].D. Hanawalt, "Phase Identification by X-Ray
`
`Powder Diffraction Evaluation of Various
`
`Techniques”, Adv. X-ray Analysis.,
`
`v.20 (1976) pp.63-73
`
`RS 1027 - 000001
`
`RS 1027 - 000001
`
`

`
`PHASE IDENTIFIC&Tl0N BY X-RAY POWDER DIFFRACTIOH
`
`EVALUAIION 0? VARIOUS TECHNIQUES
`
`J.D. Hanawa1t*
`
`University of Michigan
`
`Ann Arbor, Michigan 43109
`
`*Chairman, Joint Committee on Powder Diffraction Standards
`
`ABSTRACT
`
`Three pwder mixtures, each composed of four or more phases, were
`submitted for phase identification by x-ray diffraction. Laboratory
`technicians supplied tables of "d" values and of relative intensities
`_as obtained separately and independently by use of the diffractometer,
`the Debye camera and the Guinier camera. These tables of diffraction
`data were "solved" by utilization of the Joint Committee search manuals
`and reference to the Joint Comittee Powder Diffraction File (P.D.F.).
`The same tables of data were then submitted to the 2dTS:Diffraction
`Data Te1e—Search for a computer printout of results. Experimental data
`are also presented which provide a quantitative comparison of the
`accuracy of measurement of "d" values and of the resolution of Debye
`caeraa vs Guinier caeras, since this information is necessary for
`efficient search procedures whether by manual or computer methods»
`
`The "solutions" of the three unknown mixtures confirm the general
`experiece that only those diffraction data of the highest quality with
`respect to resolution of close lines and with respect to accuracy and
`intensity of "d" values are adequate for an easy and complete solution
`of complicated mixtures of phases.
`The Debye pattern does not entirely
`meet these requirements though for many simpler pfoblems its great USE-
`fulneas and especially its sensitivity to minute amounts of sample are.
`well known. The diffractometer is a very versatile instrument which
`can provide high quality pwder data and has as well the considerable
`advantage that it provides quantitative intensity values directly.
`Furthermore, units are now commercially available which automatically
`produce the diffraction data for immediate computer search and printout
`of results,
`thus almost completely eliminating manual labor in the
`determination of unknown phases in a pwder mixture.
`The Guinier
`camera also provides powder diffraction data of the highest quality
`though for quantitative intensities the additional step of making a
`microphotmeter trace of the film is necessary. However, for identi-
`fication purposes relative intensities as can be obtained by the use
`of a Frevel type visual density gauge are shown to be sufficient.
`
`O
`
`63
`
`
`
`\
`
`Aaqwwces M FR: Aflagifis’ J.
`
`.
`
`:10 (Hm)
`
`9% 73
`
`RS 1027—000002
`
`RS 1027 - 000002
`
`

`
`J. D. Hanswalt
`
`with respect to the involved question of computer vs manual search
`methods much more experience is necessary. But the evidence from these
`three problems is that each method can determine the phases present
`(insofar as_they are included in the P.D.F.) and that while the com4
`puter is probably faster the manual method is probably cheaper.
`
`INTRODUCTION
`
`It is almost 60 years since the "discovery" of powder diffractionby
`Debye and Scherrer and independently by Hull.
`The type of powder
`camera designed at that time is still in use today and in most x-ray
`laboratories is probably still the "work-horse" in the utilization of
`x-rays for phase identification.
`The diffractometer is a highly
`developed instrument likewise in use for the purpose of powder diffrac-
`tion.
`So also is the Guinier design of camera which is finding con-
`siderably increasing popularity within the last decade.
`
`Xrray diffraction phase identification became more broadly usable
`and important with the availability of a file of standard patterns
`and practical search systems. At present the JCPDS powder diffraction
`file (P.D.F.) contains about 3h,00O patterns and is expanding at the
`rate of about 2000 patterns each year. Computer programs are now
`available for retrieval of unknown patterns.
`
`the analyst in an x—ray laboratory has a considerable choice
`Thus,
`of techniques for obtaining the diffraction data of unknown phases and
`also a choice of methods of "solving" these data.
`The objective of the
`present paper is to make some comparison of the available techniques
`and methods and to illustrate with soe examples how limitations in
`the "quality" of the data reflect in the degree of completeness and
`reliability of the identification of the unknowns.
`
`DEBYE AND GUINIER DATA COMPARED
`
`In order to determine the degree of uniformity to be expected in
`powder diffraction data from various types of x—rey units, x~ray
`patterns of certain standard materials were solicited from a score of
`x—ray laboratories in the U.S.A., Canada, and Europe.
`It would be
`impractical to reproduce the films in this paper but in Figure 1 is
`seen the microphotometer traces of two of the best films which were
`received.
`The traces in Figure 1 illustrate the greater dispersion,
`the sharper diffraction lines and the much lower background of the
`Gtninier pattern and are typ:‘.ca.'|. ofi
`the differences ‘between Guiuier and
`Debye powder patterns.
`
`Accuracy
`
`Experience in measuring the positions of the lines in a Philips
`Vernier scale viewing box indicates that on a Debye pattern this can
`be done to about 19.1 m, while on a Guinier pattern this figure is
`about $9.05 m.
`Since for the same diameter camera the dispersion of
`the Guinier unit is twice that of the Debye,
`the accuracy of measure-
`ment of "d" values for the Guinier film is four times better. Figure 2
`
`RS 1027 - 000003
`
`RS 1027 - 000003
`
`

`
`.
`'_
`lmufldhuifl
`-
`-
`5
`_
`Figure 1. Hicrophotometer traces of powder patterfie of §§§O .
`film from Lawrence Radiation Laboratory, courtesy of D.1(. Smith.
`Guinier film from Uhivereity of Paris, courtesy of A. Guinier.
`
`Debye
`
`II
`II
`
`I I l
`
`l 5 I
`
`:.l'
`
`flmlllipmflflE!IllIn0taunt‘anus
`
`I:=§
`
`flIII‘IIIIIII|‘I5..
`IIIIIIIIIIIIIIIIIIIII
`
`Graph of values of experimentally observed accuracy and
`resolution for cameras of 114.6 mm diameter.
`
`RS 1027 - 000004
`
`RS 1027 - 000004
`
`

`
`
`
`66
`
`J. D. Hanawnlt
`
`shows this plot of ind as a function of d for 114.6 mm diameter cameras.
`Of course to obtain this degree of accuracy of absolute d values,
`the
`film must: be calibrated by use of an internal standard.
`For the par-
`ticular case of the three powder mixtures it can be seen by examining
`the tables of date that the actual accuracy of measurement for the
`Guinier films was considerably better than given by the curve of
`Figure 2.
`In the range up to about 3.5 A most all of the d values of
`the stronger lines agreed with the Bureau of Standards determinations
`to within 10.001 K or less.
`
`Resolution
`
`The question of resolution also involves the sharpness of the
`lines and the background density and for practical purposes requires
`an experimentally determined answer.
`The curves drawn in Figure 2
`represent the results of observations of films from a considerable
`number of different cameras.
`The Debye cameras were 57.3 mm, 114.6 mm
`and 140.0 mm in diameter.
`The Guinier camera diameters were 80.0 mm,
`100.0 mm, 114.6 mm and 120.0 mm.
`Interestingly. While the 140.0 m
`Debye did not make a noticeable improvement in.the resolution of a
`114.6 mm Debye, the 80.0 mm Guinier provided almost as good resolution
`as the 120.0 mm Guinier. Within the important and most comonly used
`range of d values for phase identification the resolution of the
`Guinier type camera is about five times greater than that of the Debye
`camera. This is illustrated in Figure 3 which compares Guinier and
`Debye patterns for the KN03 two lines at A, 3.?8 and 3.73, and the 0
`three lines at B, 2.662, 2.647 and 2.632.
`The separations of 0.05 A
`at A and 0.015 K at B are unresolved in the Debye pattern but clearly
`resolved in the Guinier pattern.
`
`Another illustration is given in Figure a which shows the Y and 2
`regions of Figure l but at a microphotometcr magnification of 50 to 1.
`These close lines can be clearly seen by direct visual inspection of
`the original film.
`The four lines at Y,o2.1l8, 231151, 2.10fi8nand
`2.1012 which are separated by ed 0.0032 A, 0.014 A and 0.0036 A
`respectively as seen in the Guinier pattern (these measurements are
`from Professor Andre Guinier's laboratory) are barely resolved into
`two lines in the Debye pattern. Likewise the three lines at Z, 157605,
`1.7554, and l.?509 which are separated by ed 0.0051 A and 0.0045 A
`respectively are entirely unresolved in the Debye pattern.
`(None of
`the Debye patterns received showed better resolution than this pattern
`from the Lawrence Radiation Laboratory, courtesy of D.K. Smith.)
`
`The curves of Figure 2 showing attainable resolutions of Debye and
`Guinier cameras have been generated from the above type of experimental
`observations of many films.
`It should be mentioned that the resolution
`of the lines at Y and Z in the BaS04 diffraction pattern is only
`possible if the BaS0 has been heated to above 800°C. However, with
`this qualification,
`ihese lines of BaS0 provide a quite practical and
`critical test of the excellence of adjuétment of the diffraction
`camera.or instrument.
`In Figure 5-is shown the resolution by step
`scanning with a diffrsctometer by M. Nichols at Sandie Laboratories,
`Livermore, California.
`The same resolution can also be observed with
`the diffractometer with a slow scan of about 1°29 per minute.
`In
`Figure 6 is shown the resolution of these lines of B3504 as demonstrated
`
`
`
`RS 1027 - 000005
`
`RS 1027 - 000005
`
`

`
`J.D.HsnawaM
`
`Figure 3. Microphotometer trace
`of KNO
`powder patterns. Upper
`traces Guinier,
`lower traces Debye.
`
`2
`
`Joyce Microphuiumohr Magnification 50:1
`
`Figure 4. Regions Y and Z of Fig. 1
`at higher magnification.
`
`Figure 5. Diffractmeter step
`scan of B350 at Y region.
`Courtesy of
`.C. Nichols.
`
`Neutron diffraction of
`Figure 6.
`Courtesy of
`BaS0 at Y region.
`E. sfieichele.
`
`With his neutron time-of-flight diffractometer by Dr. E. Steichele, at
`the Techn. Univ. Munich. Dr. Steichele heated the B330
`to ]_000°C and
`used 20 cc for his sample.
`The resolution by neutron diffraction illus-
`trated in Figure 6 is superior to any of the x-ray diffraction tech-
`niques which we have observed.
`It is uncertain, however, why the Ads -
`of the neutron diffraction lines are not exactly the same as the ads by
`x—ray diffraction.
`
`Dr. Steichele has called our attention to another critical test of
`resolution which is posed by the Cawb pattern.
`ghe cawo
`(scheelite)
`powder pattern has a strong line (312? at 1.5921 A which is separated
`
`BS 1027 - 000006
`
`RS 1027 - 000006
`
`

`
`
`
`63
`
`J. D. Hanawalt
`
`The 303 line is clearly
`from a weak line (303) by a cd of 0.0045 3.
`resolved by the Guinier camera in spite of the fact that the 312 line
`has about 30-fold greater intensity.
`The 303 line is not resolved by
`the Debye camera and can be detected with a diffractometer only by a
`very slow seen.
`
`In summary it can be concluded that the Guinier camera of the same
`diameter as the Debye camera provides about four times the accuracy of
`determination of d values and about five times the resolution.
`The
`quantitative plots of accuracy and resolution as a function of d value
`as given in Figure 2 are necessary for efficient search procedures
`whether by manual or by computer methods.
`one needs to know for each
`step of the search how much error to allow-for.or, in the case of Cour
`puter search, what error "window" to use.
`The analyst mst, however,
`keep in mind if the search proves difficult that not all of the patterns
`in the P.D.F. are high quality "starred" patterns.
`
`ILLUSTRAIIVE SOLUTIONS OF UNKNOWN POWDER PATTERNS
`
`Tables 1, 2 and 3 list the diffraction data from uknown mixtures
`supplied by Mr. Ron Jenkins of Philips Electronic Instruments, Inc.
`These data were produced by a graduate student, Mr. Chi-Hung Leung, who
`had no other connection with the project.
`The main steps in Mr. Leung's
`work were to mix the unknown samples with Si powder as an internal
`standard, make the 1-1/2 hour exposure to CuKu
`radiation in a 100 m
`diameter Hag; design XDC—70O Guinier camera, develop the film, use a
`Philips vernier scale viewing box to make the linear measurements of
`the diffraction lines, and type these measurements into a teletype—
`writer terminal of the University of Michigan cmputer in which had
`been stored the Lacrangian program (15) using six lines of silicon
`for calibrating the film.
`The "d" values as printed out by the com-
`puter are listed in Tables 1, 2 and 3. Mr. Leung also made micro-
`photometer traces of the films and calculated the relative intensities
`after measuring the peak height above background of each line.
`It is
`interesting to note that equivalent "d" values can be obtained by less
`sophisticated methods simply by visual reading on a Nies scale with
`careful use of the-silicon lines for calibration. Also,
`though some~
`what
`tedious.
`the visual reading of intensities with a Frevel type
`gauge (5) is satisfactory for identification purposes.
`
`Manual Solutions
`
`.I.D.
`the author,
`Beginning with these three tables of data,
`Hanawalt, "solved" the three unknown mixtures by manual search of the
`J.C.P.D.S. Mineral File for Table 1 and search of the Mini File for
`Tables 2 and 3.
`It was fortuitous that the Mini File did contain all
`of the inorganic phases present.
`If any phase had been absent in the
`Mini File,
`the search_would have been continued using the Maxi File,
`i.e.
`the total listing of inorganics in the P.D.F.
`(The Mini File is
`limited to about 2000 of the more commonly occurring inorganics in the
`P.D.F.).
`
`The results of the manual search for each of the three unknown
`mixtures are shown in the respective three tables.
`No other informs"
`tion than the diffraction data was used in the search.
`Inspection of
`
`RS 1027 - 000007
`
`
`
`RS 1027 - 000007
`
`

`
`I . D. Hnnawslt
`
`TABLE 1. Mineral #1
`
`
`
`7-321‘12°.3’35:9GibhsitePennState
`
`4.85 320
`
`4.37 50
`1.32 23
`
`20-535 F05pyzrhotiteTokyoU.
`
`2.07 180
`
`1.750 16
`
`1.689 11
`
`1.511 15
`
`Iflflhu O
`
`Hi;-ammo
`
`1.189
`1.152
`1.?58
`1.751
`1.T2fi
`l.T21
`1.682
`1.6?‘
`1.670
`1.660
`1.594
`
`an?‘
`1.535 15
`
`9-423 Curesc:~u.1a3pyr:teTorontoUniv
`
`3.03 100
`
`1.565 40
`1.354 80
`
`1.591 60
`1.573 20
`
`EED
`
`MI
`
`I
`
`19
`1D
`11
`11
`10
`35
`10
`21
`70
`21
`I9
`65
`
`4.809
`4.150
`{.382
`{.339
`4.259
`3.889
`3.???
`3.531
`3.4!!
`3.346
`3.318
`3.102
`3.040
`2.989 6
`2.335 34
`2.?2T 42
`2.645 ?
`
`2.606
`2.481
`2.451
`3.112
`2.385
`2.324
`2.280
`2.210
`2.166
`2.121
`2.104
`2.069
`2.056
`1.930
`1.870
`1.855:
`1.836
`1.819
`1.800
`1,190
`1.751
`1.753
`1.728
`1.121
`1.684
`1.6?!
`
`1.660
`1.59%
`1.515
`1.53;
`
`these three tables will show that in each case the identification of
`the phases contained in the mixtures is positive and convincing. The
`purpose in presenting these tables in detail is to permit the viewer
`to check the exactness of the agreement of "d" values from the Guiniet
`pattern iaith the ":1" values from the P.D.F.
`It will be seen t:h.at,with
`a few exceptions here and there, ":1" values agree within ;I-_ 0.001 A or
`less.
`By chance most of the standard patterns involved originated at
`the 11.5. National Bureau of Standards and were produced by their
`exacting diffractometer techniques. The other standard patterns also
`came from very reliable sources.
`It is thus demonstrated that Guiniet
`camera "d" values agree quite axlctly with the best diffractometer data.
`with regard to intensities the agreanent is satisfactory considering
`the many well known factors which may affect this quantity.
`
`RS 1027 - 000008
`
`RS 1027 - 000008
`
`

`
`70
`
`J.l.Iiana19aJl
`
`TABLE 2.
`
`.
`:2
`5
`2
`3
`-I
`N
`N
`N
`""'
`53
`7.5%..
`2.5%
`«H4232
`83
`-4-Asa
`
`
`5-?5? 8
`-
`2-511:
`4.347 12
`4.500 12
`4.424 23
`-
`:-23:3
`-
`2-23 2
`3.5:; 130
`.4
`3.393 15
`3.364 5
`.
`2.3;: :5
`2.929 3
`2.903 22
`2.804 12
`2.861 14
`2.770 5
`2.714 12
`
`2.23 4
`2.
`7
`2.545 15
`2.437 45
`2.431 10
`2.403 17
`2.373 30
`2.355 5
`2.344 5
`2.334 5
`2.296 7
`2.257 10
`2.223 0
`
`2.210 7
`2.197 34
`2.152 9
`2.119 7
`
`2.054 14
`2.057 12
`2.025 5
`2.001 5
`1.992 5
`1.392 45
`1.702 10
`1.599 32
`1.537 70
`1.555 27
`1.523 21
`1.554 4
`
`3.52
`
`100
`
`2.431 10
`
`2.373 20
`
`2.332 10
`
`1.392 35
`
`1.570 20
`
`1.555 20
`
`Inorganic #2
`OK‘
`3*‘?
`-I N
`73..
`N
`nag
`5.96
`7
`5.72
`12
`.91
`50
`=5
`25
`4.54
`55
`4,59
`4.42
`100
`1?
`. 62 17
`2-20 *5
`3.569 15
`3.390 55
`3.359 25
`2.943 50
`3.240 12
`2.923 65
`2.900 90
`
`' 3.25
`
`100
`
`2.351 45
`
`2.712 50
`
`2-6?). 15
`2.547 75
`
`2-409 42
`
`2.355 19
`2.344 20
`
`2.256 25
`2.229 25
`2.214 20
`2.209 30
`
`2.143 13
`2.115 13
`
`2.055 50
`2.023 12
`1.999 19
`1.939 12
`
`1.733 3
`
`2.437 50
`
`2.297 3
`
`2.183 25
`
`2.054 1
`
`0
`
`1.537 3
`
`1.524 20
`
`r--
`22°
`43:9-
`1
`
`5.75
`
`40
`
`4.
`‘-3:
`
`35
`
`3.43
`7
`3.40 90
`
`2.39 55
`
`2.75
`
`35
`
`2.327 15
`2.
`0 40
`2.492 7
`
`2-405 7
`
`2.135 17
`2.147 11
`
`1,992 17
`1.919 25
`
`1.564 11
`
`Computer Solutions
`
`Ihese same three tables of diffraction data for the three unknown
`mixtures were also submitted for computer solution via 2dTS.
`2dTS:
`Diffraction Data Tele-Search is designed by the Joint Committee on
`Powder Diffraction standards to make available to the user data
`retrieval and identification of multiphase mixtures via a remote com-
`puter facility. By teletypewriter, access to the P.D.F. and utilize-
`tion of the Johnson-Wand Search-Hatch program is achieved.
`£dTS prints
`out the ten best choices along with a "scale factor" which is an
`indication of the relative amounts of the phases.
`It is then necessary
`
`
`
`RS 1027 - 000009
`
`RS 1027 - 000009
`
`

`
`2"’
`
`I D.Hhn0wafl
`
`71
`
`9
`Q
`-«-I
`-53
`as
`'7-995 5
`4-706 18
`4.575 20.
`4.421 10
`4.415 10
`3.002 52
`3.222 92
`3.552 35
`3.593 2
`3-3T1 3
`3.292 100
`3.214 21
`3.011 11
`2.959 5
`2.233 5
`2.594 20
`2.593 5 -
`2.555 2
`2.540 5
`2.525 43
`2.593 34
`2:503 4
`2.458 4
`2.421 3
`2.411 3
`2.333 100
`2.325 50
`2.253 2
`2.149 29
`2.104 12
`2.040 12
`2.015 33
`2.014 54
`1.990 32
`1.941 23
`1.932 13
`1.905 54
`1.052 2
`1.290 3
`1.251 5
`1.215 3
`1.525 2
`1.550 45
`1.530 5
`1.514 21
`1.425 22
`1.424 25
`1.325 0
`1.352 4
`1.342 52
`
`O
`""|
`:0
`293
`:22
`8.02
`20
`4.22
`100
`4.57
`100
`4,42
`39
`
`r---
`EON]
`222
`252
`2.69
`4.52
`
`1
`11
`
`4..1 [Q9
`,3;
`2
`
`3.32
`
`35
`
`2.234 50
`2.595 20
`
`2.540 15
`
`'
`
`2.41.3 2
`2.335 25
`
`1.921 20
`
`1.252 2
`1.252 5
`
`TABLE 3.
`
`Inorganic #3
`
`.-4
`‘Rug
`40
`
`n ,9
`kg.»
`1.2
`
`'___
`.2...
`.-.23
`
`
`4.56
`4.45
`
`9
`4
`
`3.22
`3.59
`
`100
`53
`
`3.215 15
`3.025 4
`
`2.249 3
`
`2.555 11
`-
`2.525 24
`2.390 23
`
`2.231 5
`2.150 25
`2.104 12
`2.040 10
`2.019 21
`
`1.940 15
`
`1.559 3
`
`1.232 2
`1.205 1
`1.522 5
`1.549 4
`'1.s21 4
`
`1.325 5
`1.355 4
`1.340 4
`
`3.00
`
`15
`
`3.292 100
`
`2.333 52
`
`1.999 2
`
`1.905 15
`
`1.549 10
`
`1.514 2
`1.425 12
`
`1.345 0
`
`3.24
`3.50
`3,53
`
`25
`25
`35
`
`20
`3.10
`50
`3.01
`2.959 20
`2.232 2
`2.594 95
`2.595 35
`
`2.527 55
`2.592 55
`2.503 12
`2.455 20
`2.421 30
`
`2.357 13
`
`2.325 95
`
`2.013 100
`
`2.019 20
`
`1.424 40
`
`for the analyst to manually inspect these ten answers and decide which
`and how many of the ten are valid and correct answers to the problem.
`The 2dTS search for problem 1 was carried out on the Hineral File and
`also on the Hhxi File; for problems 2 and 3, it was carried out on th8
`H101 File and the 05221 File. For these three problems 2d'J.‘S provided
`the some answers nether the search was made on the Maxi File or on the
`smaller files. For each of the three problems the computer gave the
`same principal solutions as obtained by the manual search plus extra
`answers which must be inspected as has been mentioned.
`
`Table 4 40005 sue comparisons of the times and tests of the solu-
`tions by manual vs computer methods.
`The manual times listed give the
`total time, 1.e.
`the time for search plus the time for matching.
`Generally the time required for manua1.matching of 5;; of the lines of
`the standard pattern given in the P.D.F. is considerably more than the
`
`RS 1027 - 000010
`
`RS 1027 - 000010
`
`

`
`72
`
`J. D. Hanawfllt
`
`the search time required for locating the possible answer based on the
`eight lines in the Search Manual.
`
`The 2dTS figures in Table 4 show the seconds and dollar costs for
`central computer tie and the_time required for punching the diffrac-
`tion data on tape to feed the teletypewriter terminal.
`The figures do
`not include the elapsed time of a few minutes to get the answer back
`(or possibly more minutes if access to the computer is not immediate)
`nor the time which may be required to inspect the ten most probable
`answers printed out by the computer. when the computer search is made
`on the Maxi File rather than on the smaller specialized files the search
`is equally successful for these three problems, but the times and costs
`for the computer central processing unit are considerably greater.
`Also, as to be expected,
`the manual search method is considerably more
`tedious when the much larger Maxi Search Manual must be used.
`
`TABLE 4.
`
`Time and Costs for Solution of Unknowns
`
`Problem
`
`Manual Method
`Total Time
`
`Mineral 1
`
`2 hr 45 min
`
`Inorganic 2_
`
`1 hr 1? min
`
`Inorganic 3
`
`2 hr 47 min
`
`2dTS Computer Method
`Entry of Data
`Central Computer-
`Time
`Time
`Cost
`15 min
`67 sec
`$42.18
`
`l5 min
`
`15 min
`
`26 sec
`
`$19.04
`
`24 sec
`
`$17.96
`
`The diffraction data for these three problems were also processed
`by Mr. M.C. Nichols with the Maxi Pile using his program on a CDC 6600
`computer at Sandia Laboratories in Livermore, California. Mr. Nichols’
`program was also successful in obtaining the correct solutions to the
`problems.
`The figures are not at hand for the times and costs of Mr.
`Nichols‘ solutions.
`
`Solutions with Debye Camera Data
`
`To save space the Debye and the diffractometer data of the three
`problems are not shown.
`The diffractometer data essentially match the
`Guinier data.
`The Debye data, however, are not equivalent,
`some lines
`being missing and others having shifted "d" values in spite of the fact
`that the Debye films were carefully calibrated with internal standards.
`These diffraction data provided very unsatisfactory results in the 2dTS
`"computer printout of answers.
`For problem 1, only three of
`the ten
`listed phases agreed with the five correct answers given by the Guinier
`data.
`For problem 2, only two of the four correct answers were given.
`For problem 3 two of
`the five correct answers were given in the final
`printout though two more were included among the 50 phases with highest
`reliability ratings.
`It should be mentioned that if information
`regarding the chemical elements present or absent had been included in
`the computer input, more complete answers might have been obtained.
`It should also be mentioned that experience shows that for many simpler
`multiphase unknowns in which super—positions and interference of lines
`are absent the Debye pattern will be fully adequate for a complete
`solution.
`
`RS 1027 - 000011
`
`RS 1027 - 000011
`
`

`
`LlD.Hanawafl
`
`73
`
`SUHMARY ARD CONCLUSIONS
`
`Instruments and methods in the field of phase identification by
`x-ray powder diffraction have been evolving over the past 60 years
`(1-10) and have reached a high state of automation. The advent of
`computer programming has contributed greatly and has become quite inw
`dispensable to the continuing development of data banks and data re-
`trieval (11-17).
`
`The evidence submitted in the present paper is that with the super-
`ior quality of data provided by the Guinier camera, manual methods can
`also be very successful.
`It can be concluded that the small x-ray
`laboratory with limited equipment or the laboratory so located in the
`world as not
`to have access to large computers can with adequate per-
`sonnel take full advantage of the inherent power of x—ray powder dif-
`fraction phase identification.
`
`REFERENCES
`
`I-| -
`
`P. Debye and P. Scherrer, Phys. 2. ll, 277 (1916); 18, 291 (l9l?).
`2. A. 1:. Hull, Phys. Rev. g, 661 (1917); J. mi. chemfsoc. 4_1, 1159
`(1919).
`‘
`3. A. N. Winchell, Am. Min. $2) 261 (1927).
`4. A. Guinier, c. R. Acad. Sci. ggq, 1115 (1937); Ann. Phys. 12;, 161
`(1939).
`J. D. Hanawalt, H. W. Rinn, and L. K. Frevel, Ind. Eng. Chem.,
`Anal. Ed. lg, 1.57 (1933).
`6. A. K. Boldyrev, V. I. Mikheev, V. N. Dubinina and G- A. Kovalev,
`Ann. Inst. Mines Leningrad ll, 1 (1938).
`W. P. Davey, J. App. Phys. 19, 820 (1939).
`P. H. Dewolff, Acta Cryst. l, 20? (1948).
`W. Parrish, Am. Min. jg, 770 (1943); X-ray Analysis Papers,
`Centrex Publ. Co., Eiodhoven 1965.
`10. w. c. Bigelow and J. v. Smith, ASTM, S.T.P. 33. 54 (1965).
`11.
`D. K. Smith, Law. Rad. Lab. UCRL—7196 (i963); ibid. UCRLrS0264
`(1967).
`12. M. c. Nichols, 1JCRL—7007B, 19 (1956).
`13.
`G. G. Johnson and V. Vand,
`Ind. Eng. Chem. §2, 19 (1967).
`14.
`L. K. Frevel and C. E. Adams, Anal. Chem. £9, 1335 (1963).
`15. Lagrangian Interpolation, Fortran IV, Ingeniorsfirman Instrumentt—
`janet, Sundbyberg, Sweden (1970).
`_
`G. G. Johnson, Data Base and Search Programs, J.C.P.D.S.
`(197é).
`The Joint Comittee on Powder Diffraction Standards, 1601 Park
`Lane, Swarthmore, Pennsylvania
`19081.
`
`5.
`
`PE”I“
`
`16.
`17.
`
`RS 1027 - 000012
`
`RS 1027 - 000012