDs) OLEDs yttrium nYs high transition superconducting 6 temperature superconductors zirconium Zr Josephson junctions superconducting 7

using superconducting ultrathin films

3.4. Chemical product formulations and transport

Ammonia (NH3) may be used at chip fabrication facilities which might be sited in North Greenbush [5, pp.4-4]. It may be obtained in a variety of commercial formulations; including; liquid Spirit of Hartshorn (28-29% ammonia), compressed gas, cryogenic liquid and as powdered or granular urea. Concentrated sources of ammonia are highly toxic, whereas pressurised sources may pose explosivity and cryogenic risks in addition to toxicological risks [19]. The EIS, however, addresses neither the formulation(s) of ammonia nor the possible public health and safety significance of the selection of formulation(s) of ammonia, which might be transported to and utilised at chip fabrication

facilities in North Greenbush.

438 R.A. Michaels

3.5 Air emissions

Criteria pollutants. EIS Table | estimates potential emissions to the atmosphere of criteria pollutants, including particulate matter within a 10-micron diameter (PMyo), sulphur dioxide (SO), nitrogen oxides (NO,), carbon monoxide (CO) and volatile organic compounds (VOCs). However, the National Ambient Air Quality Standard (NAAQS) for PM has been revised by the US EPA to include PM within 2.5 microns (PM2s) comprising the smaller particles within PMj) most directly associated with adverse health effects [20-22]. The EIS does not, however, estimate these more toxicologically significant PM>; emissions.

Chip fabrication facilities described in the EIS are assumed to be heated by gas fired boilers [5, p.3-3]. The boilers are assumed to utilise natural gas for 48 of 52 weeks per year and distillate (No. 2) fuel oil as a back-up fuel for four weeks per year in winter. Indeed, residential heating needs enjoy regulatory priority over commercial needs for natural gas. Consequently, commercial boilers must be equipped to burn an alternative fuel during winter months when natural gas might be in short supply. However, the EIS assumes that natural gas might be unavailable for only four weeks in a 12 week winter. Although this might turn out to be accurate, it does not constitute a credible worst-case assumption for the north-east region. A more credible worst-case assumption would be use of the distillate oil back-up fuel for 12 weeks (2,000 hours) per year. Finally, the availability of natural gas depends also upon its price. If it is available, but only at a price well in excess of the price of distillate No. 2 fuel oil, a natural gas facility reasonably can be expected to petition for use of the less expensive back-up fuel during the shortage when prices are excessive. Indeed, in some economic circumstances, permit modifications may be requested to allow utilisation of fuel oil all year, if it is cheaper than natural gas all year.

Hazardous air pollutants (HAPs). HAPs may be produced by a variety of sources in a chip fabrication facility. Indeed, natural gas-fired boilers addressed above can emit HAPs when powered by the back-up fuel. Ironically, this is true because gas plants are extremely clean when fuelled by natural gas. They typically are built with low stacks, high stack flow rates and no pollution control equipment in the stack. When such plants are fuelled by oil they may emit trace pollutants present in the oil, such as arsenic, cadmium and nickel, often introduced in refineries during crude oil cracking.

The issue of whether HAPs might be produced from technologies directly involved in chip fabrication is addressed inadequately. Numerous existing and emerging technologies are not addressed in the EIS (exemplified earlier in Section 3.3). Each of the exemplified technologies is associated with the use of potentially toxic substances including metals and compounds of metals such as; aluminium, arsenic, barium, bismuth, copper, europium, gadolinium, gallium, lanthanum, lead, platinum, strontium, thallium, titanium, yttrium and zirconium. Some of these must be used in accordance with industry consensus standards, such as those for titanium [23] and zirconium [24] promulgated by the National Fire Protection Association (NFPA). The EIS, however, did not estimate amounts of these substances which might be used, nor describe how and with what efficiency they might be contained. The EIS also fails to estimate emission rates, or provide assurances that emission rates will be quantitatively insignificant or that they will pose only de minimis risk to public health.

Extremely hazardous substances (EHSs). In 1984 methyl isocyanate was released from a chemical plant in Bhopal, India [19,25]. The release was reported to have killed

Environmental technology transfer synergism 439

thousands of residents and irreversibly injured the eyes and lungs of approximately 20,000 [25]. Governments recognised the need to identify EHSs and assist communities to prepare for chemical emergencies. The EPA identified 366 EHSs, as required by the Community Right-To-Know Act [26]. Community emergency planning requirements followed under the Risk Management Program (RMP), in Section 112(r) of the 1990 Clean Air Act Amendments [27]. EHSs, including ammonia, may be used in chip fabrication facilities [5, p.4-4]. The issue of EHSs is addressed further in Section 3.7, below.

3.6 Adequacy of health-based standards and guidelines

Boiler emission standards. The EIS indicates that standards are unavailable for sulphur dioxide (SO) and particulate emissions from natural gas fired boilers [5, pp.4—6]. The absence of these SO, and particulate emission standards might suggest that health risks posed by natural gas fired boilers emitting these substances typically are insignificant. However, this suggestion cannot be applied to natural gas fired boilers during operations with back-up No. 2 fuel oil. This operating mode may have little, if any, public health significance. Whatever public health significance it has may be understated, however, because of the short duration of No. 2 fuel oil use assumed in the EIS.

Chip fabrication emission standards and guidelines. The EIS describes chip fabrication technologies based upon only four technical literature sources [5, Appendix A of Appendix 3], dated 1984, 1989, 1994 and 1995. Today these sources are obsolete or obsolescent relative to the rapid pace of innovation in this industry. The sources must be deemed to be even further out of date relative to emerging chip fabrication technologies which might be adopted at facilities eventually to be sited in North Greenbush. Emerging technologies more relevant to North Greenbush were exemplified earlier, in Section 3.3. Each of these examples identifies substances that might be purchased for use in chip fabrication. Most of these substances are not identified in the process description portions of the EIS. Consequently, the EIS also fails to quantify the amounts of the materials which might be transported to each facility and also fails to evaluate the fraction of each substance that might be emitted from chip fabrication processes. In addition to incompleteness of data on substances which might be emitted from modern chip fabrication facilities, the majority of the substances also appear to lack emission standards and guidelines to protect public health.

3.7 Emergency response planning

Industrial facilities exceeding threshold quantities of EHSs on site are subject to the US EPA’s Risk Management Program (RMP; [19,25]). The RMP defines Emergency Planning Zones around facilities and requires facilities to submit to the US EPA a risk management plan aimed at preventing catastrophic chemical releases and averting Bhopal-type consequences of such releases, should they occur. The cost of community emergency planning for a facility is likely to depend roughly upon the area of its Emergency Planning Zone, which in turn depends upon the square of the radius around the facility deemed necessary to attenuate a worst-case chemical release to specified airborne concentrations, termed Acute Exposure Guideline Levels (AEGLs).

440 R.A. Michaels

AEGLs for cach EHS have nothing to do with routinely acceptable exposures. Acceptable community and occupational levels are set by a host of other, more conservative parameters. Rather, AEGLs must be established to protect members of the general population in the context of “exposure at high levels. but of short duration, usually less than one hour and only once in a lifetime” [25]. AEGLs are under development by the National Advisory Committee (NAC) on AEGLs and NAC AEGL’s composition is legislatively defined to assure balance. One concern potentially is failure to achieve balance within NAC AEGL with respect to its review of specific EHSs [19], such as ammonia.

Three types of AEGL are defined [25]:

1 AEGL-1, to protect against nuisance exposure

2 AEGL-2, against irreversible injury or disability (including impairment of the ability to escape)

3 AEGL-3, against lethality.

Each type of AEGL is quantified via multiple combinations of airborne concentration and exposure duration (five minutes, 30 minutes, one hour, four hours and eight hours) to protect against effects of concern in exposures lasting from minutes to hours.

In the context of regulatory requirements for emergency planning, the EIS reasonably may be expected to address basic emergency planning issues including the following:

¢ _EHSs: identify extremely hazardous substances (EHSs) which might be used at chip fabrication facilities

e threshold EHS quantities: report threshold quantities which, if stored and/or used onsite, will trigger emergency planning requirements

© compliance with emergency planning requirements: describe how chip fabrication facilities might comply with emergency planning requirements, such as sheltering people in place or evacuating people under emergency circumstances

© — identifying emergency response resources: describe how chip fabrication facilities will identify local emergency response resources and train them to respond or participate in responding to selected emergency scenarios

¢ alerting emergency responders: describe procedures, if any, to alert emergency responders by providing adequate advance notice of plans to transport substances of potential concern.

The EIS addresses emergency planning in Vol. 1, Chap. 5, Section 5.4, titled Emergency Services and Hazardous Materials, on pages 5-18 to 5-27 [1]. EHSs are identified in Table 8, but Table 8 is based upon only a brief process description that is out of date, as documented earlier. Table 8 estimates annual consumption of EHSs but fails to quantify amounts which might be stored and used on-site at any given time. The EIS, therefore, can neither estimate emergency planning zone sizes nor map their location for the benefit of the host community, whose property values conceivably might be impacted by inclusion in an EPZ, The EIS acknowledges the need to comply with the EPA’s Risk

Environmental technology transfer synergism 441

Management Program (RMP). However, the treatment of this issue is inadequately detailed with respect to the issues identified above. The EIS does, however, identify emergency response resources in detail.

3.8 Sensitive subpopulations

Sensitive subpopulations may include pregnant women, infants, children, the elderly and those suffering from predisposing conditions, which might include pulmonary and cardiopulmonary illnesses such as asthma, chronic obstructive pulmonary disease (COPD), congestive heart failure, emphysema and reactive airway disease syndrome (RADS). Specific subpopulations which are more sensitive than average members of the population can be identified relative to specific substances of potential concern in chip fabrication. Places where concentrations of sensitive subpopulations are located may include assisted living residences, day care centres, hospices, hospitals, nursing homes and schools.

The issue of sensitive subpopulations primarily pertains to emissions to the atmosphere. Sources may include stack emissions, cooling tower emissions, ‘fugitive’ emissions (which by-pass pollution control equipment) and emissions from vehicles transporting substances to, and wastes from, chip fabrication facilities. The EIS reasonably may be expected to model chip fabrication facility emissions to the atmosphere, at least at a screening level of detail, to determine whether or not air impacts will coincide with sensitive subpopulation locations, such as the types of facilities listed above.

The EIS addresses emissions to air in Chap. 5 [1], which is supported by Appendix 3, titled Technical Support Documentation: Air Permit Application, Semiconductor Manufacturing Facility [5]. Chap. 5, however, completely fails to address sensitive subpopulations. Examination of Appendix 3 reveals the presence of detailed estimates of facility emissions based upon the assumed chip fabrication process description. Within Appendix 3 is a supporting appendix, titled Appendix D, Air Toxics Analysis.

Appendix D applies the New York State Air Guide-1 (AG-1) review process for new sources of air pollutants [28]. The appendix presents an air quality analysis based upon use of the AG-1 air dispersion computer model in conjunction with US EPA air dispersion models ISCLT2 and SCREEN3. Results are presented with respect to substances listed in Air Guide-1 whose emission rates from proposed chip fabrication facilities were quantified. Short-term and long-term impacts (airborne concentrations) of the substances are tabulated and compared with respective AG-1 acceptable limits, termed short-term guideline concentrations (SGCs) and annual guideline concentrations (AGCs). The following emissions failed the initial screening analysis, either by exceeding applicable AGCs or SGCs (indicated below via ), or by exceeding the de minimis screening criterion of | j1g/M’ (indicated below via *):

* carbon monoxide,

* hydrofluoric acid,

* nitrogen oxides,

* particulate matter,

* — sulphur dioxide,

** 2-(2-aminoethoxy)ethanol,

442 R.A. Michaels

** ethyl lactate,

** ethyl pyruvate,

** hexamethyldisilazane,

** hydroxylamine

** tetramethylammonium hydroxide.

These emissions were then subjected to refined modelling using SCREEN3 for shortterm impacts and ISCLT2 for long-term (annual) impacts. Refined modelling resulted in impacts within respective AGC and SGC limits (Tables D-8 and D-9).

Notwithstanding the acceptability of all impacts relative to standards and guidelines set forth in Air Guide-1, the air dispersion results are not mapped spatially relative to locations in and around North Greenbush. No inventory of potentially sensitive subpopulations was developed and air impacts are not related spatially to the locations of potentially sensitive subpopulations. Consequently, the acceptability of the air impacts relative to sensitive populations is entirely conditional upon the adequacy of the standards and guidelines to protect sensitive subpopulations. However, this issue appears to have been omitted from consideration in the EIS.

A final issue, also not addressed in the EIS, is the degree to which simultaneous and/or sequential exposure to multiple pollutant emissions requires adherence to more stringent standards and guidelines than those presented in Air Guide-1, which states the following:

“DEC has not developed guidelines for assessing the toxicity of mixtures of air contaminants. It is largely unknown if mixtures of air contaminants will induce additive, synergistic, potentiating or antagonistic effects. However, in the absence of information to the contrary, the methodology proposed by the American Conference of Governmental Industrial Hygienists (ACGIH) should be used (refer to ACGIH Appendix C; 1) and limitations to this approach as described therein should be acknowledged.” [28, p.C-7]

This health risk assessment methodology is not applied and not mentioned, in the EIS.

3.9 Plans to educate employees, residents and emergency responders

The EIS reasonably should be expected to address plans to educate chip fabrication facility employees, community residents and emergency responders about hazardous materials stored and used on-site and appropriate responses to emergencies involving these substances. This issue area should be addressed in connection with emergency planning, because statutory and regulatory requirements for such training activities arise from emergency planning requirements (see Section 3.7). Accordingly, the EIS should address the training issues listed above in Section 5.4, titled Emergency Services/Hazardous Materials. However, examination of EIS Section 5.4 reveals little information about the training issue area. Employees. The EIS states the following:

“Large facilities that use bulk chemicals normally have their own fire brigade/hazardous materials response team. This team is equipped and trained to control accidental releases of hazardous materials within the facility and assists local emergency services providers should their assistance be needed.” [1, pp.5—20]

Environmental technology transfer synergism 443

Training of employees other than emergency responders is not addressed. However, supporting appendices on environmental health and safety [29] and the SEMATECH report [2] clearly suggest awareness of the need to train employees to prevent releases, rather than just emergency responders to remediate releases.

Emergency responders. The EIS states the following with respect to educating emergency responders:

“In the event that local fire departments must respond to an accidental release at the facility, the facility should pre-plan such a response with local emergency services providers (including EMS and police services as well) and provide any necessary training and specialised equipment to these agencies. The facility should participate with emergency services agencies in pre-planning for emergencies, as well as drills and exercises.” [1, pp.5—21]

Specific training resources, intensities, programs, objectives and schedules are not identified.

Community residents. The EIS states the following:

“Either an on-site or transportation accident may result in an airborne hazard to the general public. Such an event will require procedures for notification of the general public as to the steps to be taken to prevent adverse health impacts (such as evacuation or in-place sheltering). The facility should assist in public notification through support for local warning systems and through public outreach programs (examples of such systems are rapid telephone dialers and tone alert receivers).” [1, pp.5—23]

3.10 Plans for facility maintenance

Examination of Appendix 10 reveals that it is out of date, as documented earlier. Further, it draws upon a meagre base of technical support, as also documented earlier. Finally, it includes no section(s) devoted to facility maintenance.

Examination of Appendix 11 (DFESH) reveals a document which includes no section(s) devoted to facility maintenance. DFESH embraces trendy concepts such as total quality management, which it embellishes fashionably as Total Quality ESH Management, with the acronym TQESHM. The document primarily is devoted to promoting this TQESHM approach, for example, by identifying where people can obtain training to become proficient in implementing it (WPCOTTBPIII) in their own semiconductor facilities.

Although the TQESHM management approach can and should result in benefits with respect to all aspects of chip fabrication facility operations, including maintenance, it

444 R.A. Michaels

does not include within its scope the specific maintenance objectives which will be adopted by a proposed facility to assure residents in the host community that the facility will remain safe as it ages. For example, the SEMATECH document assures community residents that training of personnel in TQESHM is commercially available and should be updated periodically, but it does not assure residents that bullet tanks containing pressurised anhydrous ammonia will be inspected and with what frequency and whether the inspection method(s) will be invasive or not.

3.11 Plans to monitor processes involving hazardous chemical products

The issue of monitoring is closely related to that of maintenance, addressed above. Monitoring can be remote or on-site, invasive or non-invasive, reactive or proactive. The issues of inspection, record keeping and public access to records are all intrinsic to the issues of monitoring and maintenance. With respect to monitoring, the issues of inspection, record keeping, record storage and retrieval and public access to records appear to be completely overlooked in the EIS. Yet these issues constitute the principal bases for enforcement of permit conditions and for assuring that chip fabrication facilities will be held responsible for violations of public safety and public trust. Secretive industrial culture, though important for commercial success, no longer is acceptable where public health and safety are involved. The EIS fails to address this issue, though the SEMATECH report [2] supporting the EIS [1] clearly reflects the cultural evolution of the semiconductor industry in this direction.

4 Conclusions and discussion

Approximately ten communities vied for primacy in the race to attract chip fab applicants to a preferred site. Empire State Development Authority funded preparation of an EIS for chip fabrication facilities at the Rensselaer Technology Park in the town of North Greenbush. Chip fab supporters hoped to shorten New York State’s lengthy permit process to within nine months, deemed acceptable to potential chip fab facility applicants. In contrast, opponents hoped to use the EIS to generate environmental resistance and thereby thwart local plans to attract the chip fabrication industry.

The outcome of this age-old conflict between those favouring development and those opposing it was community rejection of the plan to attract chip fab facilities. This negative outcome was determined by a complex interplay of factors. These most notably included the degree of resident suspicion about the chip fab industry and the receptiveness of residents to health risk assessment findings and conclusions about that industry. Here, this interplay has been termed technology transfer synergism. Reasons why this synergism caused project extinction rather than enhanced prospects for survival cannot be known fully. A significant factor, however, was failure to arrange for public presentation of the risk assessment, with opportunities for concerned people to express their concerns and hear responses in real time, in person. Thus, the risk assessment process was never fully engaged.

Chip fab supporters cited the conclusion that “no specific risks were identified which cannot be mitigated acceptably if addressed,” whereas opponents exploited the conclusion that “the EIS, however, was found deficient with respect to its consideration

Environmental technology transfer synergism 445

of critical risk related issues...” Most notably, the EIS had failed to address many substances associated with emerging chip fabrication technologies. These principal conclusions are set forth below.

No specific risks were identified which cannot be mitigated acceptably if addressed. The EIS, however, was found deficient with respect to its consideration of critical risk related issues. It failed to identify many, if not most, emissions of potential concern, especially from emerging chip fabrication technologies likely to be relevant for North Greenbush. It failed to address formulations of hazardous chemical products, such as ammonia, to be used. Air emissions were analysed in detail but the list of emissions was incomplete and boiler emissions were estimated based upon assuming only 672 hours of back-up fuel use, rather than a more reasonably conservative value such as 2,000 hours. Although air emissions were found to be within applicable standards and guidelines for individual substances, the EIS failed to evaluate the adequacy of these standards and guidelines for regulating an industry new to New York State. It also failed to adjust the standards and guidelines as required to account for simultaneous exposure of people to multiple emissions. It failed to identify potentially sensitive subpopulations of people in the project area and to map air impacts relative to the locations of such populations; which might include assisted living residences, day care centres, hospices, hospitals, nursing homes and schools.

The EIS was found to address several issues likely to be of public concern in only rudimentary fashion, precluding early resolution of the issues needed to expedite eventual permit review. These issues included emergency planning, plans to educate employees, residents and emergency responders, plans for facility maintenance to assure continued safety with facility age, and plans to monitor facility operations.

4.1 Substances of potential concern to be used at planned facilities

The Findings section indicates that the EIS failed to identify many if not most chemical components and commercially available chemical products utilised in chip manufacture. This is partly attributable to the paucity of data considered regarding existing chip manufacturing facilities and partly because the source documents forming the technical basis for the EIS were out of date. Substances of potential concern associated with emerging chip fabrication technologies are exemplified in the Findings section and summarised below in Table 2. Based upon these findings, I conclude that overcoming resistance to permitting chip fabrication facilities requires consideration of a full range of substances of potential concern. These would include substances whose quantitative consideration ultimately will reveal their insignificance to public health. Consideration of substance of potential concern in a health risk assessment is necessary, in the presence of risks to public health, to inform the public of such risks. Consideration of substances of potential concern also is necessary, in the absence of risks to public health, to allay public fears. Thus, the conclusion that risks are not of concern is most appropriately an output of the risk assessment, not an input.

446 R.A. Michaels

Table 2 Substances of potential concern in electronic chip fabrication Substance Formula Data source(s) Organics alkane-thiols Hen+1)Cy-SH 13 formic acid HCOOH 12 pentacene CyHis 9 polyimides [(O=C),NH], 17 tris 8-hydroxy-quinoline CsH;NO 16 Metals (and compounds) aluminium Al 11 arsenic As 15 barium Ba 6,7 bismuth Bi 8 copper Cu 6,7 europium Eu 18 fluorine F 18 gadolinium Gd 7,9, 18 gallium Ga 9, 14, 15 gold Au 8, 13 lanthanum La 6 lead Pb 7 lithium Li 18 platinum Pt 12 strontium Sr 6 thallium Tl 6 titanium Ti i yttrium ry. 6 zirconium Zr t

4.2 Potential remedies for adequate consideration of substances

Deficiencies with respect to identifying and addressing substances of potential concern can be addressed in either of two ways. One is to await a permit applicant and an application to site a facility which will utilise a specified technology. In that case, the entire onus of identifying and addressing substances of potential concern will rest upon the EIS submitted by the applicant. Although this expanded onus might not prolong the permit preparation process unmanageably for a project applicant, it will prolong the permit review process, possibly unacceptably for the applicant.

A preferable approach is to place more of the onus of identifying and addressing substances of potential concern on the EIS. The EIS can perform either of two tasks. It can survey the scope of existing and emerging chip fabrication technologies and narrowly address only those which will be eligible for permitting. Alternatively, the EIS can exemplify a broad range of technologies, reducing the onus on applicants whose technologies are exemplified in the EIS. A broader EIS survey of the field of chip

Environmental technology transfer synergism 447

fabrication technology evolution can be expected to appeal to a wider variety of applicants.

4.3 Chemical product formulations and transport

As indicated in the Findings section, ammonia may be used at chip fabrication facilities that might be sited in North Greenbush. Ammonia is highly toxic and, indeed, it is classified as an EHS. The EIS, however, failed to describe uses of ammonia at chip fabrication facilities. It also failed to identify commercially available formulation(s) of ammonia. Finally, it failed to consider the possible public health and safety significance of the selection of formulation(s) of ammonia that might be transported to, or utilised at, chip fabrication facilities in North Greenbush. These deficiencies are especially significant in the case of ammonia because it is often utilised in high volume applications, such as commercial refrigeration and has a history of being involved in transport accidents [30-34]. To remediate this deficiency the EIS should describe ammonia uses in chip fabrication facilities, identify usable commercial formulations and address the possible public health and safety significance of the selection of ammonia formulation(s) for use at chip fabrication facilities.

4.4 Air emissions

The EIS analysed air toxics emissions via New York State’s Air Guide 1 [28]. The analysis purported to demonstrate acceptability of air toxics emissions. However, two issues which were found to be addressed inadequately (see Section 3) render this conclusion premature. First, boiler emissions were estimated based upon use of back-up No. 2 fuel oil for only four weeks per year, whereas 2,000 hours per year represents a more reasonable assumption in the present context. Air toxics emission rates are higher during use of back-up No. 2 fuel oil. Second, numerous substances not addressed in the EIS are involved in emerging technologies for chip fabrication and some of these may reasonably be expected to be involved in chip fabrication facilities if sited in North Greenbush, These are exemplified in Table 2.

4.5 Adequacy of health-based standards and guidelines

The EIS indicated that standards are unavailable for sulphur dioxide (SO») and particulate emissions from natural gas fired boilers [5, p.4-6]. Substances of potential concern associated with emerging technologies also appear to lack air pollutant emission standards and guidelines for their control. Emerging chip fabrication technologies relevant to North Greenbush were exemplified earlier in Section 3.3. Each of these technology examples identifies substances that might be purchased for use in chip fabrication. These substances and the emerging technologies with which they are associated, were set forth above in Tables 1 and 2. Table 1 also indicates the qualitative toxic potency evaluation (high, moderate, or low toxicity) of the substances in New York State’s Air Guide | [28]. Air Guide 1 formed the basis for the EIS demonstration of acceptability of chip fabrication facility emissions in North Greenbush relative to standards and guidelines applicable in New York State. However, as Table 1 indicates, 19 of 25 substances identified in association with emerging chip fabrication technologies

448 R.A. Michaels

are not listed in Air Guide 1 and apparently lack specific standards and guidelines applicable to them.

Based upon the above considerations, I conclude that, although the EIS evaluated the acceptability of chip fabrication facility air emissions relative to standards and guidelines applicable in New York State, its credibility was undermined by failure to evaluate the adequacy of the standards and guidelines to regulate an industry that is relatively new to New York State. Accordingly, Federal and New York State air pollution standards and guidelines should be evaluated relative to their adequacy to protect public health in the project vicinity. A recommended approach is for the EIS to evaluate California’s procedures for regulating the chip fabrication industry in Silicon Valley and compare New York State and California to determine whether New York State air pollution standards and guidelines must be augmented, or whether they are adequate. Further, a long-term issue which should be addressed is whether clustering of chip fabrication and computer-related industries might require a long-term strategy of regulatory evolution to address cumulative impacts of these related industries.

4.6 Emergency response planning

Section 3 identifies deficiencies in the amount of detail with which the issue of emergency planning was addressed. Of greatest concern is failure to identify all EHSs which might be present at chip fabrication facilities and to estimate amounts which might be present on-site. This is important because US EPA emergency planning requirements are triggered with respect to any EHS only when a threshold quantity on-site is reached. In such cases emergency planning zones must be defined based upon air modelling of a specified worst-case scenario. The EIS should estimate EPZ sizes and show, for reasonable scenarios, which portions of North Greenbush would be included in such zones. Of special interest is whether residential areas would be included. Addressing this potentially contentious issue in advance of a permit application ultimately may streamline permit review.

The complete list of 366 EHSs should be considered. One way of demonstrating consideration of all EHSs is to tabulate them, along with respective threshold quantities, for triggering RMP requirements. In the table, each EHS may be characterised with respect to the likelihood of its use in chip fabrication (such as ‘yes’, ‘maybe, in an emerging technology’ or ‘no’). Reasonable worst-case assumptions about the size of EPZs should be made and resulting maps should be provided to determine which, if any, residential neighbourhoods might be included in an EPZ. If any residences are included the EIS should discuss the social and economic consequences. The expected use of ammonia, an EHS, should be described.

4.7 Sensitive subpopulations

The issue of sensitive subpopulations is primarily an air impact issue, but neither the air resources section of the EIS nor the detailed appendix supporting it [5] addressed sensitive subpopulations. Air dispersion modelling revealed acceptability of air impacts relative to standards and guidelines for individual substances as set forth in Air Guide-1 [28], but Air Guide-1 specifically cautions against the use of these values when multiple air contaminants are emitted. The EIS failed to apply or acknowledge the procedure recommended by NYS DEC. to address multiple emissions (set forth in Appendix C of

Environmental technology transfer synergism 449

ACGIH 1990; [35]). Further, the EIS failed to account for many if not most potentially toxic substances likely to be emitted from chip fabrication facilities, as documented above. Finally, the EIS failed to develop an inventory of locations where potentially sensitive subpopulations would be expected and failed to map air impacts relative to the locations of such sensitive subpopulations.

In conclusion, sensitive subpopulations were not inventoried, nor were plans to inventory them presented. Air emissions were characterised as being acceptable, protecting sensitive subpopulations wherever they are. However, the EIS failed to assess the acceptability of combinations of multiple air emissions as required, failed to evaluate the adequacy of standards and guidelines relative to chip fabrication emissions and failed to account for many, if not most, emissions from such facilities. Accordingly, the EIS should be enhanced with respect to these deficiencies. An inventory of potentially sensitive subpopulations should be developed. Air impacts should be mapped relative to the locations of such populations. Procedures required by NYS DEC Air Guide-1 for evaluating multiple emissions should be applied and a more realistic list of substances which might be emitted should be developed for inclusion in this analysis.

4.8 Plans to educate employees, residents and emergency responders

The level of detail with which this issue area was addressed is rudimentary. Specific training resources, intensities, programs, objectives and schedules were not identified. Justification for this brevity was apparently attempted in the EIS as follows: “A more detailed risk analysis will be part of the facility’s EPA-mandated Risk Management Program” [1, pp.5—18].

However, this approach fails to respond to the purpose of the EIS, which is to evaluate potential risks to public health and the environment so that, if risks are found to be within acceptable limits, the EIS will have accelerated the permit review process. Permit review acceleration was deemed necessary because of the protracted permit process that is in place in New York State. If the EIS reveals compatibility of chip fabrication facilities with the site as proposed, it also must assure chip fabrication companies potentially interested in North Greenbush that they will not face the likelihood of a protracted permit review process. Potentially controversial issues must be addressed to the degree possible before a permit application is received, so that permit review can indeed be accelerated. This will assure, too, that companies contemplating application for a permit in North Greenbush will be encouraged to go forward based upon their clear perception that a credible effort has been undertaken to remove potential permitting obstacles before a corporate commitment has been made.

Asserting that employees, community residents and emergency responders will be educated falls short of describing program(s) that will be implemented to train them and the schedule(s) that will be used to update their training. The EIS failed to describe the qualifications of the people who will be used to train employees, community residents and emergency responders. It failed to identify specific training resources, intensities, programs, objectives and schedules and failed to address the documentation of such training and any process(es) that would be used to certify individuals once they have received such training. It failed to address the issue of accessibility of training records to the public and the issue of possible public input into training requirements. This is in contrast to other industries, such as the airline industry, in which stringent requirements

450 R.A. Michaels

for pilot training and for updated training have been promulgated. Records documenting the successful completion of such training are available for public scrutiny by investigators when deemed necessary.

4.9 Plans for facility maintenance

The frequency of incidents requiring emergency response and the risks of such incidents occurring, will be lessened to the degree that anticipatory facility maintenance assures that containment structures and fabrication processes will continue to function safely as chip fabrication facilities age. Accordingly, plans for facility maintenance constitute an appropriate issue for coverage in the EIS. However, examination of the EIS revealed no section(s) devoted to the issue of facility maintenance. Accordingly, the EIS should be augmented to address the issue area of facility maintenance. Specific issues to be addressed should include maintenance, inspection, documentation of maintenance, public access to inspection and maintenance records and public input into inspection and maintenance requirements. One way of assuring host community residents of the responsiveness of chip fabrication companies to their concerns would be to appoint a safety committee with oversight over maintenance issues and to include representation of the host community on this committee. Other alternatives might be considered, with the overall objective of assuring that public concerns will be heard and addressed responsively throughout the life cycle of the chip fabrication facilities, not just at the intervals when they are permitted or repermitted.

4.10 Plans to monitor processes involving hazardous chemical products

Monitoring can be remote or on-site, invasive or non-invasive, reactive or proactive. Remote monitoring of emissions will require inordinate resources for full-perimeter coverage and measurement of multiple substances at unrealistically low detection limits. On-site monitoring is preferable for assuring compliance with permit conditions. Invasive monitoring may degrade containment structures, posing incremental risks. For example, highly corrosive ammonia cracks metals and breaches seams, posing risks when pressuretesting tanks. Yet vessels containing ammonia must be tested to assure continued competence to reliably contain their product. Non-invasive and less invasive methods are in development.

Proactive monitoring provides early warnings, for example, of cracking in ammonia vessels. It prevents accidents, whereas reactive monitoring may only prevent recurrences of particular accident scenarios. Clearly, a proactive approach should be taken. The SEMATECH DFESH TQESHM approach demonstrates not only that concepts require acronyms to be taken seriously, but that thinking in the semiconductor industry has evolved toward the proactive philosophy of prevention. The EIS should reflect this new thinking by committing to specific monitoring objectives, including monitoring that is on-site, non-invasive where possible and proactive.

The EIS inadequately addressed the issue of chip fabrication facility monitoring and the related issues of inspection, record keeping, record storage and retrieval and public access to records. No assurances were provided that permit conditions will reflect a corporate commitment to monitoring that is on-site, non-invasive where possible and proactive. Accordingly, the EIS should be augmented to more adequately address the issue of monitoring. Specifically, it should address the issues of monitoring on-site, non-

Environmental technology transfer synergism 451

invasively where possible and proactively. It should consider the related issues of facility inspection, record keeping, record storage and retrieval and public access to records. The issues of notification via alarms, the timing of notifications and the list of parties to receive notifications should be addressed with respect’to occasions when monitoring instruments reveal failures of structures and/or processes, with release or the potential for release of substances of potential concern.

References

10 11

12 13

EIS (1998) Draft Supplemental Generic Environmental Impact Statement (DSGEIS), Rensselaer Technology Park, North Greenbush, New York, Murman Associates, December, 4 Vols. including appendices; i. p.

EIS, Appendix 11 (1995) ‘S70 Design for Environment, Safety and Health (DFESH), Implementation Strategy for the Semiconductor Industry’, Technology Transfer #95103006AENG, Austin, Texas; SEMATECH Technology Transfer, 30 November, Vol. 113, p.30.

North Greenbush (1999) ‘Chip fab at RPI Technology Park’, Transcript of Public Hearing, Wynantskill, Rensselaer County, New York, 4 March.

Rensselaer County Legislature (1999) ‘Resolution supporting semiconductor marketing plan and authorizing appropriation of funds for the purpose of supporting the semiconductor marketing plan — economic development and planning’, Troy, New York, 5 March.

EIS, Appendix 3 (1998) Technical Support Documentation, Air Permit Application, Semiconductor Manufacturing Facility, November, i. p.

Basov, D.N., Woods, S.L, Katz, A.S., Singley, E.J., Dynes, R.C., Xu, M., Hinks, D.G., Homes, C.C. and Strongin, M. (1999) ‘Sum rules and interlayer conductivity of high-T, cuprates’, Science, | January, Vol. 283, pp.34-52.

Ahn, C.H., Gariglio, S., Paruch, P., Tybell, T., Antognazza, L. and Triscone, J-M. (1999) ‘Electrostatic modulation of superconductivity in ultrathin GdBa,Cu;O7., films’, Science, 14 May, Vol. 284, No. 1, pp.152-5.

Switzer, J.A. (1999) ‘Electrodeposited ceramic single crystals’, Science, 9 April, Vol. 284, pp.293-296.

Hong, M., Kwo, J., Kortan, A.R., Mannaerts, J.P. and Sergent, A.M. (1999) ‘Epitaxial cubic gadolinium oxide as a dielectric for gallium arsenide passivation’, Science, 19 March, Vol. 283, No. 1, pp.897-900.

Dimitrakopoulos, C.D., Purushothaman, S., Kymissis, J., Callegari, A. and Shaw, J.M. (1999) ‘Low-voltage organic transistors on plastic comprising high-dielectric constant gate insulators’, Science, 5 February, Vol. 283, pp.822-824.

AmLani, I., Orlov, A.O., Toth, G., Bernstein, G.H., Lent, C.S. and Snider, G.L. (1999) ‘Digital logic gate using quantum-dot cellular automata’, Science, 9 April, Vol. 284, pp.289-291.

Christoph, J., Strasser, P., Eiswirth, M. and Ertl, G. (1999) ‘Remote triggering of waves in an electrochemical system’, Science, 9 April, Vol. 284, pp.291-293.

Piner, R.D., Zhu, J., Xu, F., Hong, S. and Mirkin, C.A. (1999) ‘‘Dip-pen’ nanolithography’, Science, 29 January, Vol. 283, pp.661-663.

Schuurmans, F.J.P., Vanmaelbergh, D., van de Lagemaat, J. and Lagendijk, A. (1999) ‘Strongly photonic macroporous gallium phosphide networks’, Science, 2 April, Vol. 284, pp.141-143.

Zimmermann, S., Wixforth, A., Kotthaus, J.P., Wegscheider, W. and Bichler, M. (1999) ‘A semiconductor-based photonic memory cell’, Science, 26 February, Vol. 283, No. 1, pp.292-295.

26 27 28

R.A. Michaels

Aziz, H., Popovic, Z.D., Hu, N-X., Hor, A-M. and Xu, G. (1999) ‘Degradation mechanism of small molecule-based organic light-emitting devices’, Science, 19 March, Vol. 283, No. 1, pp.900-902.

Vorflusev, V. and Kumar, S. (1999) ‘Phase-separated composite films for liquid crystal displays’, Science, 19 March, Vol. 283, No. 1, pp.903-905.

Wegh, R.T., Donker, H., Oskam, K.D. and Meijerink, A. (1999) ‘Visible quantum cutting in LiGdF,:Eu** through down-conversion’, Science, 29 January, Vol. 283, pp.663-666.

Michaels, R.A. (1999) ‘Emergency planning and the acute toxic potency of inhaled ammonia’, Environmental Health Perspectives, August, Vol. 107, No. 8.

Michaels, R.A. (1996) ‘Airborne particle excursions contributing to daily average particle levels may be managed via a one-hour standard, with possible public health benefits’, Aerosol! Science and Technology, November, Vol. 25, pp.437-444.

Michaels, R.A. (1998) ‘Permissible daily airborne particle mass levels encompass brief excursions to the ‘London fog’ range which may contribute to daily mortality and morbidity in communities’, Applied Occupational and Environmental Hygiene, June, Vol. 13, No. 6, pp.385-394.

Michaels, R.A. (1997) ‘Particulate matter policy’, Science, 5 December, Vol. 278, p.1696 (letter).

NFPA (1999) Titanium Production, Processing, Handling and Storage, Quincy, Massachusetts; National Fire Protection Association (NFPA), Standard 481.

NFPA (1999) Zirconium Production, Processing, Handling and Storage, Quincy, Massachusetts; National Fire Protection Association (NFPA), Standard 482.

NRC (1993) Guidelines for Developing Community Emergency Exposure Levels for Hazardous Substances, Washington, DC; National Research Council, National Academy Press.

US Congress (1986) Superfund Amendments and Reauthorization Act (SARA), Title III. US Congress (1990) Clean Air Act Amendments, Section 112(r).

NYS DEC (1991) Air Guide 1, Albany, New York; New York State Department of Environmental Conservation, Division of Air Resources, i. p.

EIS, Appendix 10 (undated) Input on Environmental Health and Safety Impacts of Semiconductor Fabrication Facility for Incorporation into EIS for Site in North Greenbush, New York, Cambridge, Massachusetts; EnviroBusiness, Inc.

NTSB (1971) ‘Railroad accident report: Chicago, Burlington and Quincy Railroad Company, Train 64 and Train 824, derailment and collision with tank car explosion, Crete, Nebraska, Feb. 18, 1969, 6:30 a.m.’, Washington, DC; National Transportation Safety Board, Report No. NTSB-RAR-71-2.

NTSB (1974) ‘Pipeline accident report: mid America pipeline system, anhydrous ammonia leak, Conway, KA; Dec. 6, 1973’, Washington, DC; National Transportation Safety Board, Report No. NTSB-PAR-74-6, 1974.

NTSB (1978) ‘Railroad accident report: Louisville and Nashville Railroad Company, freight train derailment and puncture of anhydrous ammonia tank cars at Pensacola, FL; Nov. 9, 1977, 6:06 a.m.’, Washington, DC; National Transportation Safety Board, Report No. NTSB-RAR-78-4.

NTSB (1979) ‘Railroad accident report: Louisville and Nashville Railroad Company, freight train derailment and puncture of anhydrous ammonia tank cars, Crestview, FL; April 8, 1979, 8 a.m.’, Washington, DC; National Transportation Safety Board, Report No. NTSB-RAR-79-11.

NTSB (1979) ‘Survival in hazardous materials transportation accidents’, Washington, DC; National Transportation Safety Board, Report No. NTSB-HZM-79-4.

ACGIH (1990-1991) ‘Threshold limit values for chemical substances and physical agents and biological exposure indices’, American Conference of Governmental Industrial Hygienists, Cincinnati, Ohio; i. p.

Publication citation

Michaels, RA. Environmental technology transfer synergism: health risk assessment and community rejection of commercial electronic chip fabrication in upstate New York. International Journal of Technology Transfer and Commercialisation, 1(4):424-52, doi: 10.1504/IJTTC.2002.001797, 2002;

Read the publication

View DOI record

Robert A. Michaels on ResearchGate

Robert A. Michaels

Environmental technology transfer synergism: health risk assessment and community rejection of commercial electronic chip fabrication in upstate New York

Publication citation

Read the publication

EPA can no longer ignore risks potentially posed by hours-long PM excursions

Particulate matter: the Marines, EPA, and air quality

RAM TRAC Corporation

Contact