Portable Mass spectrometers from Engineers (Part 1)

Over the last few months I have talked to engineers at Waters, Thermo Fisher Scientific, Purspec Technologies and Microsaic about their mass spectrometers, as well as an engineer who repairs them at Thorpe Scientific. Here are some of the things I have learnt.

Team

To develop new mass spectrometers, you require a team of physicists, electrical engineers, mechanical engineers and then analytical chemists. The physicists model the electric fields, the engineers build the instruments, and the analytical chemist verifies and compares the results.

Thermos Fisher Project Manager Stellar MS

Questions

How long does a typical development cycle take, from initial conception to release? What are the most challenging parts of development / what fails when you move from the lab to production? What areas would you recommend focusing on for someone interested in potentially building instruments?

Response

I’m not sure of the development cycle for a new product, but the cycle for the next iteration of existing products (e.g. a triple quadrupole) is around 5 years.

Most challenging aspect is maintaining component quality when moving from R&D to the manufacturing floor, which rely on 3^rd party manufacturers.  If I was starting from scratch, would try to use 3D printing for all components whenever possible.  This allows tighter control on performance specific specifications.  There are some components that have very tight tolerances for some aspects, but not all aspects.  For example, a lens may have very high tolerance on the diameter of the aperture, but lower tolerance on how thick the lens is. If you don’t know what the important parameters are, then defining tolerances for 3^rd party manufacturers become impossible.  Too tight a tolerance makes a component more expensive than it needs to be.  Too low a tolerance and you’ll see big differences in performance from one instrument to another.  Making changes and ensuring consistency is much easier and quicker with 3D printers.

Another aspect is around patents.  Both patenting new innovations (last release of instrument I was involved in had 20 novel patents).  But, also to make sure that any innovations you have do not infringe existing patents.  There are a lot of patents filed, that never even make it into a commercial instrument.  But the patent is still active.

Focus on software.  The hardware for mass specs is quite mature now.  But the users tend to have less and less experience and often no interest in the detector at all.  They just want the data.  So, anything that makes acquiring, processing or reporting the data easier.

An analogy is with cars.  No-one really strips down an engine down now.  Last time I took my wife’s car to the garage, the first thing the mechanic did was plug it into a laptop to access software. 

Even when driving, the first thing people do is connect their phone to the car.  It’s an old adage, but you need to focus on what your potential customers need to satisfy their customers.  For mass specs, this is nearly always to be able to generate data and reports easily.

Questions

Is it difficult for new companies to enter MS hardware space simply because of the existing IP and the physics / engineering is mature? Are current instruments designed more for their performance vs ease of assembly vs maintenance?

 Response

Those are three very good points.  But is very dependent on which market you are focusing on. Academic research and historically pharma companies, the emphasis was always on performance, which is often focused on sensitivity.

Ease of assembly helps reduce cost.  So that makes the manufacturer more profitable and/or means they can be more competitive in the market.  And everyone loves anything that is perceived as a bargain.

For non-academic users, the mass spec system is mainly a source of revenue.  It offers a chargeable service.  So, the longer it can operate without downtime, and if it does need to be cleaned or parts replaced, the quicker it can operational again the better.  That can be a significant plus for some customers. 

Of growing importance is environmental impact.  If you can show you use less energy, generate less heat that is increasingly being part of the scoring for public tenders.  I know of more than one lab, where the limitation on number of instruments they can have in a lab, is not based on space they occupy, but whether the air conditioning can cope with the amount of heat generated.

Could you make components out of recyclable materials.  Can you reduce the weight (this reduces carbon footprint of transportation).

New companies have to offer something different to the other vendors.  What is your unique selling point that is not served by existing vendors?  Many scientists are very conservative in my experience.  They will put up with instruments that don’t do the best job, just because they are familiar with them and may have spent a lot of time and resource, finding a workflow that overcomes the limitation of their hardware.  Show them an easier solution, to something they have overcome, doesn’t always cut through.  But if you can do something easily on your new system, that is very time consuming on their current platform you might be able to interest them.  Whether that be removing the need to do extensive sample prep, or something that makes data processing easier and therefore producing a report quicker.

Thorpe Scientific  (Repairs Quadrupoles MS)

The weakest part of the instrument tends to be the turbomolecular pump, these seem to be the main reason that the instruments fail. They are costly when brand new, but like car depreciate quickly in value with time. The same goes for the instruments.

In terms of cost, most of the parts are not as expensive as you’d think. But there’s significant mark up from manufacture to distribution.

Costs of Instruments

After speaking to sales people at Microsaic and Purspec I have verified the cost of two of their products. The Cell miniature from Purspec and quad MS from Microsaic

Microsaic costs £80,000 including softwarePurspec £100,000 including software

I will need to verify waters ASAP MS as well as Advion quadrupole MS for further comparison of costs.

This will be updated when I have more information