Carbon and Formvar Extraction Replicas

Carbon and Formvar Extraction Replicas

For the study of nanometer sized precipitates, the sample preparation is critical and almost restricted to extraction replicas. Extraction involves the chemical etching (dissolving) of the matrix of an alloy without affecting the second phase so that this second phase can be collected and eventually be enriched onto the carbon foil. The general procedure is to make a replica, where a supporting foil is evaporated onto a freshly etched surface of a metallographic specimen and subsequently stripped by further etching. The precipitates stick onto the supporting foil by electrostatic forces. The replica is cleaned and picked up onto a TEM grid.
Thin foils are less frequently used because precipitates are sometimes hardly observed or distinguished from other features as e.g. dislocations, thickness fringes and bend contours. Moreover, Fe-containing precipitates cannot be studied quantitatively in the Fe-matrix. Thin foils do however have an advantage when the interaction of the precipitates with the matrix needs to be studied.

The advantages of replica’s over thin foils are:
• Pre-concentration of precipitates on the foil.
• Easy EDX analysis (no absorption effects, thin film approach applicable and no Fe-matrix effects, lower backgrounds).
• Fe-containing precipitates can be studied.
• Diffraction studies are more convenient and tilting into a prominent zone axis orientation is easier.
• No magnetic matrix effects.
• Very large analysis area’s can be selected.
• Very fast preparation time compared to thin foils.
• Phases that differ from ferrite like e.g. bainite, cementite, martensite,… can also be extracted by replica’s.

The major disadvantages of the replica versus thin films are:
• No orientation studies of precipitates in the matrix are possible.
• Texture effects can influence the etching of the sample and might influence the precipitate collection.
• Quantification of carbon is not possible when carbon is the supporting film. It is theoretically possible on thin films, but the C-redeposition effects should be taken into account. C-contamination is typically a problem at locations were an electron beam strikes the analysing area at intermediate vacuum conditions (e.g. in a TEM). Cooling the sample will reduce but not exclude this contamination problem. The accuracy of carbon quantification will therefore be inferior to other elements.

Procedure 1
1.) The precipitates were extracted from matrix by producing direct extraction replicas. The replicas were prepared by first metallographically polishing the specimens down to 1 µm diamond paste followed by ultrasonic cleaning. The specimens were then pre-etched in 2% nital for 15 seconds and coated with carbon. The carbon film with attached precipitates was released from the specimen by immersion in 10% nital, floating off the specimen in distilled water, and fragments of the carbon replica were collected on copper grids. Fresh solutions were used at each step of the replicating process in order to eliminate any cross contamination.

2.) The precipitates were extracted from matrix by producing direct extraction replicas. The replicas were prepared by first metallographically polishing the specimens down to 1 µm diamond paste followed by ultrasonic cleaning. The specimens were then pre-etched in 4% nital until the surface is dull and coated with carbon. The carbon film with attached precipitates was released from the specimen by immersion in 4% nital, floating off the specimen in methanol ad rinced in water. Fragments of the carbon replica were collected on copper grids.

3.) After metallographic polishing, the steels were etched with PICRAL in a first step and with NITAL in a second stage. A film of biodene is then deposited on the sample (wetting with acetone). After drying, the film is torn off, taking with it an extraction of precipitates. A thin carbon layer is evaporated under vacuum on the film of biodene. The biodene film is dissolved with dichloroethane (80%) and methanol (20%). The remaining carbon foil is collected onto a Cu TEM grid.

Procedure 2
1.) Another procedure using chemical etching with an alcohol solution containing 10% hydrochloric acid, 6% acetic acid and 2 g of picric acid salt; after washed and dried, a dilute solution of 1% polyvinyl formal (“formvar” resin) in chloroform is applied to surface. This deposits a thin film, which when hardened, is baked with softened cellulose acetate. Once dry, the composite film can be stripped off, coated with carbon on the impression side, cut it into small pieces, mounted on copper grids and washed in acetone. The acetate foil is dissolved by the acetone, leaving behind the extracted particles embedded in the carbon film. Both precedures for C replication are valid. Care should be taken however with strong acids in order not to dissolve the precipitates.

2.) As an alternative, the evaporated amorphous C-film is used as a semi-permeable membrane. A sample of 25 by 25 mm is polished and coated by a C-film. The sample is then mounted into the electrochemical cell unit. The matrix is electro-chemically dissolved in an organic electrolyte, AcAc or TEA. Fe will dissolve (yellow colour), diffuse through the C-layer while the replica retains the precipitates. After the electrolysis, the larger fraction of the electrolyte can be removed using a pipette or by slow decantation of the solution. The film is removed by a soft methanol spray and collected in a petri dish. The replica is washed in methanol and collected onto a copper grid. The advantage of this method is the ‘softer’ etching of the steel. No strong acids are used which means less risk of dissolving the precipitates. For chemically stable precipitates such as TiC, NbC and carbonitrides this method brings no added value, and is therefore too time consuming. Chemically unstable precipitates such as cementite and different sulphides do benefit from this method.

To enable the study of precipitated carbon, the replica must be carbon free. Therefore ARSA developed a method where the precipitates under study are extracted from the sample on a carbon and nitrogen- free AlOx replica. It is fully described in C.P. Scott et al., Quantitative analysis of complex carbonitride precipitates in steel, Scripta Mat. 47 p. 845 (2002).

Procedure 3
1.) The preparation of TEM extraction replicas based on oxides of aluminum AlxOy and bilayers Al2O3/Al films were conducted according the ARSA procedures. These films were backed with softened cellulose acetate and afterwards removal the hardened composite films. After cut it into small pieces and mounted on copper grids, the cellulose acetate foil was dissolved in acetone. The acetate foil is dissolved by the acetone, leaving behind the extracted particles embedded in the carbon film. In spite of the X ray detector used in this study allowing to detect C and N X rays, only the ratio of metallic elements in the particles was determined, with no measurements (quantification) of the C or N levels of the precipitates due to the inaccuracy of the results.

Reference: A. De Vajt, P. Paiva, L. Barrros, D. Ravaine, C. Scott, J. Fluch, Nano precipitates in steel - NANOPREC, European Commission, Technical Steel Research, Final Report, 2006, pp. 21-23.