Application of Tungsten as a Carbon Sink for Synthesis of Large-domain Uniform Monolayer Graphene Free of Bilayers/Multilayers

Wenjing Fang† , Allen Hsu† , Yong Cheol Shin§ , Albert Liao† , Shengxi Huang† , Yi Song† , Xi Ling† , Mildred S. Dresselhaus†,∥ , Tomas Palacios† and Jing Kong†,*

†Department of Electrical Engineering and Computer Sciences, §Department of Material Science and Engineering and ∥Department of Physics, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States  

Keywords: single crystal, monolayer graphene, Cu enclosure, carbon diffusion, carbon sink

DOI: 10.1039/C4NR07418A

  

As a student who has worked extensively on graphene (Gr) synthesis, I have received this paper with open arms for several reasons. First, Mildred Dresselhaus is one of my favourite characters in Carbon Science. She is now an Emerita Institute Professor at MIT with numerous awards and papers, proof of an exemplary career. Second, the growth process used is Chemical Vapour Deposition (CVD) with which I am pretty familiar. As far as I know, it is Pr. Ruoff's team which (I believe) first studied graphene synthesis on copper through a CVD process. A carbon precursor, popularly methane (CH₄), is exposed to copper at a high temperature and low pressure. A catalytic reaction then takes place, decomposing the methane and forming honeycomb-like organized crystal structures of sp² C=C bonds. Usually several crystals of this material are grown, and ideally a single crystal is desired, with as large a surface as possible.

What happened here? In short, it was found that tungsten metal can be used to trap carbon:

2W + C --> W₂C 

in their copper enclosures, which in turn allows for the synthesis of monolayer graphene.

Carbon leaks through the small gaps of the enclosures which then diffuses inside and acts as a C source. In their previous work they obtain mono and bilayer Gr inside the enclosure using this method. So far so good. The catch is that in most applications monolayer Gr is more valuable than multilayer. They found that by adding a tungsten foil in their enclosure there was no inside Gr growth, and that the outside presented only monolayer Gr!
The process is a bit more interesting than that, though. Their results expose the inverse behaviour of Gr growth on the enclosures with and without W foil. Without the foil the Gr layers grow and grow, making multilayer islands, a typical behaviour (often observed by other labs, and I still deal with it as well). However, with the foil, Gr grows outside for about 20 minutes, hinting at the formation of multilayered islands, which eventually start to disappear. Only monolayer graphene remains by the end of the 2h reaction time! But, after 5h, the multi-layer islands begin to appear again, and Gr grows inside as well. This points to some sort of saturation IMO. Some images are given in their Supporting Information.

They further investigated the effect of W on the monolayer using sequential flowing of ¹³CH₄ for 15 minutes, and ¹²CH₄ for the rest of the 2h. This isotope usage can give interesting data when using spectroscopy techniques which are dependant on the atom's mass, such as Raman, IR. Raman spectroscopy is probably the most useful carbon materials characterisation tool, which is what they used. The Raman peaks for the isotopes are easily distinguishable and they allow for ratio calculations. They discovered that only ¹³C is present in the monolayer, showing that the tungsten does not affect it, no carbon substitutions for the rest of the reaction time.
By the way, this isotope substitution technique has also been nicely used by Xuesong Li et al. from the Ruoff group on their study of graphene growth.

More data follows! Curiosity is key in science, thus they looked further into tungsten's role in this with an X-ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD) study of the chemical changes. It was discovered that W does not evaporate on the copper, nor on the Gr monolayer. However, they have found copper on the tungsten foil. Interesting, but to be expected, as it is known that copper exhibits a small degree of evaporation in thee conditions (1045⁰C being close to copper's melting point). Does this copper layer affect tungsten's behaviour? Apparently not. They took the analysis further by constructing an enclosure with one more level. A tungsten foil trapped in a copper enclosure (A) which is closed in another copper enclosure (B). This has several implications. By using a double enclosure geometry, the carbon quantity migrating towards the core of the enclosure gets limited at the inside surface of B. There, graphene starts to form, eventually completely passivating both the inside and outside surfaces. Any more carbon trying to go in will be blocked by a monolayer of the touchest material on earth. This eventually limits the amount of C arriving on the outside surface of A. At this level we have the behaviour previously presented with a single layer of Gr obtained outside, and nothing growing on the inside surface, due to carbon being consumed towards the formation of W₂C.


I find this multi enclosure design enthralling, makes me think of a sponge which filters as it absorbs deeper and deeper towards its core. Imagine having such enclosures with different entities (I'm thinking metals and their oxides, ceramics, and other inorganics) having specific interactions at each surface level. Growth of nano-rods/wires joined with graphene, and other novel nano structures seem plausible with this design. Tuning the surfaces with nucleation sites is one of the pathways I see.
The Dresselhaus group is one to be closely followed as far as carbon science goes. I am looking forward to more of their work.