Wonders Of The World

Wonders Of The World

 

 

 

 

 

Most Beautiful Buildings In the World

Most Beautiful Buildings In the World

 

 

 

 

 

Tips for Planning a Trip with Kids

If you think pulling together a complex trip is the kind of organizational nightmare that’ll make you feel like you’re in a bad remake of a National Lampoon’s Vacation movie — don’t worry.

Here are a few strategies we picked up along the way:

Don’t overplan. It’s possible to do too much research and plan for each moment. Leave some things to serendipity, otherwise your itinerary will get too cluttered and nothing about your trip will be spontaneous. And the spontaneity can be half the fun. After a few cross-country trips, we felt more comfortable winging it, and we made some great discoveries as a result. Like Deschutes National Forest, the dormant volcanic near Bend, Oregon, which we didn’t know about until we decided to pull over for a picnic lunch.

Simple is better. No matter how complicated your itinerary, there are ways to streamline the process to include the important information and exclude the extraneous details that will be insignificant until you arrive at your destination. For example, we whittled a list of recommended restaurants down to a hyperlink that we could access later, instead of trying to save everything to a document.

Share your itinerary. Letting people know where you are is helpful when you’re on a multi-stop itinerary. When your hosts know where you’re coming from, where you’re supposed to be, and where you’re going, you can avoid all kinds of trouble. That’s why car rental companies ask for your flight itinerary – they aren’t really interested in which airline you’re using. They want to track your flight so that if your plane is late, they can hold your rental vehicle.

Pad your schedule. Remember the 15-minute rule: For every one hour on the road you should plan 15 out of the car. (And not necessarily every hour; we would drive three hours and pull over for half an hour or so, give or take.) You need to stretch, eat and take a break from sitting. Also, you can never plan too much time for meals and national parks. If someone recommends half a day in a national park or forest, take a whole day. You won’t regret it.

Be flexible. Life is too short (and so’s your trip) to lock yourself into a schedule. If you see an opportunity to do something interesting, to make an unexpected detour, or to extend your trip – do it! Just don’t forget to share your new itinerary with everyone along the way. Our favorite detour? The Grand Canyon, a day-long diversion on a road trip from California back to Florida. It was way off our designated route. And so worth it.

Photo Gallery: Best Family Trips

Namib Desert

Travelers to the Namib Desert run down the steep face of a coastal dune in Namibia’s Erongo region.

________________________________________________________________________________________

Amalfi Coast, Italy

The town of Positano is one of many along the Amalfi Coast, a famed stretch known for its curving roads and breathtaking vistas.

__________________________________________________________________________________________

Great Wall of China

Colorful parasols shield children from the sun on China’s Great Wall, which has over 4,000 miles of historic sections to explore.

_________________________________________________________________________________________

Boundary Waters Wilderness Canoe Area, Minnesota

An early morning kayaker breaks the stillness of Birch Lake, part of Minnesota's Boundary Waters Wilderness Canoe Area.

_________________________________________________________________________________________

Victoria Falls, Zambia

A microlight flight brings adrenaline junkies close to the roar of Victoria Falls, which splashes between Zambia and Zimbabwe.

____________________________________________________________________________________________

Galápagos Islands, Ecuador

Galápagos sea lion (Zalophus wollebaeki) underwater at Champion Islet near Floreana Island in the Galápagos Island Archipelago, Ecuador. Pacific Ocean.

____________________________________________________________________________________________

Nags Head, North Carolina

Jockey's Ridge State Park on Nags Head, North Carolina, has the tallest natural sand dune system in the eastern U.S.

___________________________________________________________________________________________

Sequoia National Park

California's Sequoia National Park is home to some of the biggest trees on the planet. The largest measured sequoia is nearly 275 feet (84 meters) tall and more than 100 feet (30 meters) around.

__________________________________________________________________________________________

Petra, Jordan

At Petra, candles light the way to Al Khazneh (the Treasury), whose function in ancient times is still unknown.

__________________________________________________________________________________________

Yampa River, Utah

Boaters float past huge Steamboat Rock, which sits at the confluence of the Yampa and Green Rivers in northeastern Utah.

_________________________________________________________________________________________

Samburu National Park, Kenya

Elephants roam Kenya’s Samburu National Park, one of many destinations for viewing wildlife in the East African nation.

____________________________________________________________________________________________

Sayulita, Mexico

Sayulita, a small Mexican fishing village near Puerto Vallarta, is popular with surfers.

_____________________________________________________________________________________________

The World’s Top 10 Most Expensive Cars for 2012-2013

10. SSC Ultimate Aero $654,400

1

Don’t let the price tag fool you, this American made car is actually the 3rd fastest street legal car in the world with a top speed of 257 mph and reaching 0-60 mph in 2.7 seconds. It is estimated that only 25 of this exact model will be produced.

9. Pagani Zonda C12 F $667,321

2

Produced by a small independent company in Italy, the Pagani Zonda C12 F is the 9th most expensive car in the world. It promises to deliver a top speed of 215 mph and go from 0-60 mph in 3.5 seconds.

8. Ferrari Enzo $670,000

3

The most popular supercar ever built. The Enzo has a top speed of 217 mph and able to go from 0 to 60 mph in 3.4 seconds. Only 400 were produced and it is currently being sold for over $1,000,000 at auctions.

7. McLaren F1 $970,000

4

In 1994, the McLaren F1 was the fastest and most expensive car. Even though it was built more than 15 years ago, it still has an unbelievable top speed of 240 mph and reaching 60 mph in 3.2 seconds. Even today, the McLaren F1 is still top on the list and outperforms other supercars.

6. Hennessey Venom GT Spyder $1,100,000

5

What makes the 2013 Hennessey Venom GT Spyder is its price tag, a compelling 1.1 million dollars. There is no other convertible in the world today that can match-up with this car. Not to mention that it goes from 0-60 mph in 2.5 seconds.

5. Zenvo ST1 $1,225,000

6

Able to reach 60 mph in 2.9 seconds and a top speed of 233 mph. The Zenvo ST1 is from a new Danish supercar company that will compete to be the best in speed and style. The ST1 is limited to 15 units and the company even promised “flying doctors” to keep your car running.

4. Maybach Landaulet $1,380,000

7

The Landaulet is the most expensive sedan on the market and it can go from 0-60 mph in 5.2 seconds. It is one the most luxurious cars ever made, this comes with a convertible roof that fully opens at the rear. This Maybach is made especially for CEOs and Executives who have their own personal driver.

3. Koenigsegg Agera R $1,600,000

8

The Agera R can burn 0-60 mph in 2.8 seconds, reaching a maximum speed of 260 mph. It is equipped to reach 270 mph, but the supercar is electronically limited to 235 mph. You will need to sign a waiver, only then does the company unlock the speed limiter for a given occasion.

3. Lamborghini Reventon $1,600,000

9

The most powerful and the most expensive Lamborghini ever built is the third on the list. It takes 3.3 seconds to reach 60 mph and it has a top speed of 211 mph. Its rare (limited to 20) and slick design are the reasons why it is expensive and costly to own.

2. Pagani Zonda Cinque Roadster $1,850,000

10

One of the most exotic cars out there is also one of the most expensive. It can go from 0-60 mph in 3.4 seconds with a top speed of 217 mph.

2. Aston Martin One-77 $1,850,000

11

The name “One-77″ says it all: beauty and power in One, limited to 77 units. With 750 hp, it is able to travel from 0 to 60 mph in 3.4 seconds and reaching a maximum speed of 220 mph.

1. Bugatti Veyron Super Sports $2,400,000

12

This is by far the most expensive street legal production car available on the market today (the base Veyron costs $1,700,000). Capable of reaching 0-60 mph in 2.5 seconds, the Veyron is the fastest street legal car when tested again on July 10, 2010 with the 2010 Super Sport Version reaching a top speed of 267 mph. When competing against a Bugatti Veyron, you better be prepared!

The Dazzling Evolution of Energy-Efficient Lighting

Here's a bright idea: light bulbs that use 75% to 80% less energy.

Compact fluorescent bulbs (CFL) and light-emitting diode bulbs (LED) are quickly becoming the new norm. These newer bulbs have a bright future since the U.S. Government passed the Energy Independence and Security Act of 2007, which requires a systematic reduction of inefficient incandescent bulbs, culminating in 2014.

Currently, CFLs — the spiral-shaped bulbs — are the leading alternative to incandescent bulbs. LED bulbs are the longest lasting of them all, but they come at a significantly higher cost compared to CFLs.

For more information about the past, present and future of light bulbs, check out the infographic below, made by Osram Sylvania.

Why Buy LED Light Bulbs?

The incandescent light bulb has been the standard lighting option for nearly 100 years. In recent decades, the compact fluorescent lamp (CFL) has gained popularity because it's more energy efficient and lasts longer. However, the bulb contains mercury and takes a while to shine at its brightest.

LED (light-emitting diode) light bulbs are now becoming an even more practical lighting solution. On average, about a dozen watts from these energy-efficient LED bulbs provides the same amount of light as a 60-watt incandescent. This makes it so you can save on electricity without scrimping on the amount of light you have. Additionally, these lights shine their full brightness as soon as you turn them on. Gone are the days of searching for your shoes in the dark closet as the CFL bulb slowly warms up.

Some LED light bulbs last up to 50,000 hours, which means you'll be replacing them less than you would an incandescent or CFL bulb. This is nice if you have a light socket in a hard to access area, such as a high ceiling. Another great aspect of LED light bulbs is that they don't contain mercury, so when you finally replace your LED lighting, you don't need to dispose of them as hazardous material.
In our research and testing, we found Philips L Prize, Definity A19 and Sylvania 12-watt to be the best LED light bulbs. For more information, read our articles about LED lighting.

LED Light Bulbs: What to Look For

When you're shopping for an LED light bulb, it's important to find one that provides the amount of light you need, as well as the color of light you like. You'll also want to consider how the light emits from the bulb, and the dimension of the bulb and base. The bulbs in this review are comparable to a 60-watt incandescent bulb.

Features
A standard 60-watt incandescent bulb puts off about 800 lumens. The more lumens, the brighter the light is. LED light bulbs provide many lumens for few watts compared to incandescent bulbs. Since this is the case, it's better to find a bulb that has low wattage but high lumens because it will save you on your energy bill.

You also want to find an LED light bulb that offers a long lifespan. Most offer between 25,000 and 50,000 hours of light. When you look at a Lighting Facts label on the light bulbs, they typically list the lifespan in years. This is based on a standard of using the light for three hours per day each day of the year.

When you're looking for an LED light bulb, another thing to consider is the color of the light. Depending on where the light bulb lands on the color temperature spectrum, lights can be all colors: red, yellow, green, blue and shades of white. The shades of white range from warm to cool white. The lower the color temperature, the more yellow your white light will appear. If the light is higher on the color temperature scale, it will appear to cast a cool, bluish hue. Another aspect of the color is the color rendering index (CRI). The higher the CRI number, the greater the number of objects lit by the light bulb that will appear natural.

The bulb should also be UL listed and comply with part 15 of the FCC as a Class B device, meaning it won't interfere with radio frequencies. It's also an added bonus if the light bulb is Energy Star Qualified.

Design
The dimensions and weight of LED bulbs are not the same as a standard incandescent bulb. LED light bulbs are on average a quarter of an inch taller. The average diameter of these LED lights is similar to that of incandescent bulbs but varies depending on the model. LED lights are also heavier than incandescent bulbs, so you'll want to be sure that your light fixture can support the extra weight.

Most LED light bulbs can't be fully enclosed in a light fixture because heat decreases the life of the light bulb. If you plan to use your LED light bulbs outside, you'll first want to verify that they can withstand damp outdoor conditions. If you want to dim your lights with an LED light bulb, you'll need to have one that the manufacturer has specifically designed to perform as a dimmer.

The beam spread is another thing to consider. While incandescent lights put off light in all directions, LED lighting typically sends its light in one direction. The best LED light bulbs that are comparable to 60-watt incandescent bulbs distribute the light around the bulb as well as from the top.

Help & Support
Chances are you won't need to be in contact much with the manufacturer to use your light bulbs. However, having a practical return policy and warranty are importance since these light bulbs cost more than incandescent and CFLs. A good LED light bulb should come with at least a three- to five-year warranty.

LED lighting presents a new, more environmentally friendly option. It's long lasting and can save you money over an extended period. These lights are heavier than other lighting options, and the bulb is on average a little taller than a standard light bulb; however, the base can fit in standard light sockets. These lights will keep going long after incandescent and CFLs would have stopped working, and they'll save you money on your electrical bill.

If you're savvy about saving money and the environment, you've probably taken an interest in LED light bulbs because they're even more energy efficient and environmentally safe than CFL and incandescent bulbs. Many LED bulbs also have unique designs that loosely resemble traditional incandescent bulbs but have embellishments around the bulb itself.

More Light, Less Heat

Light-emitting diodes (LED) are semiconductors. As electrons pass through this type of semiconductor, it turns into light. Compared to incandescent and CFL bulbs, LED lights are more efficient at turning energy into light. Therefore, less of the energy radiates from the bulb as heat. This is why LED bulbs are cooler during operation than incandescent and CFL bulbs.

As the light-emitting diodes create light, they warm up quite a bit for their size. LEDs are heat sensitive, so it's important that the heat move away so that it doesn't damage the semiconductors. In order to do this, these lights need a system to keep cool. Most LED lights have a heat sink plate that moves the heat away from the light-emitting diodes through the heat sink plate. Manufacturers make the heat sink plate out of a variety of materials, but it's commonly made from aluminum. Frequently, the heat sink becomes part of the design of the bulb.

An LED light bulb's heat sink usually weighs several ounces and can become hot once you turn on the light. From the heat sink plate, the heat moves into the air surrounding the bulb. If you place the bulb in an enclosed fixture, it keeps the heat from effectively moving away from the heat sink, raising the temperature around the light-emitting diodes. This, in turn, causes the LEDs to overheat, shortening the life of the bulb.

Directional Light

Many LED bulbs have light-emitting diodes that all shine in one direction. This results in a bulb that directs most of its light toward the top of the bulb. If you place this type of bulb in a table lamp, you can see that most of the light ends up on the ceiling with very little refracted onto the table. In an attempt to make LED bulbs more like incandescent bulbs, omni-directional LED lighting is becoming more common. This lighting distributes the light evenly around the bulb. Commonly, it's done by bouncing the light off reflective plates inside the bulb.

An Aging Bulb

As LED light bulbs get older, they don't just burn out. Instead, they grow dimmer. The industry standard for LED light bulbs is that they should last for at least 25,000 hours with at least 70 percent as much brightness as they have when they are new. Below 70 percent is the point at which the industry decided the decrease in brightness is noticeable.

LED light bulbs are relatively new on the market, so they haven't been through the test of time – especially those that boast a lifespan of 50,000 hours. Although most of these lights last for about 25,000 hours, their warranties only cover about three years. This is perhaps because if you run the LED lamp continuously, it will only stay within the 70 percent range for a little less than three years. However, if you look on the Lighting Facts label of many LED lamps, it gives a life based on years. This is because the industry standard is three hours of use per day instead of 24 hours.

Manufacturing LEDs on large diameter substrates: What’s the holdup?

In the search to cut the costs of manufacturing LEDs, switching to large diameter (6-8-in, 150-200-mm) substrates is often brought up as a key solution. In fact, it is mentioned so often that one might assume a mass transition to large diameter has already occurred. Such a transition has been slow to happen due to both technical and logistical challenges. But as we will discuss, the advantages will ultimately be significant enough that the LED manufacturers will move to larger substrates, reduce component costs, and further accelerate the adoption of solid-state lighting (SSL).

Fig. 1.

Analysts are predicting that for 2013, less than 20% of production will be on 6-in wafers, with 8-in not even showing up in significant numbers this year. Even three years from now, large diameter is only predicted to be breaking the halfway point of all substrate sizes.

This may come as a surprise because the demand for large diameter should be very high – it has often been cited as a fundamental cost saver to drive down LED chip prices. With the industry-wide search to drive down cost significantly, and large diameter seen as a key way to accomplish this, why aren’t more chip producers making this switch?

To answer this question, we will look at three areas. First, what are the true benefits of large diameter? Second, if the benefits are truly great, then why aren’t more manufacturers switching? Finally, we’ll look at some of the potential disruptions that might bring quicker large diameter adoption, such as c-axis CHES (controlled heat extraction system) technology – along with the prospect of 8-inch substrates.

More LED chips

We’ll begin by looking at the main advantage of moving to larger wafers – more LED chips. Yes, this is the biggest advantage, but it’s also unfortunately often overstated in this way: a large diameter 6-in wafer has nine times more surface area than a 2-in wafer on which to form LED chips (Fig. 1).

While the prior statement is certainly true in regards to the simple surface area of the wafers, the suggestion that you get 9× more chip throughput by simply using 6-in wafers sounds too good to be true – and it is. What’s the real story? In order to answer that, we need to look closely at the layout of LED chips, both on the wafer and as a group of wafers in the MOCVD (metal organic chemical vapor deposition) reactor where LEDs are formed.

In addition to 6- or 8-in wafers simply being larger, we have to consider several other factors to get a true picture of the benefit of “more LED chips.” These factors are exclusion zone, LED chip shape and size, and MOCVD reactor layout. We’ll explain these one at a time, then feed them into a true comparison simulator that will give us a much more reasonable look at the number of LED chips supported by various wafer sizes.

Exclusion zone

We will first look at what’s called the exclusion zone on a wafer. During epitaxy, LED material is not properly formed in this area, meaning these chips shouldn’t be counted because they will not result in good LEDs. For our LED chip calculator, we are using an industry standard 3-mm exclusion zone, which is shown as red chips in Fig. 2. Note that the chips on the extreme edge of the wafer – that are actually hanging off the wafer if they were full rectangles – are not going to be counted at all for our simulation.

One important characteristic of the exclusion zone is that it is 3-mm from the edge regardless of wafer diameter. This fact means that the large diameter wafers have larger exclusion zone areas. However, as a percentage of the total wafer surface area, the large wafers have a smaller proportion of their area in exclusion zones.

So you can see how a 6-in wafer that has 9× more gross surface area actually has more than 9× more net area (gross area minus exclusion zone). The advantage results in 6-in wafers having 10.3× more net area, and 8-in wafers having 18.8× more, both as compared to a 2-in wafer.

Fig. 2.

We also have to account for the rectangular footprint of LEDs. They don’t perfectly fit in the round shape of the wafer – some LEDs will be lost by partially crossing into the exclusion zone. In a similar way to the exclusion zone, these losses are a higher percentage of the total for the smaller wafers. The final advantage is shown in the chart in Fig. 3 that is based on 45x45-mil (thousandths of an inch) LEDs, including the spaces between chips. The result is slightly higher gains in chip count compared with area – 10.9× for the 6-in and 19.8× for the 8-in wafers.

MOCVD reactor layout

At this point, we’ve seen that a 6-in-diameter wafer actually holds slightly more than the often-quoted 9× more LED chips compared to a 2-in wafer. But now we have to consider that LEDs are grown in groups of wafers in an MOCVD reactor.

The LED epitaxy process is one of the most expensive and time consuming of all the steps that go into the final delivery of an SSL product. The input is a group of wafers, and the output is thousands of LEDs on those wafers. What we are seeking to answer is how switching to large diameter will change that LED count after the epitaxy process. Of course, yield – a measure of chips that function correctly – matters too, but we will look at that later.

We’ve already said that you shouldn’t expect the chip count you get after epitaxy to jump by a factor of nine, and now we’ll see why. The primary reason is the fact that so many more small-diameter wafers can fit in the reactor chamber. In a typical MOCVD reactor configuration, 56 2-in wafers can be loaded. In the same reactor only eight 6-in wafers will fit. That’s a ratio of 7:1 in favor of small diameter.

So to simply break even in the final count, each 6-in wafer would need to hold 7× more LED chips than a single 2-in wafer. However, we’ve already seen that a 6-in wafer has almost 11× more LED chips. Put in other terms, the 6-in configuration results in 55% more LED chips (1.55×). This is the final true advantage we’ve been looking for. While this is much less than the 9× (900%) figure that we started with, it is still a very significant improvement in the number of LED chips you get for the same cost of time and money for an MOCVD run. You can compare a typical MOCVD layout for small and large diameter wafers and their respective chip counts in Fig. 4.

LED chip size

We do need to consider another factor, and that is LED chip size. For our calculations we’ve used 45×45-mil rectangles, including the street width, or spaces between the chips. This size – around one square millimeter – is typical for high-brightness LEDs and is therefore a good comparison. However, as chip sizes increase the advantage for large diameter wafers also increases slightly. For example, if you use a 60×60-mil rectangle, the advantage for 6-in wafers increases to 58%.

Fig. 3.

So far we have focused on comparing 2-in to 6-in wafers, but we should discuss other sizes as well. Today’s LEDs are also made on 3- and 4-in wafers in large numbers. What is the relationship with these other sizes? Moving from 2-in to 4-in only gives 14.7% advantage, using a standard layout of 14 4-in wafers in the reactor. The gain from 4-in to 6-in is much more significant at 35.2%.

For 8-in substrates, the advantage is a very large – 77% more LED chips over 2-in – and that is only from the five 8-in wafers that can fit in a typical MOCVD reactor. Comparing a move from 6-inch to 8-in directly, there is a 14% gain.

We now have an accurate view of the advantage large diameter brings to the LED chip count: 55% more for 6-in and 77% for 8-in. While these are impressive numbers, one factor we haven’t taken into account in our LED chip simulation is LED chip yield. We will look at this important factor next.

Higher yield

Each step in the manufacturing process of an LED chip has a yield loss, from the preparation of substrates through chip packaging. The yield losses at each step add up and contribute to a significant portion of the final chip cost. There is therefore a great deal of focus currently on improving yield in all of these areas.

Switching to large-diameter LED manufacturing has been linked to yield improvement in a number of stages of the manufacturing process. The potential benefits come both directly as the larger wafer size is a more uniform surface for epitaxy and indirectly through the use of better manufacturing equipment and techniques. While yield can be a complex subject, we will briefly look at some highlights of the possible benefits.

Fig. 4.

One of direct yield benefits of larger wafers comes during epitaxy. In the MOCVD chamber, any physical disturbances, such as wafer edges, can disturb the gas flow and reduce yield. Larger wafers can help here because there are fewer edges and more undisturbed surface area. The resulting higher yield during this expensive step is an important advantage.

The second component of improved yield comes from access to modern process control and automation tools, which are designed around large-diameter wafers and have been perfected in IC manufacturing. Today’s commonly used small-diameter manufacturing techniques use manual processes, requiring many human interactions, and lack sophisticated tracking that could spot yield issues. Many experts have pointed to a general need to move from a research-style production environment to a true mass production environment. Let’s look in a little more detail at what this means.

Automation primarily refers to the use of machines to handle and transport wafers – removing the human element. Wafers can be moved faster and with less damage through automated machines instead of being hand carried. The benefit is a combination of fewer skilled operators required, less loss of wafers due to mishandling, and quicker movement through the manufacturing steps.

In addition to more automation, the use of more modern tools brings better process control. Process control is the use of data analysis to detect and predict problems that cause yield losses in any area of the production process. This involves a tracking and analysis of the substrate throughout the process, extending back to the crystal growth stage. Process control also takes into account the analysis data recorded by the various production tools.

The use of process control is often cited as a necessary step in advancing the LED industry. As with automation, the tools needed for implementing process control are designed around large-diameter substrates, so the benefit to switching goes beyond just more LED chips. For additional information on the industry’s move to improve yield through large-diameter wafer production, see the article “LED wafer and automation standards are on the fast track, ready for more industry feedback” (www.ledsmagazine.com/features/8/10/9).

Market conditions

At this point, we’ve seen how switching to large diameter wafers can create more chips per MOCVD run and improve yield in several areas. Yet the industry this year is still predicted to produce over 80% of the LED capacity using small-diameter substrates. Why? The reasons come from two factors: difficult market conditions and technological challenges in supplying large diameter substrates at a competitive cost.

Fig. 5.

The price of 2-, 3-, and 4-in wafers has dropped dramatically in the past two years due to an oversupply condition and lower-than-expected demand. At the same time, a step in the manufacturing process called PSS (patterned sapphire substrate) has increased the performance of LEDs. These two factors made staying with small diameter an attractive option while waiting for higher demand. However, some large companies moved ahead despite these conditions and are today prepared for a rapid increase in production as demand grows.

Technology barriers

The second barrier to the adoption of large diameter substrates is a group of technological hurdles. It begins with challenges in sapphire crystal growth – the first step in creating substrates. As you can see in Fig. 5, today’s sapphire for HB-LEDs is typically grown on the a-axis, even though HB-LEDs require c-axis wafers. To get c-axis wafers from a-axis sapphire boules, a core must be taken sideways – wasting a large part of the sapphire.

Today’s a-axis sapphire growth technologies also result in defects that cannot be avoided when coring for large diameter applications. The volume of a 6- or 8-in core is so large that the defects become unavoidable and the cores must be shortened or scrapped. The total losses from sideways coring and defects waste over 80% of the material. For 8-in applications, the waste is over 90% and the production costs double.

Another barrier caused by a-axis growth is that the resulting wafers have a variation in stress and strain across their surface. Because the wafer is from a sideways core of the boule, and the boule is grown along the a-axis, the wafer itself has a long growth time signature across its surface. This becomes significant during epitaxial growth when the wafer is heated.

Fig. 6.

As you can see in the Fig. 6, the wafer will bow in an uneven pattern or a warp. This warping is very difficult to counteract by the MOCVD engineer and has caused several attempted workarounds, including a move to thicker wafers and the use of stress-relieving layers. These techniques add to the production cost and complexity. Without counteracting the warp, the result during epitaxy is lower LED chip yield.

The last technology barrier is in the slicing and polishing of the wafer and application of PSS techniques. Slicing and polishing are difficult processes and must be done well to generate good yield during epitaxy. Because the large wafers are 9-16× larger, the difficulty increases significantly.

PSS application faces a similar challenge, with the additional obstacle that the pattern can only be seamlessly applied to a limited size area smaller than 6 in. In order to get PSS on a 6-in wafer, a stepper (a semiconductor manufacturing tool) must apply multiple patterns, which is common in the silicon industry. However, for HB-LEDs the edges of the multiple pattern applications must be closely matched or LED yield will drop. This accuracy requirement is proving to be very challenging.

These barriers of market conditions and technology challenges have created multiple limitations for the mass adoption of large-diameter substrates. In the future, the market demand will require the throughput and yield only available using larger substrates, and as a result the technology challenges will be overcome. As proof of the possibility of overcoming the challenges, several tier-1 manufacturers have already made the switch and are positioned with an advantage over the majority of the industry.

Potential substrate disruptions

It is certain that HB-LED manufacturing will ultimately move to large-diameter substrates. The question is how quickly, and what material will be used for the substrates? In this last section we’ll briefly look at potential answers to these questions.

Fig. 7.

As we said at the beginning, adoption of large-diameter is currently low and predicted to take years to complete. However, advances in alternative substrates to traditional sapphire may accelerate this adoption or carve out niche channels for some companies. But first we’ll focus on sapphire.

The main challenges we noted were the low material utilization due to a-axis growth and the high level of defects that make larger wafers expensive. There are alternative growth technologies that can grow directly on the c-axis for much lower waste. In addition, growth technologies that avoid significant defects are also available.

Sapphire grown with these characteristics of c-axis growth (also called on-axis growth for LED applications) and low defect levels are very well optimized for large-diameter applications. As you can see in Fig. 7 depicting c-axis CHES technology, the problems of low material utilization and high defect levels are both solved at once, with the additional benefit of a near net shape boule. The result is over 75% utilization for both 6- and 8-in applications.

In addition, the problems of warp during epitaxy that we saw from a-axis-grown sapphire are reduced as the c-axis CHES wafers are grown with a single time signature across their surface (Fig. 8). Because of these advantages, it is expected that as more manufacturers move to large-diameter applications, the growth technologies will also transition to c-axis, low-defect-level growth.

Fig. 8.

Alternative substrates to sapphire, such as silicon, silicon carbide, and gallium nitride (GaN) are also being researched. A small number of LED manufacturers are even in production on each of these substrates, yet not as a cost-effective alternative to sapphire. Each of these alternatives has certain advantages over sapphire, yet multiple breakthroughs are needed for one of them to significantly displace sapphire. Of the alternative substrates, current predictions give silicon the best chance for success.

Because LEDs have such a broad potential market, there will be room for these alternative substrates along with sapphire remaining predominant. For example, an advantage GaN substrate brings is higher performance per chip – albeit at very high cost. This substrate may find a niche where a single bright LED chip is desired or required.

The next diameter past 6-in is the 8-in wafer. These wafers give another dramatic gain in LED chip count and further opportunities for yield improvement. Yet the barriers we examined earlier are the same, with the addition of the sapphire substrate costs doubling over 6-in using a-axis growth methods. Therefore the prediction is that one of these alternative technologies (sapphire grown on c-axis, silicon, or another substrate) will become dominant for 8-in wafers and beyond.

Obstacles and benefits

In summary, we found a move to large-diameter-based LED manufacturing provides a 55% increase in LED chips per MOCVD run using 6-in wafers, and a 77% using 8-in wafers. In addition to more chips, the yield would increase throughout the manufacturing process due to better epitaxy yield, automation, and process control.

The reason these advantages haven’t become common except with the largest LED manufacturers are several market and technology barriers. These include a-axis sapphire growth technology, a depressed market, and the use of PSS. But new c-axis-growth technology provides an optimized path to supplying large-diameter sapphire substrates. Other substrate materials, such as silicon, will likely find niche uses with some manufacturers. The advances of larger wafers will continue to the next step of 8-in.

Large diameter has already been proven by large tier-1 manufacturers as an important component in reducing costs and increasing performance of HB-LEDs. See “Philips Lumileds announces workhorse Luxeon T LED family” (www.ledsmagazine.com/news/9/12/7) for an example.

Yet many companies are staying with small diameter wafers until the next demand wave comes. However, companies that are planning ahead to gain a competitive edge – as is possible with large-diameter LED manufacturing – will be more efficient, more flexible to meet demand, and find success in the future.