I agree with the suggestions. Back in 1980 I routinle did this on 100 to 2000 nm ito on silicon. We deposited 4 silver pads around the perimeter for Vander Pauw measurements. Conventional tungsten or CuBe probes do not make very good contact to Ito. Metallized rubber, In wire, soft large surface area probe tips work well.

Eventhough you took care, first check with contacts and check is the contacts are ohmic or not. With addition connections also you check. Then change the source range, Current/voltage measure range and compliance. check in which mode, you are working. For sheet resistance, source function should be variable. Another thing, don't forget to make sensing mode to 4-wired.

With such a thin layers one should also take care about shape of electrodes. For sheet resistance measurement in your situation I would consider using sphere shaped contacts rather than point probes.

Yet another important point, which is often omitted, is the requirement of probe-to-sample-edge distance not smaller than ten times the pitch between probes.

I suggest depositing gold (or chromium) contact points by sputtering through a mask. Make contact with these while ensuring you do not penetrate through either the gold layer or your film. If your film is not dense or homogenous you may have trouble getting a consistent result from small contact points, in which case you may consider depositing larger patches or even strips.

Thanks everyone for your input. @Dorian. are you suggesting that i use the Van der Pauw configuration. I thought of that but I have a lot of samples (ITO, ZnO,AZO) on both glass n Si/SiO2 spacer for various Ar and O2 partial pressures and RF powers. Besides my samples are approx 2 x 2 cm in size. I am using the largest tip radius for my 4PP equipment with minimum input current of 10 nA. Its really a challenge getting consistent results. I have managed to measure all the films with most of them giving me a sheet resistance which is out of range of the test unit. I guess I should just work on thicker films. Any suggestion of how to improve the conductivity and transmission of ultra-thin TCOs films is most welcome.

@ Mike, any idea where I can order 4PP with spherical tips?

Lucas Signatone had those in offer:

http://www.signatone.com/

Maybe you are already doing it but if not, I would recommend you to use 4-wires measurements to get rid of probes and contact resistances. I do not think using Van der Pauw configuration is necessary if your samples has long strip shape.

@F.Molina-Lopez

That's exactly what 4 point probe is about.

Van der Pauw on the other hand is a method to deduce sheet resistivity from resistance measurements

I agree with the suggestions. Back in 1980 I routinle did this on 100 to 2000 nm ito on silicon. We deposited 4 silver pads around the perimeter for Vander Pauw measurements. Conventional tungsten or CuBe probes do not make very good contact to Ito. Metallized rubber, In wire, soft large surface area probe tips work well.

Thanks you all for all your suggestions. I will try to implement the Van der Pauw configuration

I respect all the above answers.

If you have probes not available with you then you can try this one.

I simply did it using 4 copper wire contact with silver paste. You can try both linear as well as vander pauw method according to your sample geometry. If it is rectangular shape linear contact is best. If it is not rectangular then vander pauw may be the best option.

@ Bruno thanks. I do have the book. Its a great resource. I will check out the papers.

Hi Jephias,

I couldn't agree more with the comments above.

Deposition of metal ohmic contacts may be difficult for your varied and large number of samples as you have expressed. In such a case you may use the ready made probes available or make one in-house. I have included some links for probes. Even though the probes are spring-loaded, care would be required as your films are too thin.

The book by Schroder will be handy for the calculation, especially the correction factors.

Coming to the problem of instability in your measurements. Kindly make a note of the instrument/SMU's limits, especially the minimum current that can be sourced, maximum voltage that can be forced, input impedance of the voltmeter. I have attached a small document on this account. You may estimate compliance limit and set the current or voltage accordingly. Modern SMUs display the compliance limit whenever the measurement goes out of limit. Assuming, that the resistance is fixed for a sample, you may decrease the current and/or increase the output voltage of SMU to resolve your out of limit issue.

Another factor for instability is time (crude term, but let me at least try to justify it). Check the NPLC (number of power line cycle) of the measurement condition. Increasing NPLC increases accuracy.

Give enough time before taking measurements, due to RC delay in highly resistive samples, the current needs some time to rise to the saturation value. Once stable, measure, say 10 readings as a particular point of measurement and then get the average value.

I hope this helps. It helped me when I was working with resistive InGaZnO samples.

Further, in order to increase the conductivity, the answer is not a single one. In ITO, the main contributor in conduction is the doping of Sn4+ to In3+, in addition to the contribution from the oxygen vacancy. However, in ZnO the origin of conduction is due to oxygen vacancies, so reducing the O2 partial pressure during deposition will increase the conductivity.

A small note:

In some TCOs, reducing the O2 partial pressure (during deposition) beyond a certain level of pressure decrease the conductivity in spite of increasing it due to generation of defect and its associated deep levels within the band gap.

Regards,

Sonachand A.

http://www.signatone.com/applications/4ptprobe.asp

http://www.jandel.co.uk/downloads/Jandel-Brochure.pdf