NIST/NBS Steel standards for microbeam analysis

[jeffchen]

Hi all,

I'm not sure if this topic has been discussed elsewhere in the forum, but does anyone know if there has been any systematic study on the suitability of using NIST/NBS steel standards as microprobe standards, i.e. homogeneity of them at microscale? I understand the composition of most of the commercial available ones were based on bulk chemical analysis method. what is your experience with them?
Many thanks already.

Jeff

[JonF]

Hi Jeff,

    I don't have a useful answer to your question other than I too am interested in finding the results of this.

Wondering whether we could expand the question to encompass other metal alloys, too: Al alloys specifically but also Ti, Mg etc

When I looked a while ago for standards, all the metallurgical standards I came across either said that weren't suitable for microanalysis or weren't homogeneous on a micron scale.

Jon

[Probeman]

Hi all,

I'm not sure if this topic has been discussed elsewhere in the forum, but does anyone know if there has been any systematic study on the suitability of using NIST/NBS steel standards as microprobe standards, i.e. homogeneity of them at microscale? I understand the composition of most of the commercial available ones were based on bulk chemical analysis method. what is your experience with them?
Many thanks already.

Jeff

Hi Jeff,
There is nothing systematic about the data below, but I threw a couple of NIST alloy SRMs that I had already mounted into the instrument over the weekend and run a bunch of elements on both. The data looks pretty good but I wasn't able to run a second set of standards for a drift correction as the filament blew during the last traverse.

I suspect the answer to your question is going to be "it depends". That is it depends on which SRM standards we are talking about.  Here is the the first one I ran which is a NiFe alloy, SRM 1159 to be exact. First the major elements:



Note that the Fe appears to vary a little, but according to the homogeneity range calculations below, the Fe is homogeneous within the 1% precision confidence interval.  The minor and trace elements look like this:



Clearly the Cu is *not* homogeneous, but the others look pretty good.

Finally I turned on the Homogeneity Range calculations from the Calculation Options dialog in the PFE Analyze! window and we get this output:

Range of Homogeneity (t-test) in +/- Elemental Weight Percent (Average of Sample):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci    .013    .000    .000    .017     ---    .000    .001    .001    .000    .008    .000
  80ci    .021    .000    .000    .026     ---    .000    .001    .001    .000    .012    .001
  90ci    .026    .000    .000    .034     ---    .000    .001    .001    .000    .016    .001
  95ci    .032    .000    .000    .040     ---    .000    .001    .001    .000    .019    .001
  99ci    .042    .000    .000    .053     ---    .000    .002    .002    .000    .025    .001

Test of Homogeneity (t-test) at 1.0 % Precision (Average of Sample):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci     yes     yes     yes     yes     ---     yes     yes     yes     yes      no     yes
  80ci     yes     yes     yes     yes     ---     yes     yes     yes     yes      no     yes
  90ci     yes     yes     yes     yes     ---     yes     yes     yes     yes      no     yes
  95ci     yes     yes     yes     yes     ---     yes     yes     yes     yes      no     yes
  99ci     yes      no     yes     yes     ---     yes     yes     yes     yes      no     yes

Level of Homogeneity (t-test) in +/- Percent (Average of Sample):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci      .0      .4      .2      .0     ---      .1      .2      .2      .1     1.3      .2
  80ci      .0      .6      .4      .1     ---      .2      .2      .2      .1     2.0      .3
  90ci      .1      .8      .5      .1     ---      .2      .3      .3      .2     2.6      .4
  95ci      .1      .9      .6      .1     ---      .3      .4      .4      .2     3.1      .4
  99ci      .1     1.2      .8      .1     ---      .4      .5      .5      .3     4.0      .6

Range of Ideal Homogeneity (t-test) in +/- Elemental Weight Percent (Average of Sample) (Meisenkothen and Donovan):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci    .095    .000    .002    .050     ---    .000    .006    .006    .000    .004    .001
  80ci    .145    .001    .003    .077     ---    .000    .009    .010    .000    .007    .002
  90ci    .187    .001    .004    .099     ---    .001    .011    .012    .000    .009    .003
  95ci    .223    .001    .005    .118     ---    .001    .013    .015    .000    .010    .003
  99ci    .294    .001    .006    .155     ---    .001    .018    .020    .000    .014    .005

Range of Actual Homogeneity (t-test) in +/- Elemental Weight Percent (Average of Sample) (Meisenkothen and Donovan):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci    .160    .000    .002    .205     ---    .000    .006    .007    .000    .097    .005
  80ci    .244    .001    .003    .313     ---    .001    .010    .010    .000    .148    .008
  90ci    .313    .001    .004    .402     ---    .001    .012    .013    .000    .190    .010
  95ci    .374    .001    .004    .480     ---    .001    .015    .016    .000    .227    .012
  99ci    .494    .002    .006    .635     ---    .001    .020    .021    .000    .300    .016

Next I will look at SRM C2402, which is a very different story!

[Probeman]

Now for SRM C2402 which is a Hastelloy material. First the major elements:



I think we can all agree that this would not be suitable as a microanalytical standard!  Next, a similar story, for the trace elements:



And again, the homogeneity range calculations from the Calculation Options dialog:

Range of Homogeneity (t-test) in +/- Elemental Weight Percent (Average of Sample):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci    .235    .273    .037    .035    .037    .003    .024    .014    .000    .002    .007
  80ci    .358    .416    .056    .054    .056    .005    .036    .021    .001    .003    .011
  90ci    .460    .534    .072    .069    .072    .006    .047    .028    .001    .003    .014
  95ci    .548    .637    .086    .082    .085    .007    .056    .033    .001    .004    .016
  99ci    .722    .839    .114    .108    .112    .010    .073    .043    .001    .005    .021

Test of Homogeneity (t-test) at 1.0 % Precision (Average of Sample):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci     yes      no     yes     yes     yes     yes      no      no     yes     yes      no
  80ci     yes      no     yes     yes      no     yes      no      no     yes     yes      no
  90ci     yes      no     yes     yes      no     yes      no      no     yes     yes      no
  95ci      no      no     yes      no      no     yes      no      no     yes     yes      no
  99ci      no      no     yes      no      no     yes      no      no     yes      no      no

Level of Homogeneity (t-test) in +/- Percent (Average of Sample):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci      .5     1.6      .2      .5     1.0      .2     2.7     2.0      .1      .4     3.8
  80ci      .7     2.4      .4      .7     1.5      .3     4.1     3.0      .2      .6     5.8
  90ci      .9     3.1      .5     1.0     1.9      .4     5.3     3.9      .3      .8     7.4
  95ci     1.1     3.7      .5     1.1     2.2      .5     6.3     4.7      .3      .9     8.9
  99ci     1.4     4.8      .7     1.5     3.0      .7     8.3     6.1      .4     1.2    11.7

Range of Ideal Homogeneity (t-test) in +/- Elemental Weight Percent (Average of Sample) (Meisenkothen and Donovan):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci    .095    .093    .061    .019    .030    .008    .009    .010    .001    .003    .001
  80ci    .145    .141    .092    .028    .046    .012    .013    .015    .002    .005    .002
  90ci    .186    .181    .119    .037    .059    .016    .017    .019    .002    .006    .003
  95ci    .222    .216    .141    .044    .071    .019    .021    .023    .003    .008    .003
  99ci    .292    .285    .186    .057    .093    .024    .027    .030    .004    .010    .004

Range of Actual Homogeneity (t-test) in +/- Elemental Weight Percent (Average of Sample) (Meisenkothen and Donovan):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci   4.588   5.329    .722    .689    .714    .062    .465    .275    .007    .034    .135
  80ci   6.990   8.119   1.099   1.050   1.087    .095    .708    .420    .011    .052    .205
  90ci   8.978  10.429   1.412   1.348   1.397    .122    .909    .539    .014    .067    .264
  95ci  10.705  12.434   1.684   1.608   1.665    .146   1.084    .643    .016    .080    .315
  99ci  14.094  16.372   2.217   2.117   2.192    .192   1.427    .846    .022    .105    .414

[jeffchen]

Now for SRM C2402 which is a Hastelloy material. First the major elements:



I think we can all agree that this would not be suitable as a microanalytical standard!  Next, a similar story, for the trace elements:



And again, the homogeneity range calculations from the Calculation Options dialog:

Range of Homogeneity (t-test) in +/- Elemental Weight Percent (Average of Sample):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci    .235    .273    .037    .035    .037    .003    .024    .014    .000    .002    .007
  80ci    .358    .416    .056    .054    .056    .005    .036    .021    .001    .003    .011
  90ci    .460    .534    .072    .069    .072    .006    .047    .028    .001    .003    .014
  95ci    .548    .637    .086    .082    .085    .007    .056    .033    .001    .004    .016
  99ci    .722    .839    .114    .108    .112    .010    .073    .043    .001    .005    .021

Test of Homogeneity (t-test) at 1.0 % Precision (Average of Sample):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci     yes      no     yes     yes     yes     yes      no      no     yes     yes      no
  80ci     yes      no     yes     yes      no     yes      no      no     yes     yes      no
  90ci     yes      no     yes     yes      no     yes      no      no     yes     yes      no
  95ci      no      no     yes      no      no     yes      no      no     yes     yes      no
  99ci      no      no     yes      no      no     yes      no      no     yes      no      no

Level of Homogeneity (t-test) in +/- Percent (Average of Sample):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci      .5     1.6      .2      .5     1.0      .2     2.7     2.0      .1      .4     3.8
  80ci      .7     2.4      .4      .7     1.5      .3     4.1     3.0      .2      .6     5.8
  90ci      .9     3.1      .5     1.0     1.9      .4     5.3     3.9      .3      .8     7.4
  95ci     1.1     3.7      .5     1.1     2.2      .5     6.3     4.7      .3      .9     8.9
  99ci     1.4     4.8      .7     1.5     3.0      .7     8.3     6.1      .4     1.2    11.7

Range of Ideal Homogeneity (t-test) in +/- Elemental Weight Percent (Average of Sample) (Meisenkothen and Donovan):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci    .095    .093    .061    .019    .030    .008    .009    .010    .001    .003    .001
  80ci    .145    .141    .092    .028    .046    .012    .013    .015    .002    .005    .002
  90ci    .186    .181    .119    .037    .059    .016    .017    .019    .002    .006    .003
  95ci    .222    .216    .141    .044    .071    .019    .021    .023    .003    .008    .003
  99ci    .292    .285    .186    .057    .093    .024    .027    .030    .004    .010    .004

Range of Actual Homogeneity (t-test) in +/- Elemental Weight Percent (Average of Sample) (Meisenkothen and Donovan):

ELEM:       Ni      Mo      Cr      Fe       W      Co      Si      Mn       V      Cu       O
  60ci   4.588   5.329    .722    .689    .714    .062    .465    .275    .007    .034    .135
  80ci   6.990   8.119   1.099   1.050   1.087    .095    .708    .420    .011    .052    .205
  90ci   8.978  10.429   1.412   1.348   1.397    .122    .909    .539    .014    .067    .264
  95ci  10.705  12.434   1.684   1.608   1.665    .146   1.084    .643    .016    .080    .315
  99ci  14.094  16.372   2.217   2.117   2.192    .192   1.427    .846    .022    .105    .414

Thanks for this John. I've got a bunch of NIST alloy standards, I suppose I'll check them through. Would be good if we could build some sorta database for them collectively.

[Probeman]

Thanks for this John. I've got a bunch of NIST alloy standards, I suppose I'll check them through. Would be good if we could build some sorta database for them collectively.

Julien Allaz, Owen Neill and Anette von der Handt have organized a standards "focus interest group" (FIG) that is having a meeting at the Baltimore MAS meeting this summer.

The first goal of the group is to document the standards that are already "out there".   I will re-post Julien's email from the list server if you missed it.
john

Edit by John: Of course this site would be happy to host such a database as well. I assume it will primarily be a "homogeneity on the microscale" database since the average compositions of these materials are already well characterized.

I wonder if NIST has any internal studies of the microscale homogeneity of these materials...

[crystalgrower]

Hello All,

The NIST data on homogeneity is located in their Special Publications list.  The pdf's all begin with SP260-xxx there are a few that specifically address microanalysis.  There was also a special edition of the Journal in 2002 that exclusively discussed microbeam standards.  Google will get you there fastest.

FYI the pucks were good for XRF because the entire surface (the right face always marked) is assayed.  So Charles Taylor didn't exactly read the manuals when he chopped them up...

I found the certificates of analysis to contain useful information and the archival certificates are on a separate webpage at nist.gov so you can download them for free.