Greetings, fellow researchers.
Lately I've been trying to grow single crystal germanium thin film (100 nm thickness) on (100)silicon wafer, the thin film crystalline will be rectified using high-resolution rocking curve of Ge(400) peak, using Bruker Advance D8, CuKa1(1.5406A), witch should locate at 66 degree under theta-2theta scan...
How ever, during quality check of our commercial single crystal silicon wafer, I've found that naked silicon wafer already reflected x-ray at 2theta=66 degree(see picture below), and under such circumstances rocking curve of thin film germanium would be totally useless, since their peaks overlaps. I'd like to know the cause of the weird 66 degree signal of naked silicon wafer, so that further experiment could be organized.
Possible reasons of this anonymous peak includes:
1. Under-filtered x-ray source: not quite possible, since CuKa2 would reflect at 2theta=69.7 degree, CuKb would reflect at 2theta=60 degree, none of them locates at 66 degree.
2. Preferred oriented SiO2 native oxide: unlikely, since native oxides of Si are often very thin (less than 30 nm)and present amorphous characters; although SiO2 (300) diffraction angle is about 65.75 degree, close to 66.
3. Signal from sample holder: could be a reason, but the naked silicon wafer sample is large(2cm*2cm) and thick(600 micron thickness), therefor also unlikely.
If it's not a Cu K-beta line you have to check K-alpha line from tungsten (W). If you have an old lamp it is possible to see such lines (W atoms can be removed from kathode as the lamp works, and deposited on your anode).
Cheers!
>>Marcin Perzanowski
Thanks for your reply, after some calculation I've found that L series of W could be the signal source...waiting for certification...
Good News! A well articulated question! Better News! Lots of room to get better. Grateful for the "Edit" tab:-)
Clarification:
- Is the XRD profile shown, for the "substrate" alone? Is that what you mean by "naked"?
- Have you compared the front and back of your wafer as well?
- What is the sampling size on the specimen surface? Slit sizes used? Beam size on specimen?
- Did you run a "known standard" yet, to calibrate?
- What deposition method? Where is the diffractogram for the film?
Suggestions:
The Bragg profile you've shown indicates "hump" due to potential "low order" phase (high defect, near amorphous). Got info? Good quality wafer does not produce Bragg profiles as show earlier. Have you checked how the profile is affected by Phi rotation about the surface normal to the sample?
https://www.researchgate.net/post/What_is_the_best_way_to_interpret_REAL_TIME_3D_X_Y_o-2TH_o_ph_ch_reciprocal_space_relative_intensity_data_from_HRXRD_rocking_curve_analyses
https://www.researchgate.net/post/What_depths_can_one_typically_probe_with_XRR_X-ray_Reflectivity_Reflectometery?_tpcectx=profile_questions
"I've found that L series of W could be the signal source...", then you should find other reflections as well from those wavelength components. Is that so? Do the relative intensities make sense as well?
In my many decades of work with XRD, I've never encountered such unusually high intensity contributions from "W deposited on the anode". It is worth confirming.
A standard reference material will unequivocally establish the veracity of any such claims.
Hi,
"In my many decades of work with XRD, I've never encountered such unusually high intensity contributions from "W deposited on the anode"."
My answer was so straightforward (regarding this L-series W contribution) since two months ago I was fighting with the same problem. In the pattern Yii-Hang Liu shows this no-name peak has intensity of approx. 10% of the main reflection. And we had the same issue. Of course, there could be another reason for this. However, if it is observed for single crystal Ge the first thing which is coming to my mind is this W peak. Especially in case when the positions match.
Well, I'm interested in the result.
Cheers!
Since both of us are interested in new information and while we are waiting for more information from Yii-Hang Liu, I'd like to push my envelope of knowledge further :-)
Marcin! At what typical tube age (use hours or fraction of remaining life) should such spurious wavelengths from cathode material be expected? Assuming constant uninterrupted use. If the tube is turned on & off often (thermal fatigue), the results would be different, I think.
BTW "decades of experience" can be easily abridged with present day XRD technology, almost instantaneously!
Ravi,
well, I do not know what is a "typical" working time. But I can say something about PANalytical Cu lamp we had. We started to observe W L-line after approximately 4 years of a constant usage. The lamp was mostly at 40 kV/30 mA (40/40 or 40/45 cases were very rare). If I remember correctly it was something like 35000 hour of working.I'm not at my institute right now so I can't remember the number of on & off switches. But I noted this somewhere.
At the beginning it was confusing. And this confusion was the same as in Liu's case. We had peaks in patters which didn't match anything. Or they gave us false information about composition. It was like - "hey guys, I've just measured pure Bi film on Si substrate and got peaks from Na. What the hell? Is there any contamination or something?" However, our software for a basic pattern analysis has a feature which allows to show Cu K-alpha1, K-alpha2, K-beta, and W K-alpha lines. And after analysis of many patterns measured over past few months we found that these "extraordinary" peaks came from W lines. We also made reference measurements (Si & Ge calibration plates, Au, Al, and Ti sheets). Just before changing the lamp to a new one the intensity (amplitiude) of these "W-peaks" was like 10% of the "main" Cu-peaks. So the situation was (almost) the same as here.
Now we have the new lamp. And after 6 months of working we're starting to observe small "W-peaks". However, in this case the lamp is propably broken from the beginning. And we have to change it once again.
Cheers! :)
Wow! Enlightening!
All the more reason to run standard samples along with (contemporaneously) each XRD profile. Otherwise, lots of futile pontification, I guess :-)
Thanks Marcin!
- Badges
- Science topic
- Similar topics
- Entomology
- Resistance