Research article
Received: 11 January 2014 Revised: 23 October 2014 Accepted: 27 October 2014 Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI 10.1002/jrs.4618Raman spectroscopic characterization of
laminated glass and transparent sheet plastics
to amplify a history of early aviation ‘glass’†
Odile Madden,a* Kim Cullen Cobba and Alex M. SpencerbA novel, non-invasive study of goggles, flight helmets, airplane windows, and canopies in Smithsonian collections is the first known
large-scale technical survey of historic aviation plastics and leverages the world’s largest air and space collection as evidence of the
materials and technologies used to create transparent plastic objects in the early-20th century. Transparent windows in these artifacts
were analyzedwith Fourier transform and portable dispersive Raman spectrometers to identify polymers and plasticizers present. The
study demonstrates the potential of Raman spectroscopy to objectively and non-destructively measure historic plastic compositions,
including formulations that havebecomeobsolete. Datawas interpreted in combinationwith archival research of historical documents
to identify windowmaterials including glass, laminated safety glass, and sheets of plasticized cellulose nitrate, plasticized cellulose
acetate, and poly(methyl methacrylate). Results are contextualized into a coherent history of the role transparent plastics played
in enclosing airplane cockpits. Published 2014. This article is a U.S. Government work and is in the public domain in the USA.
Additional supporting information may be found in the online version of this article at the publisher’s web-site.
Keywords: plastic; cellulose acetate; cellulose nitrate; PMMA; aircraft cockpit* Correspondence to: Odile Madden, Museum Conservation Institute, Smithsonian
Institution, 4210 Silver Hill Road, Suitland, MD 20746, USA.
E-mail: maddeno@si.edu
† This article is part of the special issue of the Journal of Raman Spectroscopy enti-
tled “Raman in Art and Archaeology 2013” edited by Polonca Ropret and Juan
Manuel Madariaga.
a Museum Conservation Institute, Smithsonian Institution, 4210 Silver Hill Road,
Suitland, MD, 20746, USA
b National Air and Space Museum, Smithsonian Institution, 600 Independence
Avenue SW, Washington, DC, 20560, USA
12
1Introduction
Early aviation design incorporated the most innovative plastics avail-
able at the time, and examples of these technologies are represented
in the SmithsonianNational Air and SpaceMuseum (NASM) collection.
Particularly interesting, and unexplored until now, is the co-evolution
of transparent sheet plastics and the enclosure of cockpits in
heavier-than-air aircraft of the 1920s and 1930s. A novel, non-invasive
study of goggles, flight helmets, airplane windows, and canopies in
Smithsonian collections is the first known large-scale technical survey
of historic aviation plastics and leverages the world’s largest air and
space collection as evidence of the materials and technologies used
to create plastic objects in the early-20th century. The study relied
heavily on Raman spectroscopy to non-invasively identify the poly-
mers and plasticizers present in the artifacts’ transparent windows.
Many of the artifacts studied date to the earliest years of trans-
parent plastic sheets, when compositions were experimental,
evolving, and often different from polymer formulations that are
common today. For this reason analysis was combined with study
of historical documents that describe the state of the art when
the artifacts were manufactured. This combined approach showed
that a range of plastic materials were used, including laminated
safety glass and sheets of cellulose nitrate and cellulose acetate,
all of which required low molecular weight plasticizers, and
eventually poly(methyl methacrylate) (PMMA). These findings
support our proposition that the enclosure of aircraft cockpits was
made possible by and spurred innovation in transparent plastics.
Background
In 1903 the Wright Brothers made their first powered flight at Kitty
Hawk in an open architecture aircraft. It has been estimated that theJ. Raman Spectrosc. 2014, 45, 1215–1224 Published 2014. Thtotal weight of the Wright flyer including the pilot was a mere 625
pounds, and it was moved by an 8 horsepower engine.[1] The
dangers of flying in the open were obvious as pilots could be
pitched out of the plane and killed.[2] By 1920, airplanes were more
streamlined, and pilots were given some protection from the
slipstream and elements by covering the aircraft’s skeletal frame
with fabric and leaving an opening on top for the pilot’s head
and shoulders.[3] This would have improved an airplane’s aerody-
namics by directing the air around rather than through it, but the
big hole on top was a significant source of drag. It also did little
for pilot comfort and protection. These early cockpits were located
behind the engine, and the pilot was pelted with wind, rain, ice,
oil, and the occasional bird that happened into the propeller. All
threatened the pilot’s ability to see and maneuver the plane.
Goggles and small windshields were a first defense, provided they
did not fail.[4,5] In the ensuing decades, as planes carried heavier
loads and flew faster, higher, and longer distances, the open
cockpit would have become an engineering challenge, and yet
cockpits were not enclosed consistently until the late 1930s.is article is a U.S. Government work and is in the public domain in the USA.
5
O. Madden, K. C. Cobb and A. M. Spencer
12
16The central problem was that the pilot needed to see out of the
plane in order to fly, and we posit that throughWorldWar I suitable
window materials did not yet exist. In 1903, the landscape of trans-
parent window materials would have included glass and some
polymers such as cellulose nitrate, gelatin, and shellac. A successful
window would not add too much weight to the plane, be resilient
against changes in pressure and temperature, and protect against
impacts. Plate glass was inappropriate for its weight and tendency
to shatter on impact into sharp shards that could lacerate and kill
the pilot. If a 1/4″ thick pane of glass 1 foot square can weigh from
3 to 8 lbs, and enclosing an airplane conservatively might take
twenty of these, the glass alone would add 60–160 lbs to the
vehicle weight. For a sense of scale, this would have increased the
weight of the Wright flyer by 10–26%. The carrying capacity of
planes increased dramatically in the ensuing decades, but the
weight of windows in the first decades was important.[6] Injury from
broken glass also would have been a hazard, as experience showed
with automobiles that were undergoing a similar evolution.[7,8] A
plate glass enclosure would have been unsuitable for an airplane,
where the stakes for crashes were much higher.[7–10] Newmaterials
with transparency, strength, and fracture behavior suitable for
aviation were developed in the ensuing three decades by which
time enclosed aircraft had become the norm.
Materials and methods
This study of aviation materials combined research of historical
documentswith visual examination, Raman spectroscopy, and some
X-ray fluorescence spectroscopic (XRF) analysis of historic artifacts to
create a timeline of windowmaterials used in early aviation. Archival
research helped with deciphering Raman spectra that were hard
to interpret because the formulation has become obsolete or was
a mixture.
Artifact survey and comparative materials
Eighty-seven aviator goggles, flight helmets, and aircraft windows
in the NASM collection were examined, andmany of these included
more than one window or lens. Raman spectra were collected from
76 windows and lenses from 39 artifacts, and the remainder was
identified visually. Artifacts are identified here by the 12-character
object numbers assigned by the museum (i.e. A19720891000).
Raman spectra of artifacts were compared to spectra of known
reference materials that were measured at the Smithsonian’s
MuseumConservation Institute (MCI) with a FT-Raman spectrometer.
Of the reference materials, cellulose acetate (CAS 9004-35-7), diethyl
phthalate (CAS 84-66-2), and poly(methyl methacrylate) (CAS
9011-14-7) were purchased from Scientific Polymer Products, Inc.
(6265 Dean Parkway Ontario, New York 14519). Dimethyl phthalate
(CAS 113-11-3, Aldrich number 525081) and triphenyl phosphate
(CAS 115-86-6, Aldrich number 241288) were purchased from Sigma
Aldrich (sigmaaldrich.com). The cellulose nitrate reference was an
experimentally prepared dope that was cast at MCI in 1994. The
reference spectrum for camphor ((+/
)-Camphor, 96%, CAS 21368-
68-3, Aldrich number 148075) was obtained from a commercially
available spectral library (HR Aldrich Raman library, Copyright
1996–2001, 2004 Thermo Electron Corporation for Nicolet Raman).
Raman analysis
Two Raman spectrometers were used. A research grade Fourier
transform Raman spectrometer with 1064 nm excitation (FT-Raman)wileyonlinelibrary.com/journal/jrs Published 2014. This article is a
is in the public domoffered spectra with high spectral resolution and low fluorescence
collected in a laboratory setting, while a portable dispersive instru-
ment with 785nm excitation was evaluated for analysis of artifacts
that were challenging because of their fragility, shape, size, or loca-
tion. Both instruments are owned by MCI.
A portable B&W Tek MiniRam II dispersive Raman spectrometer
was the primary tool. It incorporates a 785nm excitation laser,
1.5m fiber optic probe with 85micron analytical spot, and
thermoelectrically cooled CCD detection. An adjustable working
distance adapter (described elsewhere) was designed to facilitate
analysis of laminated glass structures.[11] Spectra were a co-addition
of 12–64 1 second scans across 160–3207 cm
1 at 10 cm
1 spectral
resolution. The instrument was controlled and data were collected
using BWSpec™ spectral acquisition software (B&WTek, Inc.). When
possible, goggles with colored or deteriorated plastic lenses that
fluoresced under 785 nm light also were analyzed in the laboratory
by FT-Raman using a NXR module coupled to a 6700 Fourier
transform infrared spectrometer (Thermo Electron Corporation).
The NXR module is equipped with a Nd:YVO4 excitation laser that
emits 1064nm radiation, CaF2 beam splitter, and a Peltier-cooled
InGaAs detector. Analyses were conducted in an enclosed sample
compartment, which restricted the artifacts that could be studied
to thosewith suitable dimensions. Each spectrumwas a co-addition
of 1024–4096 scans across 100–3700 cm
1 and 2–8 cm
1 spectral
resolution. Instrument control, data collection, and spectral inter-
pretation were managed by OMNIC 7.2a software (Thermo Fisher
Scientific). All analyses were performed in situ without removing
samples from the artifacts. Laser power was chosen on a case-by-
case basis to optimize spectral quality and minimize risk to the
artifacts, which exhibited a wide range of energy tolerances that
could not be predicted reliably from an artifact’s appearance. In
each case, the analyst started with the lowest laser power available
and increased it carefully until a usable spectrum was achieved. No
baseline correction or smoothing was applied to spectra.
X-ray fluorescence spectrometry (XRF)
Elemental analysis supplemented the Raman results and visual
examination in some cases. XRF spectra were collected with two
instruments that offer different sampling geometries: an ElvaX
energy dispersive X-ray fluorescence spectrometer (ElvaTech)
(X-ray tube with rhodium anode, Si-pin detector, 165 eV resolution)
and a Bruker ARTAX micro-XRF spectrometer (X-ray tube with
rhodium anode, silicon drift detector, 70 eV resolution).
Archival research
Trade literature, patents, company records, and published histories
were mined for information about the evolution of goggles, airplane
design and the enclosure of cockpits, and the technological history of
laminated safety glass, cellulose nitrate, cellulose acetate, and PMMA.Results
What at the outset was hoped to be a straightforward set of trans-
parent, colorless plastic windows turned out to be more complex.
The collection to which we had access included a range of glasses,
laminated glasses, and plastic sheets. Furthermore, many were
colored, chemically degraded, or otherwise damaged. Three
artifacts did not have any lenses at all. Ultimately, 76 goggle lenses,
visors, and windows from 39 artifacts were measured by portable
dispersive and/or FT-Raman spectroscopy. Summarized results forU.S. Government work and
ain in the USA.
J. Raman Spectrosc. 2014, 45, 1215–1224
Raman spectroscopic characterization of historic aviation ‘glass’the Raman analyses are provided in Table 1 (windowmaterials that
include polymers) and Table S5 (window materials that include
polymers and/or glass).Raman spectroscopy and XRF
Of 87 artifacts examined, 39 goggles had glass lenses that could be
distinguished easily with the portable Raman spectrometer. Seven
of these goggles were analyzed by Raman as a proof of concept,
and the other 32 were identified visually. The priority of the
research was to trace the evolution of plastic materials, so these
glass lenses were given only cursory attention. One interesting
discovery by XRF was the presence of ultraviolet-absorbing cerium
compounds in the glass lenses of some World War II era eyewear
(A19760039000, A19781761000, A19810332000, A20010495000,
and A20050338000).Table 1. Results summarized by window material and fabrication date. Obje
Window/lens material per Raman Object number
Plasticized cellulose nitrate sheet
CN+ camphor A19250008000
Laminated safety glass
Laminated safety glass, CN + camphor A20090121000
Laminated safety glass, CN + camphor A19500122000
Laminated safety glass, CN + camphor A19620048000
Laminated safety glass, CN + camphor A19790846000
Laminated safety glass, CN + camphor A19910095000
Laminated safety glass, fluorescent unidentified interlayer A19772777000
Laminated safety glass, fluorescent unidentified interlayer A20000878000
Laminated safety glass, unidentified interlayer A19603332000
Laminated safety glass, CN + camphor A19540007000
Laminated safety glass, CA +DMP A19610112000
Laminated safety glass, unidentified interlayer A19761327000
Plasticized cellulose acetate sheet
CA+DEP +DMP+ TPP A19720891000
CA+DEP +DMP+ TPP A19810020000
CA+DEP +DMP+ TPP A19810185000
CA+DEP + TPP A19740180000
CA+DEP + TPP T20051030000
CA+DEP + TPP T20051031000
CA+DMP+ TPP, possibly DEP A19970586000
CA+DEP + TPP, possibly DMP A19880585000
CA+DEP + TPP, possibly DMP A19950796000
TPP + herapathite (CA, DEP likely) A19910412000
CA+ TPP (or similar compound) A20050053000
Poly(methyl methacrylate) Sheet
PMMA A19602082000
PMMA A19600324000
PMMA A19781335000
PMMA A19731430000
PMMA A19810003000
PMMA A19800018000
Poly(vinyl chloride) sheet
PVC + dioctyl phthalate A20060135000
CA= cellulose acetate, CN= cellulose nitrate, DEP = diethyl phthalate, DMP =
chloride, and TPP= triphenyl phosphate.
J. Raman Spectrosc. 2014, 45, 1215–1224 Published 2014. This article is a
is in the public doTen sets of goggles had lenses of shatterproof laminated safety
glass. These consist of two sheets of glass sandwiched on either side
of, and adhered to, a thin (<1mm) plastic sheet. The fiber optic
probe of the portable Raman was customized with an adjustable
working distance adapter in order to probe the polymer layers
through the glass, and these results are described in detail
separately.[11] Eight of these laminated glass goggles were a flat-
paned style that was common during World War I (Fig. 1). Portable
Raman spectra indicated that the interlayer in six of the eight was
cellulose nitrate plasticized with camphor. The remaining two had
a red-tinted interlayer that was highly fluorescent and could not
be identified. The ninth pair of laminated glass goggles, a British
Royal Air Force style (A19761327000), had a face pad of synthetic
leather that suggests a later manufacture date, and a colorless
interlayer in the lenses with a unique but unidentified Raman
spectrum. The tenth, Japanese goggles with curved lenses thatct numbers are permanent identifiers that are searchable through NASM.
Artifact type Windows/lenses
analyzed
Fabrication date
Airplane window 1 1924
Aviator goggles 1 Circa 1920–1930s
Aviator goggles 1 Style consistent 1914–1918
Aviator goggles 2 Style consistent 1914–1918
Aviator goggles 1 Style consistent 1914–1918
Aviator goggles 1 Style consistent 1914–1918
Aviator goggles 2 Style consistent 1914–1918
Aviator goggles 1 Style consistent 1914–1918
Aviator goggles 1 Circa late-1930s–1945
Aviator goggles 1 Uncertain
Airplane window 1 1938
Aviator goggles 1 Uncertain
Aviator goggle lenses 8 Post 1943
Aviator goggle lenses 3 Post 1943
Aviator goggles 1 Post 1943
Aviator goggles 1 Post 1943
Aviator goggles 6 Post 1943
Aviator goggle lenses 13 Post 1943
Aviator goggles 1 Post 1943
Aviator goggles 1 Circa 1944–1956
Aviator goggles 1 Post 1974
Aviator goggles 3 Post 1943
Aviator goggles 2 Uncertain
Airplane canopy 3 1944
Airplane canopy 3 1945
Flight helmet 1 Circa1956
Flight helmet 1 Circa 1960
Flight helmet 2 Circa 1960
Flight helmet 2 Circa 1965
Jump goggles 1 Circa 2004
dimethyl phthalate, PMMA=polymethyl methacrylate, PVC = polyvinyl
U.S. Government work and
main in the USA.
wileyonlinelibrary.com/journal/jrs
12
17
Figure 1. World War I era aviator goggles with shatterproof glass lenses
laminated with a mixture of cellulose nitrate and camphor. The uneven
darker tint at the center of the lenses is discoloration related to age.
O. Madden, K. C. Cobb and A. M. Spencer
12
18date stylistically toWorld War II (A19603332000), also had a unique,
unidentified interlayer spectrum.
Among the goggles and helmet visors with plastic lenses, three
polymers were identified: cellulose acetate, PMMA, and polyvinyl
chloride. Sixteen goggles were in the B-8 style, characterized by a
one-piece rubber frame and a single mask-like lens (Fig. 2). The
lenses were mass produced and designed to be interchangeable.
Several of these goggles had multiple lenses, in a variety of tints,
for a total of 47. Thirty-eight of these lenses were identified by Ra-
man as cellulose acetate plasticized with a combination of diethyl
and/or dimethyl phthalate and triphenyl phosphate, and a charac-
teristic spectrum is shown (Fig. 3, Table S1). Cellulose acetateFigure 2. M-1944 goggle (1974), a version of the B-8 goggle that became
popular during World War II, consists of a molded rubber frame and a
one-piece, interchangeable plastic lens.
Figure 3. (From top) Raman spectrum of lens of M-1944 goggle (1974) col
constituent compounds, cellulose acetate, diethyl phthalate, and triphenyl pho
wileyonlinelibrary.com/journal/jrs Published 2014. This article is a
is in the public dompolymer compounded with one or more low molecular weight
phthalates and triphenyl phosphate was a common formulation
for plastic sheets and injection-molded objects from the 1930s on-
ward, and typical Raman spectra have been described by Tsang
et al.[12,13] These four compounds often can be discerned from a Ra-
man spectrum even though many peaks overlap. For undegraded,
plasticized cellulose acetate artifacts in the current study, Raman
spectra could be translated as follows. Diagnostic peaks for the cel-
lulose acetate polymer are 2939, 1435, 1379 (with shoulder at 1363),
910, 604, and 513 cm
1. The carbonyl stretches at 1738 cm
1 in
pure cellulose acetate and 1728 and 1725 cm
1 in dimethyl and
diethyl phthalate, respectively, tend to merge into an intermediate
value when these substances are compounded into a plastic. Diag-
nostic peaks common to diethyl and dimethyl phthalate that do
not overlap with cellulose acetate or triphenyl phosphate are
1601, 1581, 1041, and 403 cm
1. Dimethyl phthalate can be distin-
guished by the presence of additional peaks at 2844, 820, and a
shoulder at 393 cm
1 and the absence of peaks at 866, 848, and
351 cm
1. The converse is true for diethyl phthalate. For triphenyl
phosphate, breathing of three benzene rings gives rise to a strong
signal at 1007 cm
1. Other characteristic peaks of triphenyl phos-
phate compounded with cellulose acetate include 1593, 1232,
and 933, a partially shifted peak that spans from 726 to 719, and
617 cm
1. Peaks typical of pure triphenyl phosphate crystals
in the 270–170 cm
1 range are not diagnostic when the sub-
stance is dissolved in undegraded plastic.
Spectra of colorless lenses tended to be simpler to interpret due
to lower fluorescence backgrounds, but this was not always the
case. In most instances, colored lenses (yellow, orange, red, green,
and blue) exhibited fluorescence at 785 or 1064nm excitation,
but the strongest peaks of the polymer and/or a plasticizer usually
could be discerned at one of these wavelengths. Lenses that
contained polarizing media, imprinted ‘Polarizing’ along the upper
edge, were highly fluorescent at both laser wavelengths. Detection
of iodine in these lenses by XRF suggests the polarizing medium
may be herapathite (iodoquinine sulfate) crystals.[14]
Four American flight helmets that date from circa 1956 to 1965
were shown to have visors of PMMA using the portable Raman
system (Figs. 4 and 5, Table S2). Peaks at 1730, 1450, 1122, 985,lected with the portable instrument, compared to reference spectra of its
sphate, collected with the FT-Raman.
U.S. Government work and
ain in the USA.
J. Raman Spectrosc. 2014, 45, 1215–1224
Figure 4. United States Navy Type HGU-20/P flight helmet (Photo:
Smithsonian National Air and Space Museum (NASM 2011-00518)).
Figure 6. Two experimental German aircraft fromWorld War II that feature
transparent canopies of PMMA: (a) Horten H III h (1944), and (b) Horten H IX
V3 (1945). (Photos by (a) Dane Penland (NASM 2012-01189) and (b) Eric Long
(NASM 2000-9339), Smithsonian National Air and Space Museum).
Raman spectroscopic characterization of historic aviation ‘glass’966, 812, 600, 559, 484, 364, and 297 cm
1 were visible consistently
with a co-addition of 12 1 second scans. Three visors were
tinted, and the colorant was not identified, but this may be
the source of unidentified peaks at 639, 537, 520, and 416 cm
1 in
the spectrum of A19800018000 (Fig. 5, Table S2). One set of gog-
gles worn by the United States Army Parachute Team ‘Golden
Knights’ (A20060135000, circa 2004) was constructed of a sheet
of polyvinyl chloride plasticized with a phthalate plasticizer,
most likely dioctyl phthalate. The one-piece lens was framed
by colorless clear tubing made of the same compounds but with
a higher concentration of the plasticizer. A set of survival gog-
gles (A19750266000), still sealed in its original polyethylene
bag, had lenses that fluoresced strongly under 785 nm excita-
tion, and could not be identified. However, the transparent pack-
age (polyethylene) and nylon headband were identified with the
portable instrument.
Of the four airplanes studied, the window from the 1924 Douglas
World Cruiser ‘Chicago’ (A19250008000) was identified by FT-
Raman as a sheet of cellulose nitrate plasticized with camphor
(Figs. 7 and 8, Table S3). Camphor is the stronger Raman scatterer,Figure 5. Raman spectra of canopies of two experimental German aircraft and the visor of a United States Navy flight helmet weremeasuredwith a portable
Raman instrument and identified as PMMA. (Spectra from top to bottom) Horten H III h (1944), Horten H IX V3 (1945), United States Navy Type HGU-20/P
protective flying helmet (1965–1971), and a FT-Raman reference spectrum of PMMA.
J. Raman Spectrosc. 2014, 45, 1215–1224 Published 2014. This article is a U.S. Government work and
is in the public domain in the USA.
wileyonlinelibrary.com/journal/jrs
12
19
Figure 7. (a) Douglas World Cruiser ‘Chicago’ (1924) and (b) recent
photograph of a disintegrating utility window identified by Raman as a
mixture of cellulose nitrate plasticized with camphor (Historic photograph:
Smithsonian National Air and Space Museum NASM 88-7415).
O. Madden, K. C. Cobb and A. M. Spencer
12
20and its sharp peak at 652 cm
1 is the most prominent of the
spectrum. This was also the diagnostic peak for cellulose nitrate
interlayers in laminated glass, along with a weak cellulose nitrate
signal at 850 cm
1.[11] In the ‘Chicago’ window spectrum 37
camphor peaks are visible, plus the carbonyl stretch at 1743 cm
1Figure 8. (From top) Spectrum of utility window from the Douglas World Cru
nitrate and camphor, all measured by FT-Raman.
wileyonlinelibrary.com/journal/jrs Published 2014. This article is a
is in the public domwhich is shifted to lower wavenumbers (1732 cm
1) in the
compounded plastic. A comparable reference spectrum and inter-
pretation was published by Paris and Coupry in 2005, although
we would assign the peak at 393 cm
1 to camphor rather than
cellulose nitrate.[15] Features of cellulose nitrate also are present in
the ‘Chicago’ window spectrum, with defined peaks at ~1647,
1373, 1284, 1120, and 850 cm
1, but in general the cellulose nitrate
contribution serves more to fill out the shape of peaks groups in
which the camphor signal is more obvious.
A side window on the Sikorsky JRS-1 (A19610112000, fabricated
in 1938) was identified as laminated safety glass with a cellulose
acetate interlayer plasticized with dimethyl phthalate (Figs. 9 and
10, Table S4). The transparent canopies of the German experimental
aircraft Horten H III h (A19602082000, fabricated in 1944) and
Horten H IX V3 (A19600324000, fabricated in 1945) are PMMA
(Figs. 5 and 6, Table S2).
Proposed historical timeline
Patents and trade literature of the time combined with Raman
spectral data of NASM artifacts support our hypothesis that the
development of transparent windowmaterials was a rate determining
factor in the movement toward enclosed aircraft along with the
development of more powerful engines.
As described above, plate glass was inappropriate for early
airplane windows because of its weight and tendency to shatter.
Plastic sheets of that time would have offered the advantage of
light weight but would seem unlikely to have had great success
as windows because of their poor resilience to impact and pressure
changes. However, the artifact record shows that cellulose nitrate
windows were used in aircraft. The Douglas World Cruiser ‘Chicago’
is one of a set of four planes that flew around theworld in 1924 and
was acquired by NASM soon after (Fig. 7).[16] The cockpit was open,
but some small utility windows on the wings presumably allowed
for inspection of the equipment. One of these utility windows that
had deteriorated severely and was removed from the airplane was
analyzed for this study and identified as cellulose nitrate plasticized
with camphor (Fig. 8). This is the classic formulation commonly
described by the trade names Celluloid, Parkesine, and Xylonite.[17]
Its chemical instability, which can manifest itself in the yellowing,
opacification, and crizzling (a term used by art conservators to
describe a pattern of fine cracks, usually in chemically unstableiser ‘Chicago’ and reference spectra of its constituent compounds, cellulose
U.S. Government work and
ain in the USA.
J. Raman Spectrosc. 2014, 45, 1215–1224
Figure 9. Sikorsky JRS-1 amphibious aircraft (1938). (Photo by Dane
Penland, Smithsonian National Air and Space Museum (NASM 2012-01396)).
Raman spectroscopic characterization of historic aviation ‘glass’glass, that renders the material dull and less transparent) exhibited
in Fig. 7b, was well known by the 1920s, and it is a long-term
stability problem in Smithsonian collections.[11] Despite thewindow’s
degraded state, its Raman spectrum is easily comparable to the
reference spectra.
Cellulose nitrate also was used in windows as a laminating layer
in shatterproof safety glass. This technology remains the industry
standard for windshields in airplanes and cars today although the
polymer interlayer has evolved. The invention of safety glass
typically is attributed to French chemist Edouard Benedictus who
developed and patented ‘Triplex safety glass,’ a laminate composed
of a sheet of cellulose nitrate plastic adhered with gelatin between
two glass plates, in around 1910.[18–24] In fact, the earliest patent for
laminated safety glass for automobile windows was issued to John
Crewe Wood in England in 1905.[25] Benedictus’ first patent shows
his appreciation of safety glass’ great potential in a broad range
of applications that have since become reality.[10,20] His composite
glass was intended for windows andwindshields in vehicles includ-
ing “automobiles, cabs, carriages, omnibuses, railway carriages,
boats, and the like,” though aircraft were not specified at that early
date.[20] He also noted that his glass could be used for windows
with “quite original effects” by coloring or otherwise ornamenting
the cellulose nitrate or glass.[20]Figure 10. (From top) Spectrum of the laminated safety glass interlayer from a
constituent compounds, cellulose acetate, and dimethyl phthalate, all measured
J. Raman Spectrosc. 2014, 45, 1215–1224 Published 2014. This article is a
is in the public doBecause Triplex required non-standard thicknesses of flat, defect-
free glass, and a multi-step manufacturing process conducted in
cleanroom-like conditions, it was much more expensive than win-
dow glass, and this slowed its adoption.[9,21] World War I provided
a major incentive for its development into “airplane and automo-
bile windshields, bulletproof glass for tanks, glass for submarines,
battleship-bridge windows, and eye blanks for gas masks and
aviators’ goggles.”[10] In particular, lenses for eye protection made
economic sense because of their small dimensions. They were
adopted for civilian automobile driving and aviation, and were
specified by the United States War Department for pilots in
1917.[26] In the NASM collection, laminated safety glass is found in
aviator goggles of theWorld War I period and airplane windshields,
and a cellulose nitrate/camphor formulation was identified by
Raman in the majority of those studied.[11] The fact that the lenses
are laminated suggests that protection of the eyes from impacts
was a concern. This and other details such as a warm, fleecy face
pad (A19540007000) are consistent with open cockpit flying.
Other interesting features of these goggles include tinted interlayers
(e.g., cyan (A20090121000), amber (A19772777000, A20000878000),
and two-tone blue and clear (A19910095000)), as suggested by
Benedictus and intimates the importance of being able to see in a
range of lighting conditions.
By the 1920s the shortcomings of cellulose nitrate were well
known, and safety glass interlayers were known to deteriorate
within a few years of use.[10,21,27] Several goggles in the NASM
collection exhibit typical discoloration, shrinkage, and separation
of the interlayer (A19500122000, A19620048000, A19790846000,
and A20090121000), and a mild example is shown in Fig. 1.[11] In
time the manufacturing process was modified to include sealing
the edges of the laminate sandwich to prevent infiltration of
moisture to the plastic layer and slow degradation, but ultimately
cellulose nitrate was phased out in favor of cellulose acetate, a poly-
mer that was developed as a non-flammable substitute in the late
1920s.[10,21,28–31] Cellulose acetate was adopted for many of the
same applications, including as a safety glass laminating layer that
was more chemically stable than its predecessor.[28,29,32]
Cellulose acetate laminated glass was not encountered in the
NASM goggle collection but was found in the side windows of
the Smithsonian’s United States Navy JRS-1, a twin engine amphib-
ious aircraft built by the American company Sikorsky Aircraft in
1938 (Fig. 9). It is the one airplane in the NASM collection thatround side window of the Sikorsky JRS-1 (1938), and reference spectra of its
by FT-Raman.
U.S. Government work and
main in the USA.
wileyonlinelibrary.com/journal/jrs
12
21
O. Madden, K. C. Cobb and A. M. Spencer
12
22was stationed at Pearl Harbor on 7 December 1941 and is the sole
surviving aircraft of its kind.[33] Raman spectra confirmed that the
round side windows of this transport are laminated safety glass
with an interlayer of cellulose acetate plasticized with dimethyl
phthalate (Fig. 10). This would have been a late example of that in-
terlayer, which underwent a major change the following year. One
drawback of cellulose acetate laminated safety glass was its
rigidity.[10] When struck by an object, like a person’s head, the glass
would break but not yield, and this caused head trauma. This public
health problem was taken on in 1939 by a collaboration of chem-
ists, engineers, and manufacturers from E. I. du Pont de Nemours
and Company, Pittsburg Plate Glass Company, Libby-Owens-Ford
Glass Company, Monsanto Chemical Company, and Carbide and
Carbon Chemicals Corporation who spent $6 million to develop
an ideal safety glass laminating layer.[9,34] Their polyvinyl butyral
resin (PVB) was stable chemically, dimensionally, and mechanically;
had a broad glass transition temperature range that accommo-
dated cold and hot climates; possessed superior adhesion qualities;
and was flexible and easy to work with. When a projectile hit this
safety glass it would deform (and transfer the force of impact into
the window), but the large fragments would not detach. This resin
remains an industry standard today. PVB safety glass was not
identified in the artifacts available for this study though the authors
assume it is common in laminated glass windshields after 1939 and
that thewindowof the Sikorsky JRS-1 is an example of a technology
in its waning years.
By 1938 another sea change in airplane window glazing was un-
derway. In the 1930s, German chemist Otto Rohm and his associate
Walter Bauer experimented with PMMA as a potential safety glass
interlayer. They poured the monomer directly between two sheets
of glass and polymerized it in place with light. However, instead of
cementing the two panes together as intended, the polymer sepa-
rated cleanly from the glass in a strong, solid, transparent, optically
clear, glass-like material that could be cast, sawed, bent, and ma-
chined. They quickly recognized its potential in aircraft
glazing.[35–37] A full-page advertisement extolling its light weight,Figure 11. Raman spectra of four windowmaterials used in aviation. (From top
‘Chicago’ (FT-Raman), laminated safety glass interlayer of cellulose acetate and
with diethyl phthalate and triphenyl phosphate from M-1944 goggle lens (p
flying helmet (portable Raman).
wileyonlinelibrary.com/journal/jrs Published 2014. This article is a
is in the public domhardness, impact strength, perfect transparency, weathering, and
ease of forming appeared in the July 1937 issue of the New York
based journal, Aero Digest, where it was placed prominently oppo-
site the table of contents.[35] Windows, gun-turrets, observation
blisters, and nose cones of PMMA soon proliferated and became
popular symbols of World War II aeronautic technology.[37,38] Rohm
and Haas had operations in Germany and the United States, and
their Plexiglas was used in both markets.[32] Raman analysis shows
that PMMAwas the plastic sheet of transparent canopies on two ex-
perimental ‘Horten’ German jet airplanes constructed during World
War II; the Horten H III h and Horten H IX V3 each feature a PMMA
canopy constructed in three parts (Fig. 6). PMMA also was used
for the visors on four flight helmets used by the United States Navy
and Air Force in the 1950s through 1970s (A19731430000,
A19781335000, A19800018000, and A19810003000) (Fig. 4).
Aviation glazing was a large potential market for plastics, and a
range of competitors came on the market around this time. Cellulose
acetate, polyvinyl chloride, cellulose nitrate, and laminated safety glass
options appear in Aero Digest beginning in 1938.[39] Unpublished re-
search and development reports from the Celluloid Manufacturing
Company show that company was actively trying to compete in the
aircraft window market by 1943 with their Aero Quality Lumarith
cellulose acetate sheet, the main competitors for which appear to
have been Rohm and Haas’ Plexiglas and Vinylite polyvinyl chloride
from Carbide and Carbon Chemicals Corporation. Neither Lumarith
nor Vinylite has been encountered yet in the NASM collection.
Enclosed cockpits were the norm by the late-1930s, and transpar-
ent plastic sheet certainly revolutionized their construction.
However, these soft materials did not meet all window glazing
needs. Laminated safety glass continued to play an important role
in windows that required protection from bird strikes and other
projectiles, abrasion resistance, and activities that required optical
precision such as aiming weapons or focusing cameras.
With the enclosure of cockpits, goggles shifted from the World
War I style with laminated safety glass lenses to metal-framed gog-
gles with ground glass lenses, rubber-framed goggles with one-) cellulose nitrate sheet plasticized with camphor from Douglas World Cruiser
dimethyl phthalate form Sikorsky JRS-1 (FT-Raman), cellulose acetate sheet
ortable Raman), and PMMA sheet from the visor of a United States Navy
U.S. Government work and
ain in the USA.
J. Raman Spectrosc. 2014, 45, 1215–1224
Raman spectroscopic characterization of historic aviation ‘glass’piece plastic lenses, and sunglasses, all of which suggest that
concern about eye protection had been supplanted by optical
performance and manufacturing ease. Study of the NASM collec-
tion showed that glass and plastic lenses were offered a variety of
tints to suit different lighting conditions. Some higher tech light
filtering solutions included ultraviolet-absorbing cerium com-
pounds in glass (A19760039000, A19781761000, A19810332000,
A20010495000, and A20050338000) and polarizing plastic lenses
impregnatedwith aligned herapathite (iodoquinine sulfate) crystals
(A19910412000).[14] The biggest advance in flying goggles of the
World War II period was the Polaroid company’s introduction of
the rubber-framed B-8 google in 1943. The flexible frame offered
a one-size-fits-all option for both aviators and ground crew, and
one-piece mask-like lenses of cellulose acetate plasticized with
low molecular weight phthalates and triphenyl phosphate maxi-
mized visibility and were easily exchangeable. A descendant of
the B-8 goggle called the M-1944 is shown in Fig. 2.[40]12
2Discussion and conclusions
This work is significant for both Raman spectroscopy and the
history of technology. From the perspective of the spectroscopy,
the data shows that Raman is a valuable tool for identifying
polymers, plasticizers, and other compounds encountered in
plastic. As a result of this study, FT- and dispersive instruments
now are used routinely to sort transparent and colorless plastics
at Smithsonian and often can identify major components of inten-
tionally colored or degraded material as well. This article provides
spectra of naturally aged examples of historic cellulose nitrate and
cellulose acetate formulations, and PMMA (Fig. 11). A portable spec-
trometer with 785nm excitation and modified fiber optic probe
facilitated analysis of historic artifacts by bringing the instrument to
objects that are large, challenging to access, oddly shaped, or fragile.
As expected, fluorescence interference was a challenge for
many artifacts, particularly plastics that are colored or contain
polarizing media. Better spectra were obtained for some objects
with a 1064 nm FT-Raman spectrometer, but the instrument is
not portable and has an enclosed sample compartment, which
limited the range of artifacts that could be analyzed. This finding
underscores the potential for portable Raman spectrometers
with NIR excitation. The FT-Raman was preferred for creating
spectral reference libraries and studying multi-component
formulations because of its spectral precision, adjustable resolu-
tion, and lower fluorescent background.
From a historical standpoint, synthetic polymers are one of the
great discoveries of the last two centuries, and it is important that
we record this history through study of our cultural heritage. Plastic
technology evolved rapidly in the early 20th century, and experi-
mentation with formulations and processing methods was com-
mon. In this article we have explored that evolution by looking at
the symbiotic relationship between early plastics and aviation.
While archival research offers a valuable view of what was available
at the time, visual observation of artifacts aided by non-destructive
Raman and XRF spectroscopies shows definitively what was used.
Conversely, a spectrum can record an artifact’s composition, but it
may reflect a mixture that has become obsolete or forgotten, and
include alteration products that formwith age. A clear historical un-
derstanding of the technologies that were possible for a particular
artifact is key for reliable spectral interpretation. For these reasons,
it is vital that historical and scientific analytical approaches be ap-
plied to cultural heritage studies in unison.J. Raman Spectrosc. 2014, 45, 1215–1224 Published 2014. This article is a
is in the public doFinally, this study of plastics in early aviation demonstrates a
way in which Raman spectroscopy can be harnessed to further
research that is significant to the field of history. Studies of
cultural heritage benefit from a significant research question.
Here, Raman spectra are interpreted in the broader context of
an important technological evolution rather than as individual
objects. These early window materials were so successful that
we now take them for granted, but they solved vital public safety
issues and are the origin of important descendant technologies
including personal electronics (e.g. computers, tablets, mobile
phones), structural laminated glass used in architecture, and
explosion and projectile proof glass. Here we have demon-
strated how Raman spectroscopy contributed to a preliminary
but quite robust technological history.
Acknowledgements
This research was funded by the National Park Service and the
National Center for Preservation Technology and Training (Grant #
Mt-2210-10-NC-10). Its contents are solely the responsibility of the
authors and do not necessarily represent the official position or
polices of the National Park Service or the National Center for
Preservation Technology and Training. The B&W Tek MiniRam II
spectrometer was purchased with a generous gift from Judith
Cherwinka. Gary Gordon of NASM and Don Williams, emeritus of
MCI, fabricated two iterations of the adjustable working distance
adapter. Marion Mecklenburg, emeritus of MCI, cast the cellulose
nitrate sheet we used as a reference. Liz Garcia, Jennifer Stringfellow,
Malcolm Collum, Lauren Horelick, Jeremy Kinney, and Ellen Folkama
facilitated access to the NASM collections. Russ Lee, Chris Moore,
Robert Koestler, Paula DePriest, Molly McGath, and Dawn Planas
offered valuable discussion and analytical support.References
[1] H. S. Wolko, in The Wright Flyer: An Engineering Perspective (Ed: H. S.
Wolko), Smithsonian Institution Press, Washington, D.C., 1987, pp.
97–106.
[2] Fort Wayne Sentinel, July 2, 1912, 2
[3] L. F. E. Coombs, Control in the Sky: the Evolution and History of the
Aircraft Cockpit, Pen and Sword Aviation, Barnsley, United Kingdom,
2005.
[4] Flight, March 26, 1915, 218.
[5] Flight, July 16, 1915, 522.
[6] Flight, June 17, 1932, 540.
[7] F. Clymer, Treasury of Early American Automobiles 1877–1925, Bonanza
Books, New York, 1950.
[8] Sunset 1914, 33, 1022.
[9] E. R. Weidlein, New York Times, March 26, 1939, 56.
[10] W. L. Morgan, Ind. Eng. Chem. 1931, 25, 505.
[11] O. Madden, G. Gordon, K. Cullen Cobb, A. M. Spencer, J. Raman
Spectrosc., 2014, 45, 1318.
[12] V. Stannett, Cellulose Acetate Plastics, Temple Press Limited, London,
1950.
[13] J. Tsang, O. Madden, M. Coughlin, A. Maiorana, J. Watson, N. C. Little,
R. J. Speakman, Stud. Conserv. 2009, 54, 90.
[14] E. H. Land, J. S. Friedman, United States Patent 1,918,848, Patented July
18, 1933, (Application Serial No. 358,288 April 26, 1929).
[15] C. Paris, C. Coupry, J. Raman Spectrosc. 2005, 36, 77.
[16] M. J. Neufeld, A. M. Spencer, Smithsonian National Air and Space
Museum: An Autobiography, National Geographic, Washington, D. C.,
2010.
[17] R. D. Friedel, Pioneer Plastic: The Making and Selling of Celluloid,
University of Wisconsin Press, Madison, 1983.
[18] E. Benedictus, Glaces et Verres 1930, 18, 9.
[19] J. Wilson, J. Soc. Glass Technol. 1932, 16, 67.U.S. Government work and
main in the USA.
wileyonlinelibrary.com/journal/jrs
3
O. Madden, K. C. Cobb and A. M. Spencer
12
24[20] E. Benedictus, United States Patent 1,098,342, Patented May 26, 1914
(Application Serial No. 537,557 filed January 11, 1910).
[21] Flight 1916; 8, 66.
[22] E. Benedictus, United States Patent 1,128,094, Patented February 9,
1915 (Application Serial No. 576,356 filed August 9, 1910).
[23] E. Benedictus, United States Patent 1,182,739, Patented May 9, 1916
(Application Serial No. 633,128 June 14, 1911).
[24] E. Benedictus, United States Patent 1,206,656, Patented November 28,
1916 (Application Serial No. 712,732 filed August 1, 1912).
[25] J. C. Wood, British Patent 9,972, Issued 1905 (Application number
GBD190509972 19050511, Priority number: GBT190509972 19050511).
[26] Specifications for the Uniformof the United States Army. Special Regulations
No. 42, U. S. War Department, Washington, DC, 1917.
[27] E. R. Weidlein, Ind. Eng. Chem. 1939, 31, 563.
[28] G. J. Esselen, J. Chem. Educ. 1933, 10, 585.
[29] J. L. Meikle, American Plastic: A Cultural History, Rutgers University Press,
New Jersey, 1995.
[30] G. B. Watkins, W. Harkins, Ind. Eng. Chem. 1933, 25, 1187.
[31] G. B. Watkins, J. D. Ryan, Ind. Eng. Chem. 1933, 25, 1192.wileyonlinelibrary.com/journal/jrs Published 2014. This article is a
is in the public dom[32] W. Clark, Ind. Eng. Chem. 1939, 31, 507.
[33] “Sikorsky JRS-1,” SmithsonianNational Air & SpaceMuseum. Online col-
lections database. Available online at: http://airandspace.si.edu/collec-
tions/artifact.cfm?id=A19610112000, Accessed January 5, 2014.
[34] “Science: Softness for safety,” Time, April 10, 1939.
[35] Aero Digest 1937, 31, 24.
[36] S. Hochheiser, Rohm and Haas: History of a Chemical Company,
University of Pennsylvania Press, Philadelphia, 1986.
[37] C. B. Wooster, J. Chem. Educ. 1942, 19, 430.
[38] W. Davis, Sci. News Lett. 1941, 40, 314.
[39] A. Klemin, Aero Digest, 1938, 33, 59.
[40] M. J. Prodger, Vintage Flying Helmets: Aviation Headgear Before the Jet
Age, Schiffer Publishing, Atglen, Pennsylvania, 1995.
Supporting information
Additional supporting information may be found in the online ver-
sion of this article at the publisher’s web-site.U.S. Government work and
ain in the USA.
J. Raman Spectrosc. 2014, 45, 1215–1224