Report
The Coblentz Society Specifications for Evaluation of Research Quality Analytical Infrared Spectra (Class II) 7 6 1 Main Ave., Norwalk, Conn. 0 6 8 5 1
These specifications are designed to guide evaluators who are selecting spectra for inclusion in the Coblentz Society's catalog. They replace the earlier specifications published in Anal. Chem., 38 (9), 27A (1966) and subsequently modified in the Coblentz Society Newsletter No. 41 (1969). Class II spectra are reference spectra obtained on high quality commercial infrared spectrophotometers, operated at maximal efficiency under conditions consistent with acceptable laboratory practice. They differ from Class I spectra (Critically Defined Spectral Data) because all currently available spectrophotometers distort the spectrum to some extent, especially when operated under conditions acceptable in routine use. The specifications for Class II spectra given in this report apply only to the absorption spectra of condensed phase systems. These specifications relate to both dispersively and interferometrically measured spectra unless otherwise indicated. When used without qualification, the term "spectrophotometer" designates either a dispersive (grating) or an interferometric instrument. For interferometrically measured spectra all evaluation is based on the spectrogram and not on the interferogram. I. Spectrophotometer Operation A. Resolution. For dispersively measured spectra the spectral slit width should not exceed 2 c m - 1 through at least 80% of the wavenumber range and at no place should it exceed 5 c m - 1 (see Figure 2 of Appendix). There are technical difficulties in measuring the spectral slit width, and the resolution of the spectra submitted for evaluation will be based on a comparison of a standard spectrum of
an approved reference material with the spectrum of this reference material measured under the same conditions as the spectra submitted for evaluation (see Section IV). For interferometrically measured spectra the resolution will be judged similarly by comparison with the standard reference spectrum. The optical retardation must be at least 0.5 cm and the apodization function must be stated. B. Wavenumber Accuracy. The abscissa, as read from the chart, should be accurate to ± 5 c m - 1 at wavenumbers greater than 2000 c m - 1 and to ± 3 c m - 1 at wavenumbers less than 2000 c m - 1 . Calibration corrections within these limits are to be encouraged and should be indicated on the chart. Proof of the wavenumber accuracy will be an appended spectrum of the standard reference material. Fiduciary marks should be recorded on each chart shortly after the beginning and near the end of each uninterruptedly scanned segment of the spectrum. These marks are required to guard against errors from paper shrinkage and from mismatch between the printed chart grid and the spectrophotometer. C. Noise Level. The noise level should not exceed 1% average peakto-peak or 0.25% rms. The judgment of the evaluator is permissible. D. Energy. The spectrophotometer should be purged with dry gas or evac-
uated to ensure that at least 50% of the source energy is available at all wavenumbers (except in the regions of carbon dioxide absorption near 2350 and 670 c m - 1 ) . If the control system of the spectrophotometer permits, it is desirable that adequate purging or evacuation be demonstrated by a single beam measurement, or a measurement against a constant test signal obtained under the normal scanning conditions with no sample in the beam. E. Other Performance Criteria. These are also established by comparison with the standard spectrum of the reference material (see IV). 1. False Radiation. Apparent stray radiation should be less than 2% transmittance at wavenumbers greater than 500 cm" 1 (see IV). 2. Servo System and Recording System. Any spectrum showing evidence of dead spots or of excessive recorder overshoot should be rejected. The spectrophotometer and recorder time constant should be compatible with the scan rate (see IV). 3. Photometer Accuracy. Class II spectra are not intended to have absolute quantitative significance and, at present, it is not feasible to set specifications for photometric accuracy. Such a test might be added later, but for the present it must suffice that the shape and relative intensities of the bands of the submitted spectrum of the reference material be in reasonable agreement with those of the standard curve (see IV and Appendix). 4. Temperature. It is to be assumed that the spectrum is measured at the ambient temperature unless stated otherwise. II. Presentation A. Information to Appear on Chart. Both the structural and the molecular formulae of the compound should appear on the chart. It is also
ANALYTICAL CHEMISTRY, VOL. 47, NO. 11, SEPTEMBER 1975 · 945 A
Write For Free Booklet "HOW TO GET BETTER TLC RESULTS"
pmd PROGRAMMED MULTIPLE DEVELOPMENT . . . is gaining recognition as a new, significant approach to thin layer chromatography. Among the reasons for this are: 1. At comparable development time, PMD offers greater resolution than conventional TLC with still greater improvement for longer developments. 2. For equal resolution, PMD is faster than conventional TLC. 3. Because PMD developed spots are much narrower, PMD offers sensitivity that is 5 to 10 times greater than conventional TLC for major components, and even more for normally undetectable trace components and longer developments. The PMD story, complete with photographs and diagrams, is presented in a new booklet entitled "HOW TO GET BETTER TLC RESULTS". The booklet is free upon request. Also, each month Regis issues "PMD NEWS," a newsletter devoted exclusively to PMD applications and instrumentation. For a copy of the booklet "HOW BETTER TLC RESULTS" and/or to the free monthly newsletter, NEWS," write or phone the Regis sional Service Department.
TO GET receive "PMD Profes-
REGIS CHEMICAL CO. 8210 Austin · Morton Grove, III. 60053 Area Code (312) 967-6000
recommended that the compound name be included; preferably this should conform with the nomenclature used by Chemical Abstracts. The make and model of the spectrophotometer should be recorded as well as the date on which the spectrum was measured. For spectra obtained dispersively all changes of gratings and filters should be specified including the wavenumber at which they occur. No mechanical attenuator should be placed in the reference beam additional to the optical attenuator which is integral to some infrared spectrophotometers. For interferometrically measured spectra the apodization function must be specified (see I. A). For all spectrophotometers the wavenumber positions at which cell changes occur should be specified. The name and address of the laboratory contributing the spectrum should normally be indicated on the chart (see III. A). The physical condition of the sample should be stated (e.g., solution, liquid, mull, halide pellet matrix, etc.). For measurements on solutions the solvent used in each region of the spectrum should be recorded. The concentration and nominal path length should be given for both solutions and pellets. The nominal path lengths of liquid samples should be indicated; a very thin layer may be described as a capillary film (see III. B.l.c). In all cases the cell window or support window should be stated. B. Spectral Range. The chart should cover the range 3800-450 c m - 1 without gaps; extensions above and below this range are strongly encouraged. For such extended range spectra the wavenumber accuracy, false radiation, atmospheric absorption and resolution should be stated and must be acceptable in the judgment of the evaluator. C. Intensity Scale. It is preferred that the intensity ordinate values be expressed in absorbance units. Where the charts are plotted on a linear transmittance scale, a logarithmic ordinate grid is preferable so that the absorbance can be interpolated directly. Spectra plotted on a linear absorbance scale or in linear transmittance on a linear percent transmission ordinate grid are acceptable. Any band over 1.5 absorbance units should be reproduced on a less absorbing sample. A significant fraction of the useful bands should have absorbance greater than 0.2. At least one band in the spectrum should have absorbance exceeding 0.7. When multiple traces are required on the same
CIRCLE 191 ON READER SERVICE CARD 946 A · ANALYTICAL CHEMISTRY, VOL. 47, NO. 11, SEPTEMBER 1975
chart their number should be kept to a minimum. D. Wavenumber Readability. Sharp peaks should be readable to within 5 c m - 1 at wavenumbers greater than 2000 c m - 1 and within 2 c m - 1 at wavenumbers less than 2000 c m - 1 . Only spectra recorded with the abscissal scale linear in wavenumber are acceptable, but scale changes at designated abscissal positions are allowed. E. Recording. Recording should be continuous with no gaps in wavenumber, but it is permissible for spectra to extend over more than one chart. Discontinuities in ordinate (absorbance), if present, should not exceed 0.01 absorbance unit. Hand retraced spectra are not acceptable. F. Atmospheric Absorption. Atmospheric absorption should not exceed the allowable noise level when observed in the double beam mode (see I. D). III. Sample Identification and Preparation A. Compound Identification and Purity. Spectra should show no inconsistences with the postulated structure; any spectrum exhibiting an obvious impurity band should be rejected. Some relaxation of this requirement may be permitted in the case of isotopically labeled substances in which complete isotopic exchange cannot reasonably be achieved. In such cases the bands associated with the minor isotopic species should be indicated on the chart. Because the prime responsibility for the correctness of the structure lies with the laboratory contributing the spectrum, the name and address of the laboratory should be recorded on the chart. In exceptional cases spectra contributed anonymously may be published, but such spectra should be labeled "Chemical Structure not Authenticated". No spectrum should be published unless either: 1. The evaluator is supplied with a reasonably detailed description of the preparation and purification history of the measured sample, together with other evidence for the correctness of the assigned chemical structure. This evidence must be sufficient to satisfy an expert in the field. or 2. Two curves derived from samples obtained from independent sources
If you're in A.A. you'd do well to know IL In the rate of our sales, we think we're only second. But in our rate of innovation, we have good reason to think we're first. Here are some items that are new. 1. AA instrument with programmable calculator. Get automatic statistics on precision and linearity, sophisticated curve correction based on those statistics, peak height and peak area measurements with automatic start and stop of integration, and readout of sample concentration in any units, with sample weight and dilution factor taken into account. Meanwhile, you can still use the calculator for everything else that calculators are good for. 2. More versatility—a UV/Vis cell holder. A good AA instrument is so fast that, in many labs, it's used only a couple of hours a day. So IL has introduced a dual cell holder to replace the burner head, and convert your IL AA to a pretty useful UV7 visible spectrophotometer. 3. An ingenious sample treatment system. The IL455 Flameless Sampler comes equipped with little pyrolitic graphite boats. You can put samples into a bunch of boats, do chemical pretreatments, and then insert them one by one into the 455. Recently we have developed a mini-flame unit that uses air and acetylene, which can burn off the sample matrix while leaving the metallic elements. Our first success was in determining vanadium and nickel at low levels in crude oils. You may not even know us. At this point, you may be thinking: "Could these people possibly be second in AA sales, and I haven't heard of them?" Absolutely. Up to now, we've concentrated on certain regions and particular industries. If you want to know who we are... Write to: gjfc
Instrumentation Laboratory Analytical Instrument Division Jonspin Road / Wilmington, Ma. 01887
are available; these spectra must be in reasonable agreement. B. Sample Preparation 1. Liquid State Samples (a). For analytical purposes it is preferable that the sample be run in liquid solution, normally at concentrations in the range 5-10% weight (g) per volume (cc). Solvent bands should be compensated, but not more than 75% of the energy should be removed from the beam by such compensation and then only over a short region of the spectrum. Any solvent bands resulting from incomplete compensation should be indicated on the chart. A suitable solvent combination is carbon tetrachloride (3800-1335, 650-450 c m - 1 ) and carbon disulfide (1350-450 c m - 1 ) ; both solvents should be used at path lengths in the range 0.03-0.3 mm. Cases may arise that require the use of other solvents, and solubility limitations or other concentration dependent factors may necessitate the use of cells of longer path length. These conditions are acceptable provided the reference beam energy is not attenuated by more than 75%. (b). For documentation purposes it is desirable that the spectrum of the liquid be recorded. Solution spectra and liquid spectra are to be regarded as complementary and not as substitutes for one another. (c). The spectra of liquids not soluble in transparent solvents should be measured as capillary films (see II. A). 2. Solid State Samples (a). Solution spectra in the most transparent solvents are preferred, provided the solvents and path lengths can be chosen to leave no significant gaps due to solvent obscuration (see III. B.l.a). (b). For insoluble compounds mulls are preferred to pressed pellets unless it can be established that the pellet gives an undistorted spectrum. Solid state spectra must meet the following criteria: (i) Isotropic materials. The background absorbance should be less than 0.20 near 3800 cm" 1 and less than 0.10 near 2000 c m - 1 . No gross abnormalities should be evident in the background. Compensation in the reference beam by a blank mull or pellet should be indicated, and in no case should it reduce the reference beam intensity by more than 50%. The Christiansen effect should not be apparent, but minor distortion resulting from this effect may be permitted at the discretion of the evaluator. Interference fringes should not be apparent. Pellets should exhibit no bands
CIRCLE 1 0 2 O N READER SERVICE CARD
948 A · ANALYTICAL CHEMISTRY, VOL. 47, NO. 1 1 , SEPTEMBER 1975
THE VOICES OF CHEMISTRY ACS Audio Courses . . . • are state-of-the art courses in chemistry and chemical engineering • consist of a lecture on audiotape cassettes integrated with a course manual • can be used individually or in group situations • are suitable for industrial or academic use • can be purchased on ten-day approval Topics now available include: Applications of Orbital Symmetry Applications of Raman Spectroscopy Basic Gas Chromatography Business Aspects of Chemistry Chemical Reaction Mechanisms Fluorescence and Phosphorescence Spectrometry Fundamentals of Chemical Marketing Fundamentals of Effective R&D Management Gel Permeation Chromatography Intermediate NMR Spectroscopy Interpretation of Infrared Spectra Interpretation of Mass Spectra Introduction to Patents Ion-Selective Membrane Electrodes A Lewis Acid-Base Approach to Chemical Reactivity Modern Liquid Chromatography Modern Organic Synthesis Modern Theory of Acids and Bases Photochemistry of Organic Molecules Practical Technical Writing Principles of Heterocyclic Chemistry Use of Chemical Abstracts
I would like to learn more about the ACS Audio Courses program. Please send me your catalog. name organization address city state
Department of Educational Activities American Chemical Society 1155—16th Street, N.W. Washington, D.C. 20036
Table I. Recommended Calibration Maxima for Indene-Camphor-Cyclohexanone (98.4/0.8/0.8) Mixture"
a
Band no.
Position, c m - '
Band no.
Position, c m - 1
1 3 5 8 9 17 21 28 36 40 44 44a 44/? 47
3927.2 3798.9 3660.6 3297.8 3139.5 2770.9 2598.4 2305.1 2090.2 1915.3 1797.7 1741.9 1713.4 1661.8
49 55 57 58 60 61 64 66 67 71 73 76 77
1587.5 1361.1 1312.4 1288.0 1226.2 1205.1 1122.4 1067.7 1018.5 914.7 830.5 718.1 692.6
Band no. identifies the band in the top three panels of Figure 1 of the Appendix.
Table I I . Recommended Calibration Maxima for Indene-Camphor-Cyelohexanone (1/1/1) Mixture" Band no.
Position, c m - 1
Band no.
Position, cm "'
1 2 3 4
592.1 551.7 521.4 490.2
5 6 7 8
420.5 393.1 381.6 301.4
" Band no. identifies the band in the bottom panel of Figure 1 of the Appendix.
due to absorbed water greater than 0.03 absorbance unit. Mulls should be made with perhalogenated oils (or equivalent) for the range 3800-1335 c m - 1 , and the intensity of the over tone band near 2300 c m - 1 should not exceed 0.02 absorbance unit. Liquid paraffin (Nujol or equivalent) should be used below 1350 c m - 1 , and the in tensity of the band near 720 c m - 1 should not exceed 0.05 absorbance unit. If there are suitable bands present, they should be overlapped in the two mull spectra and one of the bands should be identified in both mull spectra; this aids in establishing the intensity ratio between the two sections of the spectrum. (ii) Nonisotropic materials. Spectra of nonisotropic materials, such as sin gle crystals or crystalline polymers, should be accompanied by a record of the orientation of the sample with re spect to the radiation beam, and in the case of dispersively measured spectra, the orientation with respect to the grating rulings should also be indicat ed. Amorphic and partially crystalline polymers of ill-defined molecular and conformational structure should not be included as pure materials. IV. Spectrophotometer Performance Checks
The approved standard reference material is a mixture containing 98.4
parts by weight of indene, 0.8 parts by weight of camphor, and 0.8 parts by weight of cyclohexanone. This is the indene mixture recommended by the International Union of Pure and Ap plied Chemistry (IUPAC) for the wavenumber calibration of medium resolution spectrophotometers when used for measurements on condensed phase systems over the range 4000600 cm" 1 . The standard IUPAC spec trum is reproduced in the Appendix. The indene-camphor-cyclohexanone mixture should be stored in sealed ampoules which should be opened im mediately before use. Information about commercial sources of this ma terial is available from the Coblentz Society. Other secondary standard spectra would be acceptable subject to their approval by the Coblentz Society and provided they are calibrated over the full wavenumber range and are com patible with the need to satisfy the criteria in Sections IV. Bl and IV. C. A. Wavenumber Calibration. The recommended wavenumber calibra tion points are the absorption maxima of the standard indene-camphor-cy clohexanone mixture listed in Table I. Suitable path lengths are 0.2 mm for the range 3800-1580 cm- 1 and 0.03 mm for the range 1600-600 cm - 1 . A mixture containing equal parts by weight of indene, camphor, and cyclo hexanone at a path length of 0.1 mm
950 A · ANALYTICAL CHEMISTRY, VOL. 47, NO. 11, SEPTEMBER 1975
may be used for the range 600-300 c m - 1 (see Table II and Appendix). B l . Dynamic Error Test for Dispersively Measured Spectra. For dispersively measured spectra the following dynamic error test is suit able for use with most grating spectro photometers. The spectrum of the indene-cam phor-cyclohexanone (98.4/0.8/0.8) mixture is remeasured from 1350 to 850 c m - 1 at one fourth of the scan rate used for the reference spectrum, with other operating conditions unchanged. The heights from the baseline of the bands at 1288.0, 1226.2, 1205.1, 1018.5, and 914.7 c m - 1 are measured in ab sorbance units on both the fast and slowly scanned charts. The peak height ratios A1226.2/A1288.0, Α 1205 .ι/ A 1226.2, and Agu.v/Aiois.s should not differ by more than ±0.02 between the slow and fast runs. B2. Aliasing Error Test for Interferometrically Measured Spec tra. The following aliasing (folding) test should be made for spectra mea sured interferometrically. The trans mission spectrum of a 0.05-mm layer of liquid water is determined over the range 3800-3000 c m - 1 . The false ener gy at 3400 cm- 1 should not exceed 0.5% T. C. Resolution Test for Dis persively Measured Spectra. In the neighborhood of 1200 c m - 1 the spec tral slit width can be determined ap proximately from the ratio Ai 205.1/ Αΐ22β.2 computed in the dynamic error test, namely:
1205.1/A1226.2
0.80 0.85 0.90 0.95 0.97
Approx. resolution (spectral slit width, cm 1) 4.0-5.0 3.0-3.5 2.0-2.5 1.0-1.5
