Site: SONY

Date Visited: October 4, 1991

Report Author: M. Thompson





T. Yamada

Senior General Manager, R&D Corp. Planning


Chief Scientist


LCD Division

A. Yasuda

Molecular Materials Research Dept.(Ferroelectric)

S. Umeya

Asst. Manager, R&D Planning

K. Watanabe

Manager, Project Promotion

I. Hiroyoshi



Sony entered the LCD field later than other companies and, therefore, has chosen a different strategy from the mainstream. It is focusing on small displays and poly-Si technology.

Sony started R&D on LCDs 20 years ago with TN direct drive segment type.

Their product history:

  1. TN - 1978 Audio Peak Level Meter in Cassette Recorder. This had a high cost and was discontinued after two years.
  2. In 1986 they produced a Laser Thermally Addressed Smectic LC light valve. The product is on the market and is high-cost and high-resolution (8,000 x 10,000).
  3. Their most recent and current product (to be announced) is the color LCD viewfinder based on high-temperature poly-Si process. They believe a-Si TFT is best for large-size AMLCD, but they are focused on small-size poly-Si displays to overcome the interconnect problems with high-resolution a-Si. Also, they have had a 10-year history in poly-Si. They produce a super-thin- film transistor (see figure Sony.1).

Figure Sony.1.

The multi-insulator structure is used to increase breakdown voltage. High- temperature processing is used to produce this device. CMOS circuits are used for the peripheral structure, and an LCD transistor structure is used for the pixel TFT to lower the leakage current. The implanted LDD regions are self-aligned. They use a three-phase switching gate to transfer RGB signals simultaneously. Therefore, the clock frequency is 1/3 and, thus, the power dissipation is reduced. However, signal delay is required, so a sample-and-hold has to be added.

They would like to make the viewfinder even smaller to enable more economic production on quartz wafers.

Device spec of "old" viewfinder:

We saw the "new" viewfinder design, which had higher resolution. They discussed the limits and issues of reducing the pixel size. Transistor size is 3x5 m. Metal lines and spaces can be reduced, but fringing fields in the LC have to be reduced by shielding with black matrix.

At present, they are evaluating projector applications of their devices, but they believe such applications may not be cost-effective and, thus, larger area low- temperature poly-Si may be the best approach.


They plan to use low-cost substrates for large-area poly-Si by using Excimer laser annealing. They need a glass substrate with 2-3 x 10(superscript -6)/degrees centigrade and low compaction < m/100m-hr.

The alternatives for low temperature poly-Si processes are:

  1. LPCVD a-Si - Deposit 550 degrees centigrade, SPE annealing @ 600 degrees centigrade
  2. LPCVD a-Si - 600 degrees centigrade
  3. PECVD a-Si - (<300 degrees centigrde)
  4. Sputtered Si - (Td<200 degrees centigrade)

Their choice is PECVD a-Si and laser annealing. With this approach, they have a-Si TFT in the matrix because of ease of obtaining low off currents and laser-annealed poly-Si for the drivers because of high on current. They laser scan the peripheral region only, resulting in a reasonable process throughput.

Deposition conditions for PECVD were:

The optimum deposition temperature for a-Si is 250 degrees centigrade, whereas for optimum laser annealing to obtain good poly-Si, the deposition temperature is 150 degrees centigrade. The annealing source is 308 nm XeCl pulsed laser:

Laser annealing can also be used for doping activation. This processing technique has only local heating and is an ultrafast process (20 ns) with high throughput.

The beam size is 5 mm and, for uniform crystallization, 10 pulses per point are used. Beam pulse overlapping gives uniform crystallization.


In a study of 7059 glass, Sony found that conventional heat treating at 600 degrees centigrade results in 15 m compaction, whereas, after this laser treatment, only 3 m of compaction occurs with further heat treatment at 600 degrees centigrade. They are concerned about the cost of 7059 glass and are considering using coated soda lime glass, in which case they must maintain the process temperature <350 degrees centigrade. They have measured the substrate temperature during laser annealing and found a maximum temperature of 150 degrees centigrade with a long decay time. They haven't measured the substrate with multiple pulses, but they claim that it is not a problem.

Sony obtains poly-Si grain size between 100 and 800 angstroms for a 200 angstrom-thick film, but they claims an even more important attribute in lattice matching between crystals, resulting in few defects.

Laser-annealed impurity doping steps:

Recrystallization would probably be accomplished first, followed by the laser impurity doping steps.

Figure Sony.2

They claim the great advantage of poly-Si is that a high aperture ratio can be achieved because the poly-Si transistor can be small compared with an a-Si TFT. They also claim the COG cost is half of the panel cost for a 1-inch product.

The main technology for the next few years, they think, will be a-Si followed by LT poly-Si. They will focus on low-temperature poly-Si for large areas and, in particular, work on Excimer beam shaping and homogenizing.



1000 x faster than TN
Wide viewing angle
Memory effect (greater than 1000-line display can be achieved)


Speed needs to be improved to obtain video rates, defect-free,
molecular orientation.

Sony's approach is to use new materials to achieve faster response and wider temperature range. SiO oblique evaporation is used for the alignment layer, and a two-pulse drive method is used to address the cell.

Sony Fluorinated Compound vs. Conventional

Temp range

-30 to 75 degrees centigrade vs. -20 to 65 degrees centigrade


33 vs. <20

Response time

56 s vs. 100 s

The response = 28 s delay + 28 s switching = 56 s. The conventional chevron structure has zigzag defects, whereas the bookshelf structure is more ideal. The oblique evaporation gives a 35 degree pre-tilt angle and a bookshelf structure which results in a contrast ratio >100. Sony's rub-free process also results in a reduction of ions.

The response time prediction is shown below:

Year : Response time (s)
'91 : 56
'93 : 32
'98 : <16

The bookshelf structure is better for shock resistance.

They believe the first application for FLC is office automation. Of the 5-10 companies working on FLC, Canon is the leader.

Their decision to make rather than buy LCDs is driven by several considerations:

  1. Performance of LCDs has to meet Sony standard
  2. Costs
  3. For small-size LCD no one has achieved the required quality
  4. Develop interdependent relationships with the suppliers

FLCs are a good candidate for large-size FPD. In the long term, 10 companies will survive in AMLCD--5 major companies and 5 second tier. Sony has a goal of first being number one in manufacturing small-size poly-Si displays, and then it will increase the size.

Published: June 1992; WTEC Hyper- Librarian