A catalytic alloy approach for graphene on epitaxial SiC on silicon wafers
Summary (2 min read)
Introduction
- The authors introduce a novel approach to the synthesis of high-quality and highly uniform few-layer graphene on silicon wafers, based on solid source growth from epitaxial 3C-SiC films.
- This work opens the avenue for the true wafer-level fabrication of microdevices comprising graphene functional layers.
- Ever since graphene was experimentally isolated about a decade ago,1 the high temperature (1300–1700 °C).
- This endeavor has proved more challenging than expected, in particular, because of the upper limit set by the melting temperature of silicon and the scarce availability of a defect-free and atomically smooth epitaxial SiC on Si(111) starting template.
- 10,11 However, their catalyst-mediated process10 allows for a reduction in the optimal synthesis temperature and for a relaxation of the strict requirements on the starting template.
II. EXPERIMENTAL
- Monocrystalline 3C-(cubic polytype) SiC films, 250 nm thick, were epitaxially grown on ,111.
- Sample foils for transmission electron microscopy were prepared via a focused ion beam liftout technique using a FEI Strata DB235 FIB/SEM with a Ga1 ion source.
- A total of eight measurements were taken for each sample by positioning the electrical probes on the sputtered metal contacts and sweeping the DC input current from a HP4145B parameter analyzer from 0 to 10 mA, which were then separately averaged into “vertical” and “horizontal” sheet resistance groups.
A. Mechanisms for catalytic alloy graphitization and physical analysis
- Figure 2 shows an overview of the ID/IG band ratios, indicative of the defect density of graphene layers,16 calculated from the Raman spectra of graphene prepared with different compositions of metal catalysts on SiC (100) and SiC(111), as measured after the removal of the reacted metal layer by wet etching.
- The error bar represents the variation obtained by measuring five different sites over the prepared samples.
- This means that asperities larger than 200 nm and smaller than 50 nm are considerably reduced with the use of Ni/Cu.
- The second important advance of this study is the demonstration that the Ni/Cu alloy graphitization achieves quality graphene on both SiC(111) and SiC (100) surface orientations, with a consistently better quality obtained from SiC(100), as inferred from the ID/IG trend in Fig.
- It is reasonable to expect that the metal-induced graphitization of SiC would proceed through an amorphization of a thin top portion of the SiC film, as the Ni helps to weaken the Si–C bonds and to release the atomic C. Given the ,1 nm thin nature of the graphene, the XPS analysis is also clearly probing the underlying amorphous layer.
C. Electrical measurements
- The authors have performed AU9sheet resistance measurements on the graphene samples prepared with the same graphitization process as the one shown in the TEM micrograph in Fig.
- The small magnification image shows a region near the top surface of the epitaxial SiC. J. Mater.
- Res., Vol. 30, No. 0, 2015 5 value of all measurements where conduction was along one axis of the samples (horizontal and vertical), so that each point on those curves is an average of four distinct van der Pauw measurements.
- By averaging measurements taken on five samples fabricated with the same graphitization procedure over separate runs, the authors can conclude that the sheet resistance of their bilayer graphene prepared on a SiC(100) layer with Ni/Cu is around 24.8 X/square 6 0.7 from sample-to-sample variation.
IV. CONCLUSIONS
- The authors have demonstrated what is, to their knowledge, the first successful approach to transfer-free, direct growth of uniform and high-quality bilayer graphene over full silicon wafers.
- The methodology relies on the use of an epitaxial SiC layer on silicon as a carbon source for graphene, and the use of a catalyst alloy of Ni and Cu, and can be used with both SiC(111) and SiC(100) surface orientations.
- The obtained bilayer graphene is around 0.8–0.9 nm thick, matching the electrical resistivity of bulk Au.
- Additionally, preliminary results indicate that the adhesion of graphene to the underlying substrate could be an order of magnitude higher than the adhesion of a graphene layer grown ex situ and transferred onto a SiO2 layer on silicon.
- Large benefit from this advance is anticipated in areas such as actuation forAU10 MEMS/NEMS and advanced onchip interconnects for micro and nanocircuits.
Did you find this useful? Give us your feedback
Citations
330 citations
Cites background from "A catalytic alloy approach for grap..."
...Catalytic growth of SiC on Si with a NiCu coating [742] allows growth of graphene on predefined locations....
[...]
...Growing an epitaxial AlN layer on Si prior to SiC growth significantly reduces Si out-diffusion and helps grow higher quality graphene [738], as well as and interface NiCu layer [742]....
[...]
290 citations
182 citations
162 citations
137 citations
References
55,532 citations
7,709 citations
"A catalytic alloy approach for grap..." refers result in this paper
...In fact, this value is comparable with the best sheet resistance reported so far for graphene by Bae et al.(31) In Fig....
[...]
...In fact, this value is comparable with the best sheet resistance reported so far for graphene by Bae et al.31 In Fig....
[...]
6,579 citations
5,673 citations
"A catalytic alloy approach for grap..." refers background in this paper
...The intensity of the 2D band around 2700 cm 1 is slightly higher than that of the G band (;1580 cm (1)), which indicates monolayer to bilayer graphene.(16) The spectrum in Fig....
[...]
3,315 citations
Related Papers (5)
Frequently Asked Questions (15)
Q2. How many contacts were used in the analysis chamber?
Sheet resistance measurements were taken using four electrical contacts in a van der Pauw configuration on the corners of AU51 1 cm2 samples.
Q3. What is the reason why SiC(100) graphene is better than SiC(?
The reason why the SiC (100) template leads in their case to an even better graphene quality than SiC(111) could be related to the combination of a considerably higher tensile stress and higher surface defectivity of the latter epitaxial SiC as discussed in their previous work.
Q4. What was the base pressure in the analysis chamber?
The base pressure in the analysis chamber was below 1.0 10 8 Torr. Data analysis was performed with the CasaXPS software and a Shirley baseline with Kratos library relative sensitivity factors.
Q5. What was the ion milling technique used to remove Ga ion damage?
Subsequent Ar1 ion milling was conducted in a Fiscione NanoMill™ to remove Ga ion damage and provide a high resolution transmission electron microscope (HRTEM) foil.
Q6. What is the resistance of graphene on SiC(100)?
This sheet resistance corresponds to a resistivity of about 2 10 8 X m for the graphene, as low as bulk Au metal; (b) sheet resistance measured on the ;23 nm Ni/Cu alloy as a reference.
Q7. What is the RMS of graphene prepared with Ni?
Note that the graphene prepared with Ni/Cu indicates a 50% decrease in RMS as compared to the use of Ni alone as a catalyst, and that the graphene roughness does not seem dependent on the orientation of the starting substrate.
Q8. What is the main advantage of the graphene graphitization process?
This allows for a fast precipitation and graphitization of C released through the Ni silicidation reaction, leading to an extraordinary improvement in the uniformity and quality of the graphene obtained through Ni/Cu.
Q9. How much is the resistance of graphene on SiC?
By averaging measurements taken on five samples fabricated with the same graphitization procedure over separate runs, the authors can conclude that the sheet resistance of their bilayer graphene prepared on a SiC(100) layer with Ni/Cu is around 24.8 X/square 6 0.7 from sample-to-sample variation.
Q10. What is the Raman spectrum for graphene?
3. Example of Raman spectrum (graphene D, G, and 2D bands region) from graphene on SiC(100) prepared with ;8 nm Ni and ;15 nm Cu. Additionally to the low D over G band intensity, indicative of a good quality graphene, note that the 2D over G intensity ratio higher than 1 is indicative of a few-layer graphene.
Q11. What is the resistance of graphene to the underlying substrate?
preliminary results indicate that the adhesion of graphene to the underlying substrate could be an order of magnitude higher than the adhesion of a graphene layer grown ex situ and transferred onto a SiO2 layer on silicon.
Q12. How does the resistance of graphene be translated into a value?
for benchmarking purposes, it is meaningful to translate the sheet resistance of the bilayer graphene into a corresponding resistivity value by using the 0.9 nm thickness revealed by TEM.
Q13. What is the TEM thickness of the graphene?
Note that as TEM observation is extremely challenging, because of shadowing induced by the topography of the sample surface (;15 nm RMS, Table I) and the necessity for 20–30 nm thick metal layer deposition on the top of thesample (and thus of the graphene) for TEM preparation, it is not surprising that the nanolayer is not clearly visible along the whole cross-sectional image.
Q14. What is the ID/IG ratio of graphene on SiC(100)?
In fact, the ID/IG ratio of graphene on SiC (100) decreases from ;2 for 5 nm Ni catalyst alone to;0.8 when the Cu catalyst layer is added.
Q15. What is the adhesion value of graphene on SiC?
these preliminary measurements show that it is unlikely that the tested samples in Fig. 7(a) contained any interface with adhesion energies as low as 0.45 J/m2, which is the adhesion value reported in the literature for a monolayer graphene transferred onto a SiO2 layer.