feature article Application of Fourier Transform Infrared Spectroscopy to Chemical Systems J A C K L. K O E N I G

Department of Macromolecular Science, Case Western Reserve U~,dversily, Cleveland, Ohio 44106

The advantages of Fourier transform infrared spectroscopy (FTIR) over dispersive infrared spectroscopy are outlined. The use of data processing with specific reference to difference spectroscopy is discussed in light of these advantages. Applications of FTIR to identification, quality control, and q u a n t i t a t i v e infrared analysis are imlicated. The use of difference spectroscopy to improve the quality of the spectra of the desired c o m p o n e n t s is illustrated for a variety of chemical systems. The role of FTIR is t h u s illustrated in the analytical chemical laboratory. Index Headings: Infrared spectroscopy; Fourier transform spectroscopy; Polymers; Biological molecules; Background stripping t e c h n i q u e .

INTRODUCTION

The application of Fourier transform techniques through the use of the interferometer in infrared spectroscopy is many years old. 1 Recently Fourier transform infrared spectroscopy (FTIR) has begun to be extensively used in the analytical laboratory. This new interest can be attributed, in part, to the increased availability of "turnkey" or complete commercial instruments. However the advent of software which allows mathematical manipulation of the spectral data is also important. As a result of this data processing capability, advantage can be taken of the improved signal-to-noise ratio available from FTIR yielding greater spectral sensitivity. Thus the applications of infrared spectroscopy to chemical problems are substantially enhanced. It will be the purpose of this article to demonstrate the use of the data processing capability for the solution of infrared spectroscopic problems. In this manner it is hoped that the principles enunciated and the examples chosen will serve as a guide for the development of new applications of FTIR and wider use in the analytical laboratory in industry, academia, and government. I. A D V A N T A G E S O F F O U R I E R INFRARED SPECTROSCOPY

TRANSFORM

A. E n e r g y T h r o u g h p u t . Like most spectroscopic techniques, the limiting aspect of infrared spectroscopy is the available energy per unit time. The infrared sources are black bodies and as such are hot wire emitters. Their lifeReceived 17 March 1975. Volume 29, Number 4, 1975

times are inversely proportional to their operating temperature. Thus, the amount of radiant energy is severely limited, and so the use temperature is limited. Additionally, infrared detectors such as thermoeouples and pyroeleetrie devices have a relatively high internal noise compared to phototubes; therefore, an appreciable infrared signal is required for operation. On this basis, infrared spectrometers should have a high energy throughput since they, in particular, suffer from energy-limited sources and noiselimited detectors. Monoehromators, using either gratings or prisms as dispersers, as the name implies, allow the observation of a narrow, predetermined, nearly monochromatic, frequency domain. Spectra are recorded by scanning the desired frequency range at successive resolution intervals. The narrow openings of the entrance and exit slits, required to achieve good resolution, limit the energy throughput, so the detector views only a small frequency portion of the energy emitted by the source. Measurable signals are achieved by increasing the time of measurement per frequency interval. As the resolution is increased by narrowing the slits, the scanning rate is substantially decreased. Geometric dispersion in monoehromators results in throwing away most of the valuable energy. The use of interferometers in infrared spectrometers produces two principal advantages--high energy throughput and multiplexing. The principal disadvantage is that an interferogram is produced rather than a spectrum and little information is available from an interferogram without further processing. A Miehelson interferometer (Fig. 1) consists of two mirrors--one stationary and one movable (at right angles to each other) and a beam splitter bisecting the angle between the two mirrors. No slits limit the incoming energy, and the energy throughput depends on the size of the mirror. In general, the energy throughput is 80 to 200 times greater than for a dispersion instrument depending on the resolution. B. Multiplex Advantage. The incoming light is divided at the beam splitter and is reflected to the two mirrors. Upon reflection and recombination at the beam splitter, the intensity of the light depends on the differences in the path lengths of the divided light beams. Constructive interference occurs when the light paths traveled are the same; and for a white light source, a strong center burst of energy is viewed by the detector. The interferometer is scanned by driving the movable mirror. Since the interference depends on the path difference and the frequency APPLIED SPECTROSCOPY

293

Fixed Mirror

i

Beam Splitterd Movable Mirror

i. ~

Detector

Source MICHELSON INTERFEROMETER FIG. 1. Diagram of Michelson interferometer. of the light, all of the frequencies undergo interferences differently as a function of the path length change. Consequently, each frequency has a particular interference pattern independent of all other frequencies. The detector views all of the frequencies all of the time during the mirror displacement giving rise to what is termed the multiplex advantage. The multiplex advantage is proportional to the number of frequency elements scanned and is given by (N) m where N is the number of frequency elements. For a resolution of 1 cm-1, and a 4000 em -1 scan, the multiplex advantage is 63. C. Speed. The intensity falling on the detector is given by

I(x) =

F

B(~,) cos (21rt,x) d~,

(1)

where I(x) is the intensity at the detector, B(v) is the spectrum of the source, and x is the mirror displacement. An example of an interferogram, I(x), is shown in Fig. 2. The spectrum, B@), is obtained by calculating the Fourier transform pair of the above equation

B@) =

I(x) cos (21r~x) dx

(2)

o0

Coupling the advantages of high energy throughput with the multiplex benefit, the scanning speed of the mirror can be rapid, and so the interferogram is obtainable in seconds. This speed represents another advantage of the F T I R system and is particularly important when multiple co-added spectra are desired. D. Computer Availability. Although the interferogram contains all of the spectral information, it is not readily discernible in the form obtained. The conversion of the interferogram to a spectrum requires the use of a computer. This requirement limited the acceptance of interferometrie spectroscopy to the very rich (costly computer time) and the very patient (2 or 3 days of processing). Although the interferogram could be quickly recorded, it often took days of computer time to process the data and be rewarded with a spectrum. But in 1965, the CooleyTukey algorithm was published, 2 and the computation time for Fourier transforms was reduced by several orders of magnitude. Simultaneously, minicomputers were developed, and the coupling to the Michelson interferometer of the computer resulted in the capability of on-line short 294

Volume29, Number 4, 1975

time Fourier transformations. With this final advance, turn-key systems were made available and FTIR moved into the analytical laboratory to stay. With the subsequent development of software, the computer could be used to advantage for processing the spectral data in refined ways. Infrared spectroscopy as an analytical tool benefited from these developments as researchers began to utilize the advantages of speed, sensitivity, and data processing offered by F T I R for the solution of chemical problems. E. Multiscanning Capability. One of the first uses of the computer was for co-adding of successive interferograms to increase the signal-to-noise ratio. Since the signal increases with the number of scans and the noise with the square root of the number of scans, the signal-to-noise ratio should increase with the square root of the number of scans. With the time for driving the mirror only a few seconds, the increase in signal-to-noise (S/N) ratio is fast and, since no increase in computational time is involved, easily accomplished. In order to co-add interferograms, precise positioning and activation of the data accumulation system is required. With the Digilab FTS 14, this is accomplished through the use of a second internal interferometer using a laser and a white light source. The laser ensures accuracy of position and the center burst of the white light source activates the data system. We have verified the theoretical predictions for improvement in signal-to-noise ratio by using a highly energy-limited system--carbon-black filled rubber in a KBr pellet. Fig. 3 shows the increase in S/N ratio as a function of the number of scans. The increase is apparent even after 800 scans. With appropriate software, which is currently available with the Digilab FTS 14, this accumulation process can be interrupted to examine spectral results and continued to the number of scans desired. In this manner, the minimal number of scans for a predetermined S/N ratio can be easily found for samples with different levels of energy transmission without loss of time or data. Further improvement in the "esthetics" of the presentation of the spectra can be accomplished by using the computer software package to "smooth" the data. F. Linear Absorbance Scale Expansion. Another

INTERFEROGRAM

~A

FIG. 2.

Interferogram.

(# SCANS)~/2 0 40--

312

identification and calibration procedures are considerably simplified. There is probably no better way to compare two digital spectra than with a computer to detect and amplify spectral differences.

/

ox / /

II. A P P L I C A T I O N OF D A T A P R O C E S S I N G CHEMICAL SPECTROSCOPY

30x

o

~

nc 20-

o

o

//// / /)~

o z

6

/ / ix~

Z

0

0

/

260

A. D i f f e r e n c e S p e c t r o s c o p y - - R e q u i r e m e n t s . Difference spectroscopy provides a sensitive method for detecting small changes in samples, whether in composition or physical state. Common features in the spectra cancel, and bands which are recorded can be interpreted in terms of differences between the samples. The requirements have been stated well 3.

"Speetrophotometer requirements for difference measurements are considerably more stringent thaT] for conventional techniques. Measuring small differences between two solutions with large initial absorbances requires an instrument with high photometric accuracy, expanded ahsorbance ranges, good resolution and low stray radiation energy (SRE) or stray light."

'7 m¢

TO

480

600

860

# SCANS

FIG. 3. Change in signal-to-noise ratio with mmfi)er of scans. immediately obvious but invaluable use of the computer system is the storage and scale expansion of the data either frequency-wise or intensity-wise Once the spectra are computed and in digital form, the data can be plotted in any desired manner. The most useful aspect of this result is magnification of the desired absorbanee signal to the limit of the signal-to-noise ratio. This magnification, without reseanning is completely linear. With dispersion spectrometers, the linearity of the comb, in the particular transmission region being examined, was always questionable for no more reason than the possible accumulation of dust. With the computer, the magnification of the signal is linear, although the computer cannot verify that the signal is the proper o n e - - t h a t is still the domain of the operator. However, the computer can be made to "read" the absorbmme value thus removing the possibility of subjective readings by the operator. G. S t o r a g e . Finally, the memory banks of the computer can be used for storage of spectra as well as software programs. When additional memory is required, fixed head discs and magnetic tape milts can be added to the computer to store spectra. With the cost of paper increasing, it may soon become cheaper to store spectra on magnetic tape rather than as hard copy. Additioimlly, the spectra can be recalled for use or replotting at any future time. How many times have we wished we could redo the spectra; now we can! The first use of the storage capacity of the computer was the development of double beam capability by ratioing the spectra obtained in the sample to that of an empty reference beam. Thus, the interference of atmospheric absorbanee like water and CO2 can be minimized but not eliminated. More importantly for the analytical spectroscopist the spectra of the "control" sample or the "standard" sample can be stored subject to immediate recall for comparison with the "unknown" sample. In this manner,

withoutrescanning.

Erickson and Bramston-Cook 3 show that a stray light level of 0.1% produces only a 2.1% error for an absorbance of 2.0, while the same stray light level of 0.1% produces a 10.1% relative error in the measurement of the difference between an absorbance of 2.0 and 2.1. These observations, although applied by the authors of uv spectroscopy, explain the limited success of difference spectroscopy using conventional dispersion infrared spectroscopy. Scattered light is always a problem in monoehromators while it is minimized in interferometers. Double beam spectroscopy has been invaluable in conventional spectroscopy but the uses are limited to dilute solutions with matched cells and solid samples where suitable wedges can be produced to match thicknesses. Absorbanee scale expansion is limited with dispersion spectroscopy due to the loss of energy in the two beams. Increased noise usually results from highly absorbing samples in the reference. With a computer, the difference spectra can be expanded to enhance readability and the background can be placed at any desired level-usually zero by subtraction of a constant factor throughout the frequency range. Perhaps the most important benefit to F T I R is the fact that any two stored spectra can be compared at any time separation between the actual measurement of the spectra and their subtraction, whereas with dispersion spectroscopy without digital capability only two specific samples can be compared and only at the time of measurement. B. D i f f e r e n c e S p e c t r o s c o p y - - M e t h o d . Difference spectra can be accomplished using a digital absorbanee subtraction method. The absorbanee procedure to be used will be illustrated mathematically. For spectrum 1, the following equation can be written for each frequency v

A~, = A;,. + A~, + A~,

(3)

where A ~ = total absorbanee of all components at frequency v in sample 1, A.~ = absorbance of component x, v A~, = absorbanee of component y, and A+, = absorbance of component z. Similarly, it can be written for spectrum 2 7'~ = Ax~ + Ay~ + . ~ APPLIED SPECTROSCOPY

(4) 295

I t is desired to subtact A ) , from A~, as A," = A L

-

at a frequency vv characteristic of component y (formerly overlapped by x)

l c A ~ = ( A ~ -- lcA~,)

(5) + (Ay, -

leAve) + ( A , , -

kAy2)

where A,' = absorbance of subtracted spectra at frequency v and k = adjustable scaling parameter. Let us consider several examples of difference spectra. 1. P a t h Length Differences. If the path lengths are different, an internal thickness band or solvent band is sought and absorbances are scaled to a predetermined thickness. A;l~ -- lczA;~2 = 0 ( ~ = internal thickness band)

(6)

tel = A h / A z~ = alClll/a~C21~

If an internal thickness band is selected, it is required that a~C~ = c~2C2;so

(7)

lcz = l~/l~

where 12 can be considered a standard thickness. In this way, solvent interferences can be eliminated as well as nonanalytieal differences associated solely with amount of sample being viewed by the beam. This scaling is particularly important for solid samples such as powders, films, or K B r pellets where matching of amount of samples in the beam is nearly impossible and was a limiting factor for classical dispersion measurements. The difference spectra can be written as As = (A~i - /c~ A~) + (A,~ -- i:~ A~)

(s)

+ (A~, - k~ A.~) If any of the x components are not present in sample 2 (C~ = 0), then A,= = 0 Ao = A~ + (Av, - let As2) + (A,, -- ]c~ A,2)

(9)

and for a specific frequency of absorbance uniquely due to component x; A ? = A; x

(v~: n ; ~ = n : • = 0; C,~ = 0)

(10)

For this simplified ease, the difference spectrum gives a direct measurement of the absorbance of component x with appropriate correction for differences in thickness. 2. Interfering Absorbances. When more than one component absorbs at the same frequency, elimination of the interfering absorbanee is desirable to isolate a particular absorbanee due to a component. For this ease, one selects a frequency region due to the interfering component. For example, A , = (Ax, -

1,,A,2) + (Ay, -- lcAy2) + (Az~ - lcAz2)

(11)

The removal of the absorbance due to component x at all frequencies can be accomplished using the following subtraction criterion A ;~ -

lcxA ;~ = 0

(~2)

]c.A~) (,y:A~, = 0~A~, # Aye)

A : ~ = (A~,~ -

and any interfering or overlapping of absorbance due to component x has been removed. A similar second difference spectrum can be computed to eliminate any additional interferences to isolate absorbance of component desired. 3. Spectral Separations. In the above simple manner, spectral purifications can be accomplished and the spectra of "pure" components obtained. Subtracting interferences due to absorbanees of solvents is an obvious application of the difference spectra technique. But probably an area with greater potential is the separation and purification of components not separable by physical or chemical techniques. Table I lists a number of these areas where spectral separations of this type would appear to be useful. 4. Molecular Changes and Interactions. One of the purposes of obtaining difference spectra is to isolate molecular changes and interactions. When chemical changes produce new functional groups, new absorbance bands appear which may be weak and overlapped by bands arising from similar chemical types present in the original sample. Difference spectra can be produced, eliminating the interference of the original sample. Scale expansion of the difference spectra allows detection of small chemical changes. Interactions arising from chemical or physical effects such as hydrogen bonding, hydrophobie effects, and intermolecular forces generally produce either a shift in the absorbing frequency or a change in the absorptivity of the bands. Either of these spectral differences can be isolated using the difference spectra technique. The shift in frequency with physical interactions makes it impossible to subtract the affected absorbance bands while the unaffected bands cancel. The difference spectrum will show a positive denotation followed by a negative denotation similar to the appearance of a differentiation df a band with a maximal or minimal value. When changes in absorptivity occur, residual positive or negative absorbance will appear in the difference spectra for the affected modes. Thus interaction effects can easily be detected using difference techniques. TABLE I. FTIR spectra p u r i f i c a t i o n t h r o u g h d a t a processing.

1. Chemical rx ~ - - - - - - ' - - - - - - ~ reacted unreaeted eolnponents ~ s o l v e n t solute

2. Solu t ions ~--------~'----~

interaction effects

3. Iteterophase systems

crystalline

4. Surface effects

amorphous

adsorbed species >surface functionality ----------------------*bulkcomposition

5. Mechanical effects ~ -

_......_..__....)orientation )geometric isomerism strMn effects

los = A ,~/ A ,~

so at all other frequencies As = (Ay I -

296

/cxAy2) @ A z t -

Volume29, Number 4, 1975

6. Composites ~ - - - - - a'xAz 2)

(13)

(14)

"

interface

qnatrix eomponents~filler

5. Applications. a) Identification of samples. The uses of difference spectroscopy are manifold. An overview of the utility is presented in l?ig. 4, where various applications are summarized. When the difference spectra refleet no bands above the noise level, the spectra are identical, which implies that the samples are identical to the limits of the sensitivity of the spectra. Even in courts of law this spectral comparison would constitute an unequivocal identification and evidence of the identity of samples. With the storage capability of FTIR, many chemically complex systems can be compared to known systems such as drugs, natural products, additives, and other chemicals. The potentials of this technique are numerous for industrial espionage of competitors' products. Unambiguous identification of one's own product can be achieved by seeking an identifying component. The future holds the possibility of purchasing tapes containing digital FTIR spectra of series of compounds such as plasticizers, antioxidants, flame retardants, etc. in the same maturer as is presently available for catalogues of spectra on cards or microfilm. b) Quality control. Another use for difference spectra is quality control by either a chemical producer or user. When the spectra of an acceptable product, be it polymer with the associated receipe package or a mixture such as paint, adhesives, or solvents, which is recorded and stored, it can be compared with any subsequent production run.

The magnitude of the absorbance differences can be determined, which gives rise to an acceptable product meeting predetermined specifications. Whenever a subsequent product is compared with the standard product, a decision can easily be made about the acceptability of the current product. Most importantly, when the difference spectra deviate from acceptable specifications, one can identify the component which is problematic and the extent of concentration variation producing the unacceptability of the product. Hence, using FTIR for quality control in some cases, not only yields a "go or no go" decision but also directs attention to the problem areas for negative decisions. For those samples requiring a short sample preparation time, rapid analysis can be attained. In fact, for specific cases, software programs can be written to automate the spectrophotometric procedure and print the results. c) Quantitative infrared analysis. Difference spectra can be used for quantitative analysis as previously indicated with proper calibration. The advantage over ordinary FTIR spectra is the isolation of analytical absorbances allowing computer expansion of the absorbanee scale. This gives higher sensitivity. Additionally, separation of components is not required since a number of interfering absorbances can be removed. A fringe benefit of the subtraction is the possibility of detecting unexpected contaminants or interchain effects.

SPECTRUM

SPECTRUM

OF

OF

UNKNOWN

STANDARD

I

IFFERENCE

J

I

J~ _ i

i

~A=O

I IDENTIFICATION

il

I

QUALITY CONTROL

l

I

11

Jk

ft L V

/k_

÷ _

k/

3A = a L~,C

~A

~L,~C

$

I

i

QUANTITATIVE DETERMINATION

II -I MOLECULAR CHANGES II

FIG, 4. Applications of infrared difference spectroscopy.

APPLIED SPECTROSCOPY

297

III. FTIR SEPARATIONS THROUGH DATA PROCESSING A. Separation of Reacted and Unreacted Species. 1. Irradiation Damage of PolyethyleneA The infrared spectrum of a polyethylene film irradiated for 50 h in N2 is shown in Fig. 5 compared with the spectrum for an unirradiated sample of polyethylene. There are some obvious spectral changes with irradiation, These changes are mainly the appearance of absorbance bands at 1716 cm-', attributed to the stretching vibration of ketonic earbonyl i

,..A°,A,,. . ,

,,

"' JJ i , . . .

.

POLY|THYL|N

2400

,

I

1900

1400

'

900

'

CM"1

I

FIG. 5. Infrared absorbance spectra for polyethylene film irradiated in nitrogen. Bottom, unirradiated polyethylene; middle, polyethylene irradiated for 50 h in nitrogen; top, difference spectrum (irradiated polyethylene-polyethylene).

groups and at 965 c m -1, due to the C - - H out-of-plane deformation of trans-vinylene groups, trans R C H ~ - C H R ' . The effects of radiation exposure in nitrogen on polyethylene can more easily be seen in the difference spectrum presented at the top of Fig. 5. The spectral changes that occur upon irradiation are more easily seen in Fig. 6 where the difference spectra for polyethylene films irradiated in both nitrogen and air for 50 and 100 h. The spectra represent changes that occur upon irradiation for an equivalent amount of polyethylene. In this way, quantitative comparisons could easily be made among various samples. In the difference spectrum, absorbance bands above the baseline reflect an increase in a particular chemical species while those bands below the baseline reflect a decrease in an absorbing species due to the irradiation process. The increase of the absorption bands at 1716 and 965 cm -1 due to the irradiation process is evident in the difference spectrum. Moreover, ill the difference spectrum there is a decrease in absorbance of the vinyl end groups R - - C H - ~ CH2, at 1642, 991 and 909 cm-k These effects--namely, the increase of ketonic earbonyl and trans-vinylene double bonds and the accompanying decrease of R C H = C H 2 groups--have previously been reported from infrared studies of irradiated polyethylene. 5 The prominent changes that occur upon irradiation are also presented in Table II along with the tentative band assignments. From the difference spectra, changes can be seen that were not evident before subtraction. A band at 1410 cm -1 appears and increases in intensity as the irradiation conditions become more severe. This band has previously been assigned to a deformation influenced by an adjacent carbonyl group O R - - C H 2 - - C - - C H 2 - - R ' . An increase in methyl content in

1716 1410

POLYETHYLENE

965 I

llal

100 HR AIR

50 HR AIR

/ I/

| 0 0 HIE N 2

SO H i N 2

1742 1642

1750

I

1500

I

1250

!

1000 909 l

750

cM I

,,M

FIG. 6. Difference spectra for polyethylene fihns irradiated ill both nitrogen and air for 50 and 100 h. 298

Volume 29, Number 4,

1975

TABLE II. Observed c h a n g e s in infrared spectra of irradiated p o l y e t h y l e n e films.

Frequency (cm- ~)

Assigmnents

Intensity change upon irradiation~

O 1742

R -C---H ~(C=O) O

1716 1642

R-- .C-- R' u(C=()) R--CH=CH2 ~(C=C) O

1410 1378 1368 1353 1308 1261 1131 1068 991 965 909 800

R--CII2-..C -CH2---R' ~(CH2) R--CH3 &~(Ctt3) %~(CH2)--amorphous phase 3,,~(CH:) -amorphous phase 7w(CII~)-amorphous phase 'y,~(CH~)--amorphous phase ~,(C--O) ~,(C--O) R--CH~CII~ tra~ls-R--Cti=CH -I1' R -Ctt=CH2 Amorphous phase

II

+

II + + -

-

+ + +

The symbols + and represent an increase and decrease in band intensity upon irradiation, respectively. -

the polyethylene films also occurs, as reflected by an increase of the symmetric methyl deformation mode at 1378 cm-L Bands at 1131 and 1068 cm -~ increase upon irradiation and are indicative of C - - O stretching v i b r a t i o n s / T h e amorphous phase in polyethylene--specifically, as refleeted in the 1368, 1353, 1308, 1261, and 800 cm -1 vibrational bands--are affected by irradiation in preference to the crystalline phase. Comparison of the difference spectra for polyethylene films irradiated in nitrogen and air shows that the decrease of aldehydic carbonyl and vinyl groups and the increase of trans-vinylene double bonds occur approximately to the same extent in nitrogen or oxygen. However, with irradiation in air there is a more drastic change in the absorption bands associated with the ketonic carbonyl groups, the methyl end-groups, and the amorphous phase. These chemical changes are indicative of oxidative degradation and the accompanying chain scission. 2. Oxidation of Polybutadiene/ The useful life of many polymeric systems is limited by their susceptibility to oxidation degradation. Rubbers are particularly vulnerable to oxidation, and as a consequence considerable research has been carried out in an effort to determine the mechanism of this oxidative chemical reaction/ Infrared has been used to study this chemical process but has been limited because the reaction must proceed to a substantial degree before detectable absorbances appear. With the difference spectra technique, considerable improvement results. In Fig. 7, the infrared absorbance spectra of a sample of unoxidized polybutadiene and a sample held in air at 30°C for 10 h. Using the band at 740 cm -1 as the basis for subtraction, the scale expanded (25 × ) difference spectrum is also shown. The decrease in the band at 300

em -1 is associated with the cis to trans isomerism, which is also reflected by the increase in the band at 975 cm -1 assignable to the trans portion of the molecule. The band at 1065 cm -I is probably associated with an oxidation product containing a C--O band, but the specific nature of the functionality has not been determined at this time. The changes occurring as the time of reaction proceeds to 235 and 640 h are shown in the difference spectra of Fig. 8. The appearance of bands due to hydroxyl (3300 cm -1 carbonyl (1770, 1720, and 1700 cm -1) are now obviously reflecting the formation of these oxidation products. 3. Vulcanization of Rubbery F T I R can be used to study the changes occurring in complex chemical reactions induced by light and heat such as vulcanization. The subtraction technique allows the elimination of the unreacted components and the isolation of the reacting sites and products. Vulcanization reactions involving several compounds and a series of reactions can be monitored, for example. A film of cis-1,4-polybutadiene sample prepared with 5 parts mercaptobenzothiazole-disulfide (MBTS), 5 parts sulfur, and 2 parts by weight ZnO was cast on K B r windows using chloroform as a solvent. The chloroform was allowed to evaporate completely. Then the film was heated at 150°C for 1 h. The infrared spectra of the unvuleanized (u) and vulcanized (v) samples are presented in Fig. 9. The unvulcanized spectrum was digitally subtracted from the vulcanized by using the 3065 cm -~ band as an internal standard. The band at 3065 cm -~ is assigned to the C - - H stretching mode of H on the cis double bond. The subtracted spectrum is also shown in Fig. 9. Bands showing a positive displacement reflect an increase due to formation of a new product by vulcanization. Bands showing a negative displacement reflect a decrease due to the vulcanization. A strong decrease in the 3007 cm -~ is attributed to isomerization of the elastomcr, cis-1,4-Polybutadiene on vulcanization in the presence of air gives up to 25 % of the trans isomer. We also see a decrease in the 2935 cm -~ and the 2915 cm -1 bands. These bands are indicative of a decrease in the allylic hydrogen which is a to the double bond. A decrease in the 1650 cm -t band is indicative of a 1065

¢-0

-- .j, ~

~975 I

C-H

T RAhlS C-H CI5

74O

3007

A. OXIDIZED

3500 |

Pll

27oo 20o0 a

.

•

•

8oo I

I

I

i

n

n

ii

FIG. 7. Oxidation of polybutadiene for 10 h. APPLIED S P E C T R O S C O P Y

299

30°¢

O07

OXIDATION

STUDY

10 HR

1065 C - O , I 97S ¢=1:1.11_

¢'H~ ¢lS

23S HR

640 HR

sorbance does not occur, the water spectrum can be subtracted. For biological systems where water is the only interesting solvent, this improvement can be utilized and demonstrated for protein solutions. The aqueous solution infrared spectrum of hemoglobin obtained by F T I R is shown in Fig. 10 (spectrum 2). No useful detail can be observed until the water spectrum (spectrum 1) is st, btracted to yield spectrum 3 where the classical amide I and II modes at 1657 and 1547 cm -~ are observed. A comparison of the infrared absorbance spectrum of hemoglobin as a cast film (less water) is shown in Fig. 11. The secondary structure of protein can be determined using conformation~lly sensitive frequencies. '2 C. S e p a r a t i o n o f S o l u t e s U n d e r g o i n g R e a c t i o n s in S o l u t i o n . 1. Silanes Undergoing Hydrolysis. 13 The problem of obtaining the spectrum of a reactive solute is a further complication. However, in many cases the chemical product occurs only in solution and cannot be isolated. Under these circumstances the need to obtain the spectra of the reaction product is essential to understand the system. Silanes are used as coupling agents to promote adhesion between glass fibers and the polymeric matrix. They are applied to the fiber from aqueous solution where they undergo the hydrolysis reaction: R'-Si(OR)3 + 3H20 ~ R'-Si(OH)3 + 3ROH

UNOX PB

1

I

35OO

I

i

I

2700

I

I

I

2000

I - - I

I

I

800 CM'I

FIG. 8. Difference spectra for oxidation of polybutadiene for 10, 235, and 640 h at 30°C. decrease in the cis isomer. The decrease in the 1466 cm -~ band indicates a decrease in the total quantity of the M B T S and the polysulfides which are formed in the vulcanizate. This result is attributed to their consumption in the formation of the rubber-bound intermediates which leads to a decrease in the total amount of thioureide linkages present. The 1450 and 1433 cm -1 bands are assigned to methylene-bending mode and the deformation of the H - - C - - C bond angle. The intensity decrease is a result of the decrease in the a-methylenic hydrogen with reaction. Using subtraction techniques in F T I R , one can observe the following effects during accelerated sulfur vulcanization: (1) a decrease in the allylic hydrogen which is a to the double bond; (2) decrease in the cis and increase in the trans isomer; (3) decrease in the total amount of free M B T S and M B T P observed in the vulcanizate. These three effects are consistent with the proposed mechanism ~° wherein the active sulfurating agent is a polysulfidic compound. B. S e p a r a t i o n o f S o l u t e f r o m S o l u t i o n . 1. Aqueous Solutions of Biological Molecules. ~1 The removal of interfering absorbance due to solvents is an important application. A classical limitation of infrared analysis has been the study of aqueous solutions due to strong, broad infrared absorbance of water. With F T I R , if the aqueous solutions are examined such that total ab300

Volume 29, Number 4, 1975

In this case, the problem is further complicated by the equilibrium reaction, the presence of the alcohol, and the aggregation or polymerization of the hydrolyzed species. In Fig. 12 the absorbance spectrum (no. 1) of water in a thin cell is shown, while the spectrum (no. 2) of a 5 % solution of 7-methacryloxypropyltrimethoxysilane (UC 174) in water. Spectrum 3 shows the results of subtracting spectrum 1 from 2 and scale expanding the absorbance

v-u

t/'"

.............................. 4._.)

IlJl,l _ #t [

.....................................

I

a2so

~

t,.. .......,

\

_J ,,.k..~_ .., ' \ U \. I

4so CM -1

FIG. 9. Tile vulcanized, unvulcanized, and subtracted spectra of cis-l,4-polybutadiene (100 parts), MBTS (5 parts), zinc oxide (2 parts), sulfur (2 parts).

D. Separation of Interaction Effects in S o l u t i o n . I. Plasticized Poly(vinyl chloride); 4 The infrared ab-

sorbance spectra for three PVC films plasticized with different amounts of DOP are shown in Fig. 14 along with spectra for unplasticized PVC and the DOP plasticizer. It is easy to see the difficulty involved in analyzing the plasticized PVC spectra because of the overlapping spectral bands of PVC and DOP. This task is simplified by appropriate processing of the spectral data for this twocomponent system. 2.

lsoo I

14oo I

moo I

6oo i

c~1""

2.

FIG. 10. Aqueous solution infrared spectrum of henmglol)in obtained with Fourier transform spectrophotometer. 1, absorb~mce spectrum of H~O; 2, absorbance spectrum of aqueous hemoglobin solution, 3, absorbance spectrum of hemoglobin in aqueous phase (spectrum 2 - spectrum 1). A1657

1098

1455 1390

1800

1400

1000

I

I

I

I

FIG. 12. Infrared spectra of T-methacryloxypropyltrimethoxysilane. 1, I)ure writer; 2, 5~ sohltion; 3, subtracted spectra; 4, spectra of neat material. 1721

1089

6

,ill1,I

818

1298

1106

1.

600 C.1Mi.

1547 1455 1400 1308

13

1454 ,1255

I

4 41

,780 760

1017

1800 I

14.00 I

1000 I

600 I

Ch,-I

1. 1698

FIG. 11. Infrared absorbance spectra of hemoglobin. 1, aqueous solution (pH = 4.8).

scale of the resultant spectrum. Spectrum 4 is of the pure, unhydrolyzed silane. The spectral differences are shown in more detail in Fig. 13 where spectrum 1 is the hydrolyzed product and spectrum 2 is the unhydrolyzed pure com. pound. The differences reflect the result of the hydrolysis, particularly the loss of the strong bands at 1190, 1167, and 1089 cm-~ due to Si--O--C motions and appearance of the band at 920 em-~ due to Si--O--H motions.

/

o

1198¢~186

H1634 ,14o91i I *

1800

t

1400

I

1000

I

600

I

cM1

FIG. 13. Comparison of hydrolyzed (1) and unhydrolyzed (2) "r-methacryloxypropyltrimethoxysilane.

APPLIED SPECTROSCOPY

301

Analysis of the plasticized PVC spectra shown in Fig. 14 can be approached by removal of the DOP contributions to the total spectra. The DOP spectrum was subtracted from each of the plasticized PVC spectra, and the results are shown in Fig. 15. The criterion for subtraction was the elimination of the 1601 and 1581 cm-1 in-plane ring vibrations. As can be seen from the difference spectra in Fig. 15, insufficient subtraction of the DOP spectral contribution occurred at several positions, each marked with an arrow, because of changes in DOP band intensities and positions which resulted from interactions between DOP and PVC in plasticized films. The subtraction has enabled the conformationally sensitive carbon-chlorine stretching region to be examined. There is a decrease in frequency for the SHe modes from 695 to 602 em-I as the plasticizer content is increased. This frequency decrease reflects an increase in a more stable conformation in the amorphous region. The spectral contribution from the PVC was also removed from the original plasticized spectra of Fig. 16. These difference spectra, shown in Fig. 16, were obtained by subtraction of the PVC spectrum of Fig. 15 from the

3200

I

2600

I

1800

1200

I

600

I

I

¢M'1

Fro. 14. Infrared absorbance spectra of plasticized poly(vinyl chloride) films.

plasticized PVC spectra. The accurate subtractions of the PVC bands that are obtained demonstrate the essentially identical PVC crystalline and amorphous spectral components just mentioned. The frequencies and relative intensities for the DOP bands as a function of plasticizer content are listed in Table III. The DOP bands that change in position are shown in Table IV. The C ~ O and C--O modes, especially the C-~O group, are affected by plasticization while the methylene and methyl modes are essentially unaffected.

E. Separation of Heterophase Systems. 1. Crystalline Vibrational Spectrum of trans-1 ,g-Polychloroprene.15 One of the more formidable problems occurring in the theoretical study and interpretation of the vibrational spectra of semicrystalline polymers concerns the assignment of crystalline and amorphous bands. We have been able to obtain the "purified" spectrum of the preferred conformation of a predominantly trans-l,4-polychloroprene (trans-CD) polymer from the semicrystalline polymer. The infrared spectrum, designated 1, of this sample, was obtained at room temperature on a Digilab model FTS-14 Fourier transform spectrophotometer. The sample was then heated to 80°C for 15 rain in the spectrophotometer and another infrared spectrum, designated 2, was obtained under the same conditions. These two samples should differ only in the degree of crystallinity attained (barring the possibility of slight oxidative degradation which would be readily observable in the carbonyl region of the spectrum if present in significant concentration). Finally, the sample was allowed to cool to room temperature and another spectrum was recorded and stored after 120 rain. An absorbance subtraction routine was employed to obtain the crystalline vibrational spectrum of trans-CD. In Fig. 17 the absorbance spectra 1 and 2 are shown together with the difference spectrum 3 obtained by subtraction of the less crystalline spectrum 2 from the more crystalline spectrum 1. The subtraction was performed on the basis of elimination of the amorphous component of the semicrystalline trans-CD spectrum with the resultant

1581 pop

_

~

.-

._j ~ _ J ~ J L J

Pyc/o.$oooP -- ooll

.~____~II'V¢/O,20OO41-- oo~ P J ~

_P.V¢/O.O DOP $ --DOlt, PVC

3200

I

2600

I

1800

I

1200

I

600

I

¢M*1

FIG. 15. Differencespectra:plasticized poly(vinyl chloride) spectra - plasticizer spectrum. 302

Volume 29, Number 4, 1975

1287 1123 1072

~

pv

172S

112610S4

C/0210DOP-- PVC .

J 1279

1723

]1211

1280

1722

1128

3200

2600

1800

1200

600

I

I

I

I

I

CM'I

FIG. 16. Difference spectra: plasticized poly (vinyl chloride) spectra - poly (vinyl chloride) spectrum. TABLE I I I . DOP frequencies and relative i n t e n s i t i e s for plasticized films.

Relative peak intensity ~ Fre}uency cm- ')

DOP

PVC/ 0.50 DOP

PVC/ 0.20 DOP

PVC/ 0.05 DOP

1730 1601 1581 1464

11.2 0.82 1.0 3.2

13.9 0.84 1.0 3.3

15.1 0.81 1.0 3.0

16.1 0.85 1.0 2.9

1381 1287 1274 1123 1072 1040 743

1.9 8.9 9.1 6.0 5.3 I .4 2.3

2.0 11.4 11.2 6.3 5.3 1.2 2.6

1.9 11.8 11.4 5.8 4.7 0.9 2.6

2.0 12.5 11.9 6.1 4.3 0.8 2.7

Tentative assignment

sample 1; AAt is the absorbance of amorphous trans-CD component in sample 1 ; A c I is the absorbance of crystalline trans-CD component in sample 1; and A M] is the absorbance of other components in sample 1, such as 1,2 and 2,4 placements. Similarly, it can be written for spectrum 2 A~2 = An2 + At2 + AM2

v(C=O) In-plane ring In-plane ring $(CH~),

~(CH~)

I t is desired to subtract AT2 from A ~ as follows

A,~ = AT, -- kAr~

(17)

where A a is the absorbanee due to the subtracted spectrum and Ic is an adjustable scaling parameter. Substitution of Eqs. (15) and (16) in Eq. (17) gives

~,(CH3) v(C -O) v(C O) v(C---O) v(C O)

A,~ = (Hal + Ae] -~- AM1)

CH ottt-ofplane ring

Intensities are normalized to intensity at 1581 cm-L

(18) -

-

l¢(Aa2 + At2 + AM~)

R e a r r a n g e m e n t of Eq. (18) gives

A.s = (An, -- ~AA~) + (Act - ]¢Ae~)

TABLE IV. DOP frequencies that change upon p l a s t i c i z a DOP

PVC/0.50 DOP

PVC/0.20 DOP

PVC/0.05 DOP

Tentative assignment

1730 1287 1274 1123 1072 743

1725 1289 1278 1126 1074 744

1723 1290 1279 1128 1075 745

1722 1290 1280 1128 1075 745

v(C=O) v(C O) v(C -O) v(C--O) Ctt grit-ofplane, ring

difference spectrum 3 representing the spectrum of crystalline trans-CD. The absorbance procedure employed will be illustrated mathematically. For spectrum 1 the following equation can be written at each frequency (15)

where AT, is the totM absorbance of all components in

(19)

+ (AM1 -- /,:AM~)

tion.

AT] = A~ 1 + Acl + AM~

(16)

In this procedure it is desired to obtain the spectrum of crystalline trans-CD by elimination of the contribution of the amorphous trans-CD to the s u b t r a c t e d absorbanee spectrum, As. In other words, the criterion for subtraction of A ~ from At, used was AA, -- lea AA, = 0

(20)

kA = AAi/AA2

(21)

where

The criterion employed for elimination of the amorphous component was the disappearaene of bands at 602 and 1227 cm -1. Earlier work b y Mochel and Hall ~6 has shown t h a t the broad 602 c m - ' band is insensitive to crystallization. More recently, the amorphous band at 1227 e m - ' has been used in the determination of crystallinity in polychloroprenes; the results of this study agree with independent dilatometric results, Application of the subAPPLIED S P E C T R O S C O P Y

303

1127

826

671

3. 2018

1318

1660

~

1167

~12a7.t .

95a

1444 14~31

1448

1431

T31~303 • ,~2,

~,1..~

F--3200 =

,

2800 I

1600 I

,,

1400 I

1000 ,..|

600 I

CM'I

,,

FIG. 17. Infrared spectra of lrans-Z,4-polychloroprene ( - 2 0 ° C polyI .r). 1, absorbance s p e c t r u m at room t e m p e r a t u r e ; 2, absort)ance s p e c t r u m at 80°C; 3, absorbance s p e c t r u m of crystalline v i b r a t i o n a l Lands of trans-l,4-polychloroprene (spectrum 1 - s p e c t r u m 2).

traction criterion, shown in Eqs. (20) and (21), can be applied to the total absorbance subtraction in Eq. (19): A~ = (Ac, - kAAez) + (AM~ -- kAAM~)

(22)

According to Eq. (22), this subtracted spectrum now contains the spectrum of the crystalline trans-CD plus the spectrum of the other structural modifications: cis-l,4; 1,2 and 3,4 placements. Since the latter modifications are only present in relatively small concentrations, they will not make a mQor contribution to the subtracted spectrum. In the actual subtraction employed, there is the implicit assumption that the infrared spectrum of the amorphous component in the heated sample is not significantly different from that of the sample at room temperature. A difference spectrum was also obtained from two room temperature spectra taken at different times of crystallization and thus at different degrees of crystallinity. This difference spectrum of the crystalline trans-CD component was essentially identical to that shown earlier in Fig. 17. The similarity betweeen the two difference spectra indicates that the effect of the temperature range used in this study on the infrared spectrum of the amorphous component is minimal. The difference spectrum which results from the subtraction of the spectrum recorded at 80°C from the spectrum recorded after the sample has been allowed to cool to room temperature and recrystallize showed no significant frequency shifts from spectrum 3. In conclusion, this type of procedure represellts an excellent method to study vibrational spectra of semicrystallinc polymers. When polymers polymerized at different temperatures are compared in the above manner, the difference spectra show a shift iIl the crystalline frequencies as indicated in Table V. Such changes could result from the inclusion of structural defects in the crystalline domain. The presence of such defects would contribute to the vibrational spectrum of the crystalline regions. The increase of structural defects as a function of polymerization temperature, ~s with the resultant inclusion of a larger concentration of 304

Volume29, Number 4, 1975

T A B L E V. C r y s t a l l i n e v i b r a t i o n a l f r e q u e n c i e s o f trans.1,4p o l y e h l o r o p r e n e as a f u n c t i o n of p o l y m e r i z a t i o n t e m p e r a ture.

Crystalline v i b r a t i o n a l frequencies (era -~) for tratts-l,4-polychloroprenes polymerized at the following t e m p e r a t u r e s -- 20°C

0°C

40°C

1660 1449 1318 1250 1167 1127 1083 1007 958 826 780 671 577

1660 1448 1316 1252 1167 1127 1083 1005 954 826 779 671 576

1660 1447 1313 1254 1167 1127 1083 1004 953 826 778 671 576

Change in frequency from the - 2 0 ° C to +40°C polymer (era -1) 0 -2 -5 +4 0 0 0 -3 -5 0 --2 0 - 1

defects, could manifest itself in vibrational changes in the crystalline spectrum. 2, Crystalline Vibrational Bands of P V C A 4 The crystalline spectrum was obtained for the normal commercial PVC and a PVC sample polymerized st -37°C. The low temperature polymer has a higher degree of stercoregularity and therefore crystallinity. So, its crystalline spectrum will be more representative of the extended syndiotactic conformation of the PVC chain. Infrared spectra for this PVC polymer are shown in Fig. 18. The bottom and middle spectra are for a cast film and a quench film, respectively. The spectrum of the quenched film is still partially crystalline. However, the important point is that the two spectra have different degrees of crystallinity. The top spectrum is a difference obtained by absorbance subtraction1 of the quenched film spectrum from the more crystalline case film spectrum. The procedure employed for absorbance subtraction has been outlined above. In this specific case, the ideal criterion for subtraction would

be the elimination of the noncrystalline components of the cast film spectrum by subtraction of the desired amount of the more amorphous quenched film spectrum. The remaining spectrum can then be attributed to the crystalline regions of the PVC sample. The methylene deformations at 1435 and 1427 em -1 have been assigned as amorphous and crystalline bands, respectively. ~7 The actual criterion for subtraction was the desired elimination of the 1435 em -~ amorphous band with the necessary requirement that the subtraction does not "distort" any region of the spectrum, such as CH stretching bands (2800 to 3000 cm -~ region) being made negative. The top spectrum in Fig. 18 was obtained in this manner. While an exact quantitative subtraction was not performed, this difference spectrum is POLY[VINYLCHLORIDE]

a qualitative spectrum representing the crystalline regions of the PVC, prepared at - 3 7 ° C by irradiation. The crystalline vibrational spectrum for the commercially prepared polymer was obtained in a similar manner. Spectra of the cast and quenched films are shown in Fig. 19 together with the crystalline difference spectrum. In comparison with the PVC polymerized at - 3 7 ° C (Fig. 18), notice the expected lower degree of crystallinity in the normal PVC spectra as evidenced by the smaller differenee in intensities between the 1427 and 1435 cm -1 methylene deformations. The two difference spectra are presented together in Fig. 20. The two spectra are extremely similar, and the frequencies are identical within experimental error and are listed in Table VI. The crystal-

| 1236

(-37 °C]

1427

~

604 9S6

1336

638

1230

CAST" QUENCHED

_J 1427

615

QUENCHED

1437

1500 I

lOOO I

5~o cM-! u

FIG. 18. Infrared absorbance spectra of poly(vinyl chloride) prepared at -37°(? by irradiation. Bottom, east fihn; middle, quenched film; top, crystalline regions (cast fihn - quenched ilhn). POLY[VINYLCHLO*,DE] [lO3~,]

1427

1256

956

1433 1427

6311

604

/

431S

lsoo I

looo I

soo cM-I ! ~..~

Fro. 19. Infrared absorbance spectra of normM commercial poly (vinyl chloride). Bottom, east film', middle, quenched film; top, crystalline regions (cast film - quenched film). APPLIED S P E C T R O S C O P Y

305

1256

1427 /Itl

":Z22,%o2:o'2

[

956

1500 11

1000 I

604

638

.

.

.

.

.

.

500 CM-1 !

FIG. 20. Crystalline vibrational spectra of poly(vinyl chloride) polymers. Bottom, normal commercial PVC; top, PVC polymerized at -37°C. line frequencies reported for syndiotactic PVC polymerized in urea-complex ~7 and those calculated for an isolated chain in the extended syndiotaetic conformation '9 are also included in Table VI for comparison. The close agreement between the three sets of frequencies indicates t h a t the chains in the crystalline regions of normal PVC have the syndiotactic conformation. As a measure of comparison between the crystalline spectra obtained here by FTS and t h a t previously obtained by dispersive spectroscopy, ~7 the crystalline frequencies t h a t were previously observed ~7 are labeled by an asterisk in Table VI. I t is seen t h a t a definite i m p r o v e m e n t has been obtained with the F T S method. TABLE VI. Frequencies of c r y s t a l l i n e p o l y ( v i n y l chloride) Difference spectrum b

Urea-complex spectrum¢

Normal coordinate analysisa

1427*~ 1381 1354 1336' 1256" 1230' 1214 1104 1089' 1024 956* 832 638* 604*

1428 1387 1355 1338 1258 1230 1195 1105 1090 ~1030 960 835,840 640 604

1445 1404 ]311 1322 1278 1233 1169 1122 1076 1022 1002 834 639 619

• Frequencies in cm-~. ', Data for difference spectra in Fig. 19. Data from Krimm et al. ~7 d Data from Tasumi and Shimanouchi. ~° Frequencies labeled with an * are those crystalline frequencies detected by Witenhafer. 2' 306

Volume29, Number 4, 1975

F . S e p a r a t i o n o f S u r f a c e h ' o m B u l k E f f e c t s . An area of extreme importance is the study of surfaces which involves a variety of fields including catalysis, adhesion. adsorption and lubrication. Although a variety of experimental techniques have been used to study surfaces, m a n y questions still remain. Infrared spectroscopy has been widely used 22 in the past. but F T I R promises to enhance this field of application since the interference due to the bulk phase can be eliminated and the spectrum of the surface or absorbed species magnified. As an example, F T I R has been used to study the surface modification of polystyrene used for cell culture and the adsorption of proteins on the modified surfaces. Cell culture is remarkably affected by the surface of the culture vessel. 23 1. F T I R of Chemically Treated Surfaces. Polystyrene Petri dishes are commonly used for cell culture, but the low surface energy leads to poor wetability by water and to poor adhesion of amniotic cells. In order to improve the properties of polystyrene surfaces, polar sites were created at the surface of the polymer by sulfonation34 The surface properties of this chemically modified polystyrene were investigated using F T I R ? ~ Fig. 21 shows the transmission spectra of three different polystyrene films (~--~31/mils); untreated polystyrene (A), sulfonated polystyrene treated 1 h in 5 % fuming acid at room t e m p e r a t u r e yielding 5.54 X 10 m ionic sites/era 2 (B); and protamine-absorbed on sulfonated polystyrene (C). Only v e r y subtle changes in peak intensities are observed in the regions of 3300, 1650, 1540, and 1200 cm -1. Fig. 22 shows t h a t the scale-expanded difference spect r u m of the sulfonated polystyrene is digitally subtracted using the band at 1600 c m - ' for subtraction. I n this spect r u m the characteristic absorption peaks of the sulfonic acid group on the surface are observed; t h a t is, the bands at 1350 and 1172 cm -1 arise from the ( - S O , H ) and the

A

THICKNESS

~'oo

3.~oo

3doo

2~bo

7~oo

WAVE

NUMBER

1800

31 micron

1400

800

(CM i )

FIG. 21. Infrared spectrum of polystyrene. A, untreated; B, sulfonated for 1 h; C, absorbed protamine on treated polystyrene. I134 .

1218

SULFONATED POLYSTYRENE

determine the influence of these forces on the structure of the polymer. Electron paramagnetic resonance studies2G of deformed and fractured polymers have shown that free radicals are formed as the result of chain rupture that occurs under the applied stress. The number of radicals observed for polyethylene is ~bout 10~ cm-3 at fracture. Further reaction of the free radicals can give rise to vinyl, methyl, and c~rbonyl groups in the deformed polymer. Zhurkov and Korsukov27 have identified these chemical groups by analysis of drawn polyethylene by differential infrared spectroscopic techniques. A Fourier transform infrared spectroscopic study on the formation of these chemical groups in polyethylene during the drawing process has been made. The infrared spectra for an undrawn polyethylene fihn and a drawn film with a nominal draw ratio of 20 are shown in Fig. 24. There was no internal standard band on which the subtraction could be based. The difference spectrum, also shown in Fig. 24, was obtained by elimination of the absorbance at 1303 cm-L This subtraction also eliminated the absorbance bands at 1368 and 1353 cm-L These three absorbancc bands are attributed to methylene

1007

16b0

PROTAMINE SULFATE 3342

i

,',.tA X MIN

0037 OOI2

I(630

/'

IS;~l

Ills 1250 I

j MAX 0 009 t~L N -o 008

i i

38(]0

3000 WAVE

7000 NUMBER

1600 (CM')

1200

000

FIG. 22. I n f r a r e d s p e c t r u m of the surface of sulfonated polystyrene.

bands at 1218, 1140, and 1040 cm -~ due to the ionized ion ( - S O t ) . The peak at 1007 cm-1 reflects the parasubstitution on the benzene ring. The infrared spectra accurately reflect the structural change occurring at the surface by superficial sulfonation of the polystyrene. 2. F T I R Characterization of Adsorbed Species. The difference spectrum of adsorbed protamine compared to the treated polystyrene is shown in Fig. 23. The characteristic absorption of the surface species has been removed, revealing the absorbance spectrum of protamine adsorbed on the polystyrene surface. The amount of adsorbed protein could be determined using labeling techniques and was determined to be 2 ~g/cm 2. The interaction of the protein with the surface is reflected in the shift of the amide A and B from 3285 and 3230 cm-~, respectively, in a cut film to 3342 and 3202 cm-1 for the absorbed species. Additional changes in the amide 1 and II reflecting a change in the conformation from $ to random as a result of the adsorption on the polar polystyrene surface.

G. Speetral

Changes Arising from Mechanical

Effects. In use, most polymeric systems are subjected to some mechanical stresses. It would appear to be useful to

!

t I

3800

3000 WAVE

2000 " NUMBER ( C M ' )

1600

'

]200

"

800

FIG. 23. hlfrared spectra of protamine ~(tsorhed on treated polystyrene.

/11

POL~'EtHTLENE

iLr-' ~>\1

Wo

,~o,0

°"'

,yo

81o